Observations of a Naturalist in the Pacific Between 1896 and 1899, v. 2

By Author	[ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z \| Other Symbols ]
By Title	[ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z \| Other Symbols ]
By Language

Download this book: [ ASCII ]

Look for this book on Amazon

We have new books nearly every day.
If you would like a news letter once a week or once a month fill out this form and we will give you a summary of the books for that week or month by email.

Title: Observations of a Naturalist in the Pacific Between 1896 and 1899, v. 2 - Plant-Dispersal
Author: Guppy, H. B. (Henry Brougham)
Language: English
As this book started as an ASCII text book there are no pictures available.

*** Start of this LibraryBlog Digital Book "Observations of a Naturalist in the Pacific Between 1896 and 1899, v. 2 - Plant-Dispersal" ***

Transcriber’s Note

When italics were used in the original book, the corresponding text has
been surrounded by _underscores_; bold or larger characters (such as the
large numbers in the index) have been surrounded by =equal signs=.
Superscripted characters are preceded by ^. Mixed fractions have been
displayed with a hyphen between whole number and fraction for clarity.
When a fractional number has been used in a range (i.e. one and a half
to two and a half), the dash separating the two numbers has been spaced
(i.e. 1-1/2 - 2-1/2).

Some corrections have been made to the printed text. These are listed in
a second transcriber’s note at the end of the text.

OBSERVATIONS OF A NATURALIST IN THE

PACIFIC BETWEEN 1896 AND 1899

[Illustration]

[_Frontispiece._

[Illustration: Selala (Vanua Levu, Fiji) (1/3).]

[Illustration: Fruits with seedlings nearly ready to fall from the tree
of Rhizophora mangle (shortest) and Rhizophora mucronata (longest). From
Vanua Levu, Fiji. (1/4 of the true length.)]

[Illustration: Rhizophora mangle (Vanua Levu, Fiji) (1/4).]

[Illustration: Rhizophora mucronata (Vanua Levu, Fiji) (1/5).]

OBSERVATIONS OF
A NATURALIST IN
THE PACIFIC BETWEEN
1896 AND 1899

BY
H. B. GUPPY, M.B., F.R.S.E.

VOLUME II
_PLANT-DISPERSAL_

London
MACMILLAN AND CO., LIMITED
NEW YORK: THE MACMILLAN COMPANY
1906

_All rights reserved_

RICHARD CLAY AND SONS, LIMITED,

BREAD STREET HILL, E.C., AND
BUNGAY, SUFFOLK.

Dedication

TO THOSE NUMEROUS PERSONS TO WHOM I WAS INDEBTED FOR
GREAT KINDNESS AND ASSISTANCE DURING MY
SOJOURN IN HAWAII AND FIJI

PREFACE

ALTHOUGH this volume contains a great amount of original material, I am
largely indebted to the labours of my predecessors for its present form;
and a scheme that at first was limited only to my own observations in
the Pacific has gradually extended itself to the general subject of
plant-dispersal. The farther I proceeded in my work the more I realised
that the floras of the Pacific islands are of most interest in their
connections, and that the problems affecting them are problems
concerning the whole plant-world. Deprived of the writings of Seemann,
Hillebrand, Drake del Castillo, and other botanists, several of whom
have lived and died in the midst of their studies of these floras, and
without the aid of the works of Hemsley and Schimper, generalisers who
have mainly cleared the way for the systematic study of
plant-distribution and plant-dispersal, it would not have been possible
for me to accomplish such an undertaking.

My interest in plant-dispersal dates back to 1884, when, whilst surgeon
of H.M.S. _Lark_, in the Solomon Islands, I made some observations on
the stocking of a coral island with its plants, which were published in
the _Report on the Botany of the “Challenger” Expedition_. In 1888 I
followed up the same line of investigation during a sojourn of three
months on Keeling Atoll, and during a journey along the coasts of West
Java. But realising that as yet I had barely touched the fringe of a
great subject, and that several years of study would be required before
one could venture even to appreciate the nature of the problems involved
and much less to weigh results, I took advantage of the circumstances of
my life to make, between the years 1890 and 1896, a prolonged
investigation of the plants of the British flora, mainly from the
standpoint of dispersal by water. This involved the study of the
seed-drift of ponds and rivers and of the plants supplying it, a study
which brought me into close relation with aquatic and sub-aquatic
plants. This line of investigation led me into contact with many other
aspects of plant-life; and as time went on my field of interest extended
to the plants of dry stations and to the bird as an agent in
plant-dispersal. Only a few of these results have been published, as in
the journals of the Linnean Society and of the Royal Physical Society of
Edinburgh as well as in the pages of _Science Gossip_. They lie for the
most part still within my note-books, and fitly so, since I regarded
such studies chiefly as a preparation for the investigation of the
general question of plant-dispersal.

When again, in October, 1896, I found myself once more in the Pacific,
the subject was taken up again with zeal; but my larger experience had
only increased my diffidence, and the unknown looked so overwhelming
that I settled down for the next three years content with merely making
experiments and recording observations. Here again the main problem was
attacked through the study of seed-buoyancy, and gradually it led me to
the systematic study of the mangroves and of the beach-plants, whilst my
inland excursions brought me into familiarity with the plants of the
interior. My geological exploration of the island of Vanua Levu, in
Fiji, greatly assisted me by giving a method to my botanical examination
of the island.

Whilst working out my geological collections in England, in the years
1900-1902, I devoted an hour or two daily to the elaboration of my
botanical notes and to a consideration of the problems concerned. During
a winter in Sicily I took up again the subject of the beach-plants; and
after the publication of the volume on the geology of Vanua Levu I was
able to accomplish a plan, for years in my dreams, of visiting the
eastern shores of the Pacific. During a period of three months from
December, 1903, to March, 1904, I examined the littoral flora of the
west side of South America at various localities between Southern Chile
and Ecuador; and finally completed this investigation by comparing the
shore-plants on the Pacific and Atlantic coasts of the isthmus of
Panama. Returning to England with a fresh collection of data, I passed
many months in elaborating and arranging all my notes, waiting vainly
for a clue to guide me in framing a scheme by which I could bring the
results of many years of work into some connected form. At last I
decided once again to take the floating seed as my clue, and without any
prearranged plan I allowed the work to evolve itself. Now that it is
finished, I can see some obvious defects; but if any other plan had been
adopted I scarcely think that I should have been more successful in
piecing together in a single argument materials resulting from so many
years of research and relating to so many aspects of plant-life.

Yet the final object of a naturalist would be but a sorry one, if his
aim were only to write a treatise and append his name to it. His
personal faith lies behind all his work; and no one can pursue a long
line of study of the world around him without rising from his task with
some convictions gained and some convictions lost.

As far as the observation of Nature’s processes at present in operation
can guide us, the world presents itself to us only as a differentiating
world. We can perceive, it is true, a progressive arrangement of types
of organisms from the lowest to the highest, and we can perceive a
development of varieties of the several types; but the only process
evident to our observation is that concerned with the production of
varieties of the type. Nature does not enlighten us as to the mode of
development of the type itself. We can, for instance, detect in actual
operation the process by which the different kinds of bats or the
different kinds of men have been developed; but there is no principle in
Nature evident to our senses that is concerned with type-creation.
Though we can supply it by hypothesis, we cannot discover it in fact. On
the other hand, the evidence of differentiation is abundant on all sides
of us, both in the organic and in the inorganic worlds. The history of
the globe has ever proceeded from the uniform to the complex; and in the
closing chapter of this book an endeavour is made to connect the
differentiation of plant and bird with the differentiation of the
conditions of existence on the earth. But this leaves no room for the
development of new types of organisms; and so far as observation of the
processes of Nature at present working around us can guide us, each type
might well be regarded as eternal. We can never hope to arrive at an
explanation of the progressive development of types by studying the
differentiating process; and since the last is alone cognisable for us,
evolution, as it is usually termed, becomes an article of our faith, and
of faith only.

In illustration of this argument, let me take the case of the races of
men. We see mankind in our own day illustrating the law of
differentiation all over the globe, as far as physical characters are
concerned. Just as the ornithologist would postulate a generalised type
in tracing the origin of various allied groups of birds, so the
anthropologist, guided by his observation of the changes now offered by
man in different regions, would postulate a generalised original type as
the parent-stock of mankind. Observation of the processes of change now
in operation by no means leads us to infer that such a generalised type
was an anthropoid ape, or even simian in character. In so doing we
should be forming a conclusion not warranted by the observation of
existing agencies of change, and we should be confusing the two distinct
processes of evolution and differentiation, or rather of progressive and
divergent evolution, of which the last alone comes within our field of
cognition. The study of variation can do no more than enable us to
ascertain the mode of development of different kinds, we will say, of
birds or of men. The origin of the type lies outside our observation.
“Given the type, to explain its origin”: this is the problem we can
never solve, and Nature aids us nothing by the study of her ways. On the
other hand, there is the subsidiary problem.... “Given a type, to
explain its varieties” ...; and here Nature’s processes are apparent to
us in a thousand different shapes.

It might seem that the presumptive evidence connecting man in his origin
with the monkeys is so strong that, supposing his simian descent were
regarded as a crime, a jury would without hesitation pronounce his
guilt; but until some observer of the processes followed by Nature can
bridge over the gap that divides man from the ape, until indeed he can
offer a legitimate illustration of how it is accomplished in similar
cases in our own day, the gap remains. Those who have read the recent
work of Prof. Metchnikoff on the Nature of Man will properly regard his
chapter on the simian origin of man as a brilliant argument advanced by
a most competent authority. Yet he fails to complete his case by
bridging over this gap, and can only appeal to the results of the now
famous researches of De Vries concerning the mutations of the evening
primrose (Œnothera). It is probable, he says, that man owes his origin
to a similar phenomenon (English edition, p. 57). Several objections
could be raised against this illustration from the plant-world, the most
important of them lying in the circumstance that these mutations could
only be urged as instances of the sudden development of new species of
the evening primrose type. They merely illustrate the process of
differentiation from a given type, and by no means represent the process
of progressive evolution from a simian to a man.

However, look where we may—and this is the great lesson I have learned
from my researches in the Pacific islands—Nature does not present to our
observation any process in operation by which a new type of organism is
produced. The processes involved lie hidden from our view. The channels
by which impressions from the outside world reach us are comparatively
few; and although it seems likely that the future development of man
will be mainly concerned with the acquirement of additional
sense-channels, no newly acquired sense will enable him to be at once an
actor in and a spectator of the great drama presented in the organic
world. That a creature should be able to get at the back of its own
existence, or, in other words, to penetrate the secret of its own
creation, is unthinkable. Outside the limited field of observation that
immediately surrounds us extends the region where reason alone can guide
us, and beyond lies the realm where reason fails and faith begins.

H. B. GUPPY.

_November 8th, 1905._

LIST OF SOME OF THE PRINCIPAL AUTHORITIES QUOTED IN THIS VOLUME,
WITH AN ENUMERATION OF THE AUTHOR’S BOTANICAL PAPERS

BURKILL, I. H., “The Flora of Vavau, one of the Tonga Islands,” Journal
of the Linnean Society, vol. xxxv., Botany, 1901.

CHEESEMAN, T. F., “The Flora of Rarotonga,” Transactions of the Linnean
Society, 2nd Ser., Botany, vol. vi., part 6, 1903.

DRAKE DEL CASTILLO, E., “Flore de la Polynésie Française,” Paris, 1893.

“Remarques sur la Flore de la Polynésie” (Mémoire couronné par
l’Académie des Sciences), Paris, 1890.

EGGERS, BARON H. VON, “Die Manglares in Ecuador,” Botanisches
Centralblatt, No. 41, 1892.

“Das Küstengebiet von Ecuador,” Deutsche Geographische Blätter, heft
4, band 17, Bremen, 1894.

EKSTAM, O., “Einige blütenbiologische Beobachtungen auf Novaja Semlja,”
Tromso Museums Aarshefter, 18, 1895.

“Einige blütenbiologische Beobachtungen auf Spitzbergen,” Tromso
Museums Aarshefter, 20, 1897.

GUPPY, H. B., “The Dispersal of Plants as illustrated by the Flora of
Keeling Atoll,” Journal of the Victoria Institute, London, 1889.

“The Polynesians and their Plant-Names,” Journal of the Victoria
Institute, London, 1896.

“The River Thames as an Agent in Plant-Dispersal,” Journal of the
Linnean Society, Botany, vol. xxix., 1891-93.

“River-Temperature,” part iii., Proceedings of the Royal Physical
Society of Edinburgh, 1896. (The first two parts deal principally with
the temperature of ponds and rivers, whilst in the last part the
thermal conditions are discussed especially in connection with the
life of aquatic plants.)

“On the Postponement of the Germination of the Seeds of Aquatic
Plants,” Proceedings of the Royal Physical Society of Edinburgh, 1897.

“On the Temperature of Springs as especially illustrated by the
Wandle and other Tributaries of the Thames.” (This paper, of which I
have no copy, was published in the Journal of the Royal Meteorological
Society, about 1895. It throws light on the thermal conditions of
plants in springs.)

“Water-Plants and their Ways,” _Science Gossip_, Sept., Oct., Nov.,
1894. (The various modes of dispersal of land as well as water plants
are here dealt with, their thermal conditions are discussed, and in
the November number are given the results of four years’ observations
on the life-history and life-conditions of Ceratophyllum demersum.)

“Caddis-Worms and Duckweed,” _Science Gossip_, March, 1895. (A short
note.)

“Stations of Plants and Buoyancy of Seeds,” _Science Gossip_, April
and May, 1895.

“Irregularity of some Cotyledons,” _Science Gossip_, September,
1895.

“Plants of the Black Pond, Oxshott,” _Science Gossip_, October,
1895.

“On the Habits of Lemna minor, L. gibba, and L. polyrrhiza,” Journal
of the Linnean Society, Botany, vol. xxx. (This paper contains the
results of three years’ systematic observations of these plants.)

“The Distribution of Aquatic Plants and Animals,” _The Scottish
Geographical Magazine_, January, 1893.

HEMSLEY, W. B., “Report on the Botany of the _Challenger_ Expedition,”
1885.

“The Flora of the Tonga Islands,” Journal of the Linnean Society,
Botany, vol. xxx.

HILLEBRAND, W., “Flora of the Hawaiian Islands,” Heidelberg, 1888.

HORNE, J., “A Year in Fiji,” London, 1881.

KOLPIN RAVN, F., “Om Flydeevnen hos Froene af vore Vand-og Sumpplanter,”
Botanisk Tidsskrift, 19 bind., 2 hefte, Kjobenhavn, 1894 (“On the
Floating Capacity of the Seeds of Aquatic and Marsh Plants”). (A
_résumé_ in French is appended to the paper.)

MARTINS, CH., “Expériences sur la Persistance de la Vitalité des Graines
flottant à la Surface de la Mer,” Bull. Soc. Botanique de France tome
iv., 1857.

NADEAUD, J., “Enumération des Plantes indigènes de l’Ile de Tahiti,”
Paris, 1873.

PENZIG, O., “Die Fortschritte der Flora des Krakatau,” Annales du Jardin
Botanique de Buitenzorg, 2 ser., tome 3, Leide, 1902.

PERKINS, R. C. L., “Fauna Hawaiiensis,” vol. i., part iv. (Vertebrata)
Cambridge University Press, 1903.

REINECKE, F., “Die Flora der Samoa-Inseln,” Engler’s “Botanische
Jahrbücher,” band xxv., heft v., Leipzig, 1898.

SCHIMPER, A. F. W., “Die indo-malayische Strandflora,” Jena, 1891.

SEEMANN, B., “Flora Vitiensis,” London, 1865-73.

SERNANDER, R., “Den Scandinaviska Vegetationens Spridnings-biologi,”
Upsala, 1901.

THURET, G., “Expériences sur des Graines de diverses Espèces plongées
dans de l’eau de Mer,” Archives des Sciences (Phys. et Nat.) de la
Bibliothèque Universelle, tome 47, Geneva, 1873.

TREUB, M., “Notice sur la nouvelle Flore de Krakatau,” Annales du Jardin
Botanique de Buitenzorg, 1888.

_Note._—Amongst the works quoted which are not specially particularised
in the text are Scott Elliot’s “Nature Studies,” 1902, and Beal’s
“Seed Dispersal,” Boston, 1900.

CONTENTS

PREFACE _Pages_
vii—x

LIST OF SOME OF THE PRINCIPAL AUTHORITIES QUOTED, _Pages_
WITH AN ENUMERATION OF THE AUTHOR’S BOTANICAL xiii—xv
PAPERS

LIST OF ILLUSTRATIONS _Page_
xxvii

ADDITIONS AND CORRECTIONS _Page_
xxviii

CHAPTER I

INTRODUCTION

The study of insular floras.—Their investigation in this work from the
standpoint of dispersal.—The significance of plant-distribution in the
Pacific.—The problems connected with the mountain-flora of Hawaii.—The
persistence of dispersing agencies at the coast, their partial
suspension on the mountain-top, their more or less complete suspension
in the forest, and the effect on the endemic character of plants.—The
connection between the endemism of birds and plants.—The relative
antiquity of plants of the coast, forest, and mountain-top.—The
genetic relation between coast and inland species of the same
genus.—The ethics of plant-dispersal.—Evolution takes no heed of modes
of dispersal.—The seed-stage is the price of Adaptation.

_Pages_ 1-11

CHAPTER II

THE FLORAS OF THE PACIFIC ISLANDS FROM THE STANDPOINT OF DISPERSAL
BY CURRENTS

The initial experiment.—The proportion of littoral plants.—The two great
principles of buoyancy.—The investigations of Professor Schimper.—The
investigations of the author.—The great sorting process of the
ages.—Preliminary results of the inquiry into the buoyancy of seeds
and fruits.

_Pages_ 12-22

CHAPTER III

THE LESSON OF THE BRITISH FLORA

Results of observations on the buoyancy of over 300 British plants.—The
small proportion of plants with buoyant seeds or seedvessels.—Their
station by the water-side.—The great sifting experiment of the
ages.—Summary.

_Pages_ 23-30

CHAPTER IV

THE LESSON OF THE BRITISH FLORA (_continued_)

The choice of station of the water-side plant possessing buoyant seeds
or seedvessels.—Determined by its fitness or unfitness for living in
physiologically dry stations.—In the internal organisation of a plant
lies the first determining influence of station.—The grouping of the
British strand-plants.—Whilst the Xerophyte with buoyant seed or fruit
finds its station at the coast, the Hygrophyte similarly endowed makes
its home at the river or pond side.—The grouping of the plants of the
river and the pond.—Summary.

_Pages_ 31-39

CHAPTER V

THE FIJIAN STRAND-FLORA

The inland extension of the beach-plants.—The grouping of the
coast-plants.—Their modes of dispersal.—The zone of change.—Summary.

_Pages_ 40-46

CHAPTER VI

THE TAHITIAN STRAND-FLORA

(_From materials supplied mainly by the work of Drake del Castillo_)

Lacks the mangroves and their associated plants.—Possesses mainly the
plants of the coral beach.—Predominant agency of the currents.—Inland
extension of shore-plants.—Summary

_Pages_ 47-50

CHAPTER VII

THE HAWAIIAN STRAND-FLORA

Its poverty.—Its negative features.—Their explanation.—The subordinate
part taken by the currents.—The Oregon drift.—The inland extension of
the beach-plants.—Summary

_Pages_ 51-60

CHAPTER VIII

THE LITTORAL PLANTS AND THE CURRENTS OF THE PACIFIC

The working value of the currents as plant-dispersers.—The relation
between the currents and the distribution of shore-plants.—The clue
afforded by the American plants.—Two regions of tropical shore-plants,
the American and the Asiatic.—America, the home of the cosmopolitan
tropical shore-plants that are dispersed by the currents.—Hawaii and
the currents.—Summary

_Pages_ 61-75

CHAPTER IX

THE GERMINATION OF FLOATING SEEDS

Germination in the floating seed-drift of tropical estuaries.—A strain
of vivipary.—Abortive germination of seeds in warm seas.—A barrier to
plant dispersal.—The borderland of vivipary.—Summary

_Pages_ 76-87

CHAPTER X

THE RELATION OF THE BUOYANCY OF SEEDS AND SEEDVESSELS TO THE DENSITY
OF SEA-WATER

The general principles concerned.—The subject assumes a statistical
character.—Seeds and seedvessels are as a rule either much heavier
than sea-water or much lighter than fresh water.—The present littoral
plants with buoyant seeds or seedvessels could be equally well
dispersed by currents in oceans of fresh water.—Seed-buoyancy has no
relation either in the present or in the past to the density of the
sea.—Though an accidental attribute, the specific weight of seeds has
had a profound influence on plant-distribution.—Summary

_Pages_ 88-98

CHAPTER XI

ADAPTATION AND MEANS OF DISPERSAL

Nature has never concerned herself directly with providing means of
dispersal.—Fleshy fruits not made to be eaten.—Nor “sticky” seeds to
adhere to plumage.—Nor prickly fruits to entangle themselves in fur
and feathers.—The dispersal of seeds a blind result of the struggle
between the intruding Evolutionary power and the controlling influence
of Adaptation.

_Pages_ 99-103

CHAPTER XII

THE CAUSES OF THE BUOYANCY OF SEEDS AND FRUITS OF LITTORAL PLANTS,
WITH ESPECIAL REFERENCE TO THOSE OF THE PACIFIC ISLANDS

The classification of buoyant seeds and fruits.—The first group, where
the cavity of the seed or seedvessel is incompletely filled.—The
second group, where the kernel is buoyant.—The third group, where
there is air-bearing tissue in the seed-tests or fruit-coats.—The
buoyant seeds and seedvessels of the littoral plants of the British
flora.—Summary

_Pages_ 104-118

CHAPTER XIII

ADAPTATION AND SEED-BUOYANCY

The question of the operation of Natural Selection.—Are there two
principles at work?—The presence of buoyant tissue in the seed-tests
and fruit-coats of inland plants, both wild and cultivated.—Useless
buoyancy.—The buoyancy of seeds and fruits is not concerned with
Adaptation.—Summary.

_Pages_ 119-129

CHAPTER XIV

THE RELATION BETWEEN LITTORAL AND INLAND PLANTS

Professor Schimper’s views.—Great antiquity of the
mangrove-formation.—Problem mainly concerned with the
derivation of inland from littoral plants.—Grouping of
the genera possessing both coast and inland
species.—Scævola.—Morinda.—Calophyllum.—Colubrina.—Tacca.—Vigna.—Premna

_Pages_ 130-139

CHAPTER XV

THE RELATION BETWEEN LITTORAL AND INLAND PLANTS (_continued_)

Inland species of a genus developed from littoral species originally
brought by the currents but no longer existing in the
group.—Illustrated by the Leguminous genera Erythrina, Canavalia,
Mezoneuron, and Sophora, and by the Apocynaceous genus Ochrosia.—The
Hawaiian difficulty.

_Pages_ 140-154

CHAPTER XVI

THE RELATION BETWEEN LITTORAL AND INLAND PLANTS (_continued_)

The Fijian difficulty.—Inland species of a genus possessing
fruits not known to have any means of dispersal through
agencies now at work in the Pacific.—Pandanus.—Its remarkable
distribution in oceanic groups.—To be attributed perhaps
to extinct Columbæ or extinct Struthious
birds.—Barringtonia.—Guettarda.—Eugenia.—Drymispermum.—Acacia
laurifolia.—Conclusions to be drawn from the discussion.—Summary of
Chapters XIV., XV., XVI.

_Pages_ 155-169

CHAPTER XVII

THE STORIES OF AFZELIA BIJUGA, ENTADA SCANDENS, AND CÆSALPINIA
BONDUCELLA

Afzelia bijuga.—The African home of the genus.—The double station of
Afzelia bijuga, inland and at the coast.—The nature of the buoyancy
of its seeds.—Summary relating to Afzelia bijuga.—Entada
scandens.—Its station and distribution.—Darwin’s opinion of the
plant.—The dispersal of its seeds by the currents.—Summary relating
to the plant.—Cæsalpinia bonducella and C. bonduc.—Their station and
distribution.—Their characters in various Pacific groups.—The
parents of inland species.—Their dispersal by the currents.—The
germination of their seeds.—A dream of vivipary.—The causes of the
seed-buoyancy.—Summary of results

_Pages_ 170-197

CHAPTER XVIII

THE ENIGMAS OF THE LEGUMINOSÆ OF THE PACIFIC ISLANDS

Leguminosæ predominate in tropical littoral floras.—The anomalies of
their distribution in the Pacific islands.—They conform to no one rule
of dispersal or of distribution.—Strangers to their stations.—The
American home of most of the Leguminous littoral plants.—Summary

_Pages_ 198-203

CHAPTER XIX

THE INLAND PLANTS OF THE PACIFIC ISLANDS

PRELIMINARY COMPARISON OF THE PHYSICAL CONDITIONS OF HAWAII,
FIJI, AND TAHITI

Introductory remarks.—The tranquil working of the winds and currents
contrasted with the revolutionary influence of the bird.—The Hawaiian,
Fijian, and Tahitian groups.—Their surface-areas and elevations.—Their
climates.—The mountain climate of Hawaii.—The rainfall of the three
groups.—Summary

_Pages_ 204-219

CHAPTER XX

THE ERAS IN THE FLORAL HISTORY OF THE PACIFIC ISLANDS

THE AGE OF FERNS

The eras in the plant-stocking.—The age of ferns and lycopods.—The
relative proportion of vascular cryptogams in Hawaii, Fiji, and
Tahiti.—The large number of peculiar species in Hawaii.—The mountain
ferns of Hawaii.—The origin of peculiar species.—Dr. Hillebrand’s
views.—Their origin connected not with greater variety of climate in
Hawaii, but with isolation.—Summary

_Pages_ 220-230

CHAPTER XXI

THE ERAS OF THE FLOWERING PLANTS

THE ERA OF THE ENDEMIC GENERA

THE AGE OF COMPOSITÆ

The islands of the tropical Pacific as the homes of new genera and new
species.—The significance of a large endemic element.—Synopsis of the
eras.—The era of endemic genera.—The endemic genera of
Compositæ.—Their affinities and mode of dispersal.—The mystery of the
suspension of the dispersing agencies.—Mr. Bentham’s views.—The
remnant of an ancient Composite flora in the tropical Pacific.—The
dispersion of the Compositæ antedates the emergence of the
island-groups of the Fijian region at the close of the Tertiary
period.—Summary

_Pages_ 231-249

CHAPTER XXII

THE ERA OF THE ENDEMIC GENERA (_continued_)

THE AGE OF THE TREE-LOBELIAS

The distribution of the arborescent Lobeliaceæ.—On the upper flanks of
Ruwenzori.—The Lobeliaceæ of the Hawaiian Islands.—The Lobeliaceæ of
the Tahitian or East Polynesian region.—The capacities for
dispersal.—The explanation of the absence of the early Lobeliaceæ from
West Polynesia.—The other Hawaiian endemic genera.—The Fijian endemic
genera.—Summary

_Pages_ 250-267

CHAPTER XXIII

THE ERA OF THE NON-ENDEMIC GENERA OF FLOWERING PLANTS

THE MOUNTAIN FLORAS OF THE PACIFIC ISLANDS AS ILLUSTRATED BY
THE NON-ENDEMIC GENERA

The mountain-flora of Hawaii.—A third of it derived from high southern
latitudes.—An American element.—Compared with Tahiti and
Fiji.—Capacities for dispersal of the genera possessing only endemic
species.—Acæna, Lagenophora, Plantago, Artemisia, Silene, Vaccinium,
&c.—Capacities for dispersal of the genera possessing non-endemic
species.—Cyathodes, Santalum, Carex, Rhynchospora.—Fragaria chilensis,
Drosera longifolia, Nertera depressa, Luzula campestris.—Summary.

_Pages_ 268-288

CHAPTER XXIV

THE ERA OF THE NON-ENDEMIC GENERA OF FLOWERING PLANTS (_continued_)

THE MOUNTAIN-FLORAS OF THE TAHITIAN AND FIJIAN REGIONS

The mountain-flora of the Tahitian region, as illustrated by the
non-endemic genera.—Derived chiefly from high southern
latitudes.—Weinmannia, Coprosma, Vaccinium, Astelia, Coriaria,
Cyathodes, Nertera depressa, Luzula campestris.—The mountain flora of
Rarotonga.—The mountain-flora of the Fijian region, as illustrated by
the non-endemic genera.—Weinmannia, Lagenophora, Coprosma, Astelia,
Vaccinium, Nertera depressa.—The Fijian Coniferæ.—Dammara, Podocarpus,
Dacrydium.—Not belonging to the present era of dispersal.—The age of
dispersal of the Coniferæ in the Pacific.—Earlier than the age of
Compositæ and Lobeliaceæ.—The first in the Mesozoic period.—The last
in the Tertiary period.—Summary

_Pages_ 289-306

CHAPTER XXV

THE ERA OF THE NON-ENDEMIC GENERA OF FLOWERING PLANTS (_continued_)

THE AGE OF THE MALAYAN PLANTS AS REPRESENTED IN THE LOW-LEVEL FLORA OF
HAWAII AND IN THE BULK OF THE FLORAS OF THE FIJIAN AND TAHITIAN
REGIONS

THE AGE OF WIDE DISPERSAL OVER THE TROPICAL PACIFIC

The widely dispersed genera which possess only peculiar species in
Hawaii.—Pittosporum.—Reynoldsia.—Gardenia.—Psychotria.—Cyrtandra.—Freycinetia.—Sapindus.—Phyllanthus.—Pritchardia.—Summary.

_Pages_ 307-332

CHAPTER XXVI

THE ERA OF THE NON-ENDEMIC GENERA OF FLOWERING PLANTS (_continued_)

THE AGE OF MALAYAN PLANTS (_continued_)

THE AGE OF WIDE DISPERSAL OVER THE TROPICAL PACIFIC (_continued_)

The widely dispersed genera that are as a rule not entirely
represented by endemic species in any
archipelago.—Elæocarpus.—Dodonæa.—Metrosideros.—Alyxia.—Alphitonia.—Pisonia.—Wikstrœmia.—Peperomia.—Eugenia.—Gossypium.—The
last stage in the general dispersal of plants of the Malayan
era as illustrated by the widely-dispersed genera having
as a rule no peculiar
species.—Rhus.—Osteomeles.—Plectronia.—Boerhaavia.—Polygonum.—Pipturus.—Dianella.—Summary.

_Pages_ 333-358

CHAPTER XXVII

THE ERA OF THE NON-ENDEMIC GENERA OF FLOWERING PLANTS (_continued_)

THE AGE OF MALAYAN PLANTS (_continued_)

THE AGE OF LOCAL DISPERSAL

Synopsis of the Chapter given on page 359

_Pages_ 359-410

CHAPTER XXVIII

THE POLYNESIAN AND HIS PLANTS

Identity of the problems presented by the indigenous plants and the
peoples of the Pacific islands.—The food-plants of the Polynesians and
the pre-Polynesians.—Their weeds.—The aboriginal weeds.—The white
man’s weeds.—Weeds follow the cultivator but are distributed by
birds.—The general dispersion of weeds antedates the appearance of the
Polynesian in the Pacific.—Weeds of little value to the
ethnologist.—Aleurites moluccana.—Inocarpus edulis, Gyrocarpus
Jacquini, Serianthes myriadenia, Leucæna Forsteri, Mussænda frondosa,
Luffa insularum.—Summary

_Pages_ 411-428

CHAPTER XXIX

BEACH AND RIVER DRIFT

In the south of England.—On the coast of Scandinavia.—In the
Mediterranean.—Southern Chile.—Very little effective dispersal by
currents in temperate latitudes.—Cakile maritima.—In tropical
regions.—River drift.—River and beach drift of Fiji.—Musa Ensete.—The
coco-nut.—River and beach drift of Hawaii.—Comparison of the beach
drift of the Old and New Worlds.—Summary

_Pages_ 429-439

CHAPTER XXX

THE VIVIPAROUS MANGROVES OF FIJI

RHIZOPHORA AND BRUGUIERA

_Rhizophora._—Represented by Rhizophora mucronata, Rhizophora
mangle, and the Selala, a seedless intermediate form.—Their
mode of association and characters.—The relation of the
Selala.—Polyembryony.—The history of the plant between the
fertilisation of the ovule and the detachment of the seedling.—Absence
of a rest period.—Mode of detachment of the seedling.—Capacity
for dispersal by the currents.—_Bruguiera._—The mode of
dispersal.—Peculiar method of fertilisation.—Length of period between
fertilisation and the detachment of the seedling.—Mode of detachment
of the seedling.—Summary

_Pages_ 440-467

CHAPTER XXXI

A CHAPTER ON VIVIPARY

The significance of vivipary.—The scale of germinative capacity.—A lost
habit with many inland plants.—The views of Goebel.–-The shrinking in
the course of ages of tropical swamp areas.—The variation in the
structures concerned with vivipary.—Abnormal vivipary.—Summary.

_Pages_ 468-473

CHAPTER XXXII

THE WEST COAST OF SOUTH AMERICA

The littoral floras of the West Coast of South America.—The Convolvulus
soldanella zone of Southern Chile.—The plantless or desert zone of
Northern Chile.—The Sesuvium zone of Peru.—The Mangrove zone of
Ecuador and Colombia.—The two varieties of Rhizophora mangle, the
“mangle chico” and the “mangle grande.”—The floating vegetable drift
of the Guayaquil River.—The Humboldt current and the climate of the
West Coast of South America.—The advance northward of the arid
climatic conditions of the Peruvian sea-border.—The retreat of the
mangroves.—Evidence of ancient coral reefs on the coast of Peru.—The
shore plants and stranded seed-drift of the Panama Isthmus.—Summary.

_Pages_ 474-501

CHAPTER XXXIII

SEED-DISPERSAL AND GEOLOGICAL TIME

The shifting of the source of Polynesian plants from the New to the Old
World.—The floral history of Polynesia stated in terms of geological
time.—The suspension of the agencies of dispersal in later
periods.—Parallel differentiation in the course of ages of climate,
bird, and plant.—New Zealand.—Insects and bats as agents in
plant-dispersal.—The effective agency of sea-birds in other
regions.—The observations of Ekstam.—The Spitzbergen controversy.—The
efficacy of ducks as distributors of aquatic plants.—Summary

_Pages_ 502-514

CHAPTER XXXIV

GENERAL ARGUMENT AND CONCLUSION _Pages_ 515-523

APPENDIX _Pages_ 525-605

LIST OF ILLUSTRATIONS

PLATE.

The Fijian species of Rhizophora _Frontispiece._

FIGURES.

TO FACE PAGE

Diagrams illustrating some of the causes of 111
seed-buoyancy

Figures illustrating the development of the seed 452-453
and the germinating process of Rhizophora and
Bruguiera

Diagrams illustrating the structure of the growing 574
seeds of Barringtonia

Diagram illustrating the prevailing 585
cloud-formations of Mauna Loa

MAPS.

Oceania 12

The Ocean Currents 61

Trade routes of the Pacific Ocean (intended to 66
illustrate the distances traversed by floating
seeds in that ocean)

The West Coast of South America 474

Rough plan of the Gulf of Guayaquil 484

ADDITIONS AND CORRECTIONS

Page 5 and subsequent pages. _For_ Ipomea _read_ Ipomœa.

Page 68. _For_ Hippomanes _read_ Hippomane.

Page 68. _For_ Conocarpus erecta _read_ Conocarpus erectus.

Page 122. Sir W. Buller includes the fruits of the Puriri tree (Vitex
littoralis, according to Kirk) amongst the food of the New Zealand
fruit-pigeons.

Page 177. _For_ Entata, in the head-line, _read_ Entada.

Page 266. The fruits of Oncocarpus vitiensis have been found in the
crop of a Fijian fruit-pigeon (Carpophaga latrans). _See_ Hemsley’s
_Bot. Chall. Exped._, Introd., 46, and iv. 308; also Newton’s
_Dictionary of Birds_, p. 724.

Page 368. Sernander (p. 185) observes that the fruits of Naias marina
have little or no floating power.

Page 416. For the first eight lines read as follows:—“Of these, 22
occur in Continental regions on both sides of the Pacific; 12 are
found in the Old World alone; one is peculiarly American, and two are
confined to the Australian and Polynesian regions. A few of these can
be regarded as exclusively American in their origin, though the bulk
of them hail evidently in the first place from the Old World. But
from the circumstance that all or most of the other species of the
genus concerned are confined to America, it may legitimately be
inferred that Waltheria americana, Ageratum conyzoides, and Physalis
angulata are American-born species. Teucrium inflatum is a peculiar
instance of an American weed collected in Polynesia before apparently
it had been recorded from the Old World.”

Page 438. _For_ Conocarpus erecta _read_ Conocarpus erectus.

Page 417. Add after Cardiospermum halicacabum.... “Its seeds, as my
experiments show, possess little or no capacity for dispersal by
currents, since they sink at once or within a few days, even after
drying for months.”

Page 455. Omit the reference to figure 6 in the centre of the page.

Page 498. _For_ Hippomanes _read_ Hippomane.

Page 508. Amongst my Solomon Island collections identified at Kew were
the fruits of a species of Litsea from the crop of a fruit-pigeon
(Hemsley’s _Bot. Chall. Exped._, IV. 295.)

Page 533. _For_ Commelyne _read_ Commelina.

Page 539. At foot of page, _for_ Thames sea-drift, _read_ Thames
seed-drift.

Page 581. _For_ Crambe maritimum _read_ Crambe maritima.

Page 618. Under Mascarene Islands add Myoporum to the plants linking
them to the Pacific Islands.

------------------------------------------------------------------------

OBSERVATIONS OF A NATURALIST
IN THE PACIFIC

CHAPTER I

INTRODUCTION

TO proceed from the general to the special is the only method of dealing
with insular floras. A broad and comprehensive grasp of
plant-distribution, such as is only acquired by a life-time of research
aided by travel and the handling of large collections, is a necessary
foundation for the study; yet in the nature of things such
qualifications can be possessed by but a few. To direct an inquiry in
the opposite direction, and endeavour to attack the problem of
continental floras through the insular floras would result merely in the
investigation of a few of the many questions connected with
plant-distribution.

The panoramic sketch of the surveyor on the mountain-top aids him in a
thousand ways when after months of tedious labour he plots the details
in his chart. Without such a panoramic view of the plant-world in his
mind’s eye, an observer like myself can only look for guidance to the
writings of those who have generalised on the foundations of a far
broader experience, such as those of Bentham, De Candolle, Gray, Hooker,
Schimper, and others.

It would be quite possible for a botanist possessing a profound general
acquaintance with the plant-world to dispense altogether with actual
observation and experiment on modes of dispersal. It would be quite
possible for him to arrive at conclusions, which, even if they did not
always come into line with results of observation and experiment, we
should be compelled to prefer. It is only from his more elevated
position that a general can follow the course of a battle; whilst the
private with his experience confined to a limited area of the field of
conflict may form the most erroneous ideas of the progress of the fight.
So it is with observers whose employment it is to struggle with the
details and secondary principles of plant-distribution, and so it is
with the generaliser who has already roughly mapped out the principal
features of the main problem.

When Mr. Bentham in 1869, remarking on the paucity of species common to
tropical Asia and America, characterised them either as plants wholly or
partially maritime and spread by the currents, or as weeds dispersed by
cultivation over the warm regions of the globe, he mentioned amongst the
plants in the former category, Gyrocarpus jacquini. This tree presents
one of the mysteries connected with the Pacific islands; and I don’t
imagine that this eminent botanist could have known anything except
inferentially as regards the mode of dispersal of its fruits. Yet
experiment shows how well founded the inference was, whilst behind it
lay a life-time of botanical research.

The author thus approaches the subject of the floras of the Pacific
islands rather as a plotter of detail than as a delineator of great
designs. However much we may study the means of dispersal, we have
behind them the great facts of distribution, serving like the main
stations of a trigonometrical survey, and with these we have to make our
lesser facts and observations square. One is conscious all the time that
much of what seems new in one’s researches has already been foreseen by
the generaliser, and that one can do little else than assist in
confirming some of his results. This is all that I can lay claim to in
this work.

The floras of the islands and coasts of the tropical Pacific are here
regarded entirely from the standpoint of plant-dispersal. The fruits and
seeds rather than the flowers have been the subject of my
investigations; and although there is much to please the eye in the
flora of a Pacific island, it was always with a sense of disappointment
that I turned away from some pretty flowering plant that failed to
present me with its seed. Amongst the wonders of the plant-world rank
the Tree Lobelias of the Hawaiian Islands; yet their greatest charm to
me lay not so much in their giant-flowers and their arborescent habit as
in the mystery surrounding the home of their birth and their mode of
arrival in these islands. When I first stood under the shade of the
lofty Dammara vitiensis, the Kauri Pine of Fiji, all my interest lay in
its cones lying on the ground; and I remember how eagerly I handled my
first specimen, and how anxiously I watched its behaviour when
experimenting on its capacity for different modes of transport. When a
strange plant presented itself on a beach, my first care was to
ascertain the fitness of its fruits or seeds for transport by the
currents; and all inland plants with fruits likely to attract
frugivorous birds were at once invested with a special interest for me.

The mangrove swamps were always great places of interest, and months of
my sojourn in the Pacific must have been passed in exploring their
creeks and in examining their vegetation. Botanists usually avoid these
regions; but the observation of the germination of the Rhizophora fruits
on the trees and the inquiries connected with their methods of
distribution over the oceans were pursuits so engrossing that I ignored
the numerous discomforts connected with the exploration of these gloomy
regions. The magnificent mangrove forests of the Ecuador coast of the
Pacific will live longest in my memory, though the risks were
considerably greater and the discomfort of existence extreme. But the
mangrove swamps present us with glimpses into the conditions of plant
life during the warmer epochs of the earth’s history, when perhaps the
seed-stage was largely dispensed with, whilst an atmosphere, laden with
moisture and screening off much of the sun’s light, enveloped most of
the circumference of the globe.

The plant world viewed only from the standpoint of dispersal may lack
much that is pleasing to the eye, though it abounds with small and great
problems fascinating to the reason. Matters of great moment are here
involved, and in the case of the Pacific islands they concern not only
the source of the oceanic floras, but the story of the islands
themselves; whilst behind these there rise up questions of yet deeper
import, questions that are bound up with the beginnings of genera and
species, and with other mysteries of life on the earth. The distribution
of plants presents something more than a problem of means of dispersal,
or a problem of station, or a problem of plant migration connected with
climatic changes. It is something a great deal more than all three,
since it is indissolubly connected with a past, of which unfortunately
we know very little. Let us take it to be a question of means of
dispersal, and then in imagination transporting ourselves to the
Scandinavian coast, let us gather up the stranded West Indian beans of
Cæsalpinia, Mucuna, and Entada, that have been drifted there for ages by
the Gulf Stream, and lie in some cases semifossilised in the adjacent
peat-bog. Was ever dispersal so utterly purposeless as this? Yet here
lies a principle of plant-dispersal that is fundamental. We see it in
the thistle-seed floating seaward in the wind. Nature never intended its
pappus for such an end. It was formed for quite another purpose, yet it
aids largely the dispersion of the plant. What can be more significant
than that?

Or let us take it to be a matter of station. Given time and the
recurrence of the same conditions, with others I once imagined that we
could explain most things in plant-distribution, whether of plants at
the coast or of plants inland, whether of plants of the alpine peaks or
of plants of the plains, or of plants of the river or of the pond. Time,
it was held, had long since discounted the means of dispersal, and
distribution became merely an affair of station. But the supplanting of
many indigenous species of a flora by introduced species is a common
story in the plant-world; and such a view needs no further discussion
here. Nor is distribution only concerned with plant-migration. Any
theory of the origin of alpine floras on tropical mountains will have to
explain the presence of the temperate genera, Geranium and Sanicula, not
alone on the summits of the mountains of Equatorial Africa and
Madagascar, but on the uplands of Hawaii in mid-Pacific, where also are
found Ranunculus, Vaccinium, Fragaria chilensis (the Chilian
strawberry), and Drosera longifolia.

Taking genera of different stations each in their turn, and following up
the clues thus afforded, it would be possible to find support for all
the reputable views relating to plant-distribution. The wide range of
aquatic plants under conditions that completely change the character of
the terrestrial vegetation, such, for instance, as Myriophyllum and
Ceratophyllum, might be plausibly attributed to the relative uniformity
of the conditions of aquatic life both in time as well as space. The
occurrence of Vaccinium on mountain-tops over most of the world, even on
the highlands of Samoa, Tahiti, and Hawaii in the Pacific Ocean, would
be rightly regarded as evidence of active dispersal of the seeds through
the agency of birds from one mountain-summit to another, whether in
mid-ocean or in the centre of a continent. The prevalence of the same
beach-plants over most of the globe in the same climatic zones would
point unmistakably to the predominant agency of currents. But with many
plant-genera, some of which range the world, whilst others again may be
restricted to a single group of islands in the Pacific, there is often
no question either of means of dispersal, or of station, or of
plant-migration, and problems of a very different nature are opened up.

When we leave the beach and the mountain-top, the river and the pond,
all the troubles of distribution begin; and since but a small proportion
of plants in a typical flora belong to these stations, it follows that
difficulties will dog our steps with the large majority of the plants.
The agencies of dispersal now working around us, the current, the wind,
the insect, the bird, and the bat, will explain many of the features of
littoral and alpine floras and of the vegetation of ponds and rivers.
Here we have in so many cases wide-ranging genera with the means of
dispersal ready to hand. We can connect the wide range of Vaccinium with
the wide range of birds of the grouse and other families that feed on
the berries. We can associate the great areas of aquatic or sub-aquatic
genera, like Potamogeton and Sparganium, with the migratory habits of
the ducks in the stomachs of which we find their seeds. We can connect
the great ranges of beach plants like Ipomœa pes capræ in the tropics,
and Convolvulus soldanella in the temperate regions with the currents,
and the almost cosmopolitan range of many ferns and lycopods with the
winds and other agencies.

When, however, we enter the forests we find genera that are often much
more restricted in their areas, and species that are yet more limited in
their range. There is very little dispersal going on here. The birds are
strange. Their distribution is usually very local. They look lazily down
at us from the branches, as they disgorge the seeds and stones of the
fruits they have eaten, which cover the ground around. We can almost
fancy that they say:—“Our work is done. We rest from the toil of our
ancestors. They carried seeds to far-distant Hawaii, Tahiti, and Savaii.
Our work is done.” And as we walk through those noiseless forests, where
the machinery of species-making is ever in silent motion, we become
aware that we are treading one of Nature’s great workshops for the
manufacture of species and genera. Outside the forest all is bustle and
hurry. We are in the streets, or rather in the distributing areas of the
plant-world. We hear the noise of the breaker, the roar of the gale, the
cry of the sea-gull, the flapping of a myriad pairs of wings of some
migrating host overhead, and we know that the current, the wind, and the
bird are actively at work; but their operations are confined mainly to
the beach, the mountain-top, the river, and the pond.

Let us take a well-wooded Pacific island several thousand feet in
height. We find on its beaches the same littoral plants that we have
seen before on the tropical shores of Malaya, of Asia, of Africa, and of
America. We find in its ponds and rivers the same species of
water-plants, such as Ceratophyllum demersum, Ruppia maritima, and Naias
marina, that are familiar to us in the cool and tepid waters of much of
the globe. On its level summit, if it remains within the clouds we find
in the boggy ground, where Sphagnum thrives, genera that are represented
in Fuegia, New Zealand, and the Antarctic islands, such as Acæna,
Lagenophora, and Astelia, and the world-ranging Drosera longifolia. In
other elevated localities we find Ranunculus, Geranium, Sanicula,
Artemisia, Vaccinium, and Plantago, chiefly genera of the temperate
regions of the northern hemisphere; whilst there are also found Gunnera,
Nertera, and Uncinia, all hailing from the south and belonging to the
Antarctic flora characterising all the land-area around the globe in the
latitude of New Zealand and Fuegia. The Hawaiian species of Nertera and
of Uncinia occur also in New Zealand, and the first-named is found also
in Tristan da Cunha and in South America. In the Hawaiian uplands there
is also to be seen Deyeuxia, a genus of grasses found in the Tibetan
highlands and in the Bolivian Andes at elevations of 16,000 to 19,000
feet; and the same species that exists in Australia may be found in the
mountains of Hawaii. Here also, both in Hawaii and Tahiti, occurs Luzula
campestris.

In making the foregoing remarks on the alpine plants of a Pacific
island, I have had Hawaii in my mind, but we find the elements of a
similar widely-distributed mountain-flora in the less lofty peaks of
Tahiti and Samoa, and traces even in Fiji, where the mountains, however,
have only a moderate elevation. But the point I wish to lay stress on is
the cosmopolitan yet temperate character of the mountain-flora of an
island lying in the midst of the tropical Pacific. As he shifts his
station on this mountain-summit, the observer might at different times
imagine himself in the Sierra Nevada of California, on a Mexican
tableland, on a peak of the Andes, or in the lowlands of Fuegia. Other
plants that I have not mentioned, such as Coprosma, would bring back to
him New Zealand. He might even be on a mountain-top in Central Africa,
or on a Madagascar plateau; whilst in the boggy region of an elevated
Hawaiian tableland he would meet with not only the physical conditions,
but also several of the plants found on the higher levels of Tristan da
Cunha.

It is, however, to be noted that although these mountain-tops in the
mid-Pacific have been stocked with genera from the four quarters of the
compass, the species as a rule are restricted to that particular
archipelago. Whilst the beach and the river in most cases possess plants
that have very wide ranges over the earth, a good proportion of the
species on the mountain-summit are not found elsewhere. This implies a
partial suspension of the means of dispersal on the mountain-top, whilst
the currents and waterfowl are still actively distributing the seeds of
the littoral tree and of the aquatic plant. We here get a foreshadowing
of another great principle, or of another line along which Nature has
worked in stocking these islands of the Pacific with their plants, a
subject concerning which much will be said in later pages.

Hitherto, we have dealt only with a small proportion of the flora, and
with but a small portion of the area of the island. We have yet to deal
with the intermediate region between the sea-border and the summit of
the island, or, in other words, with the forested mountain slopes. This
is the home of many of the peculiar species and peculiar genera, both of
plants and birds; and it is with this zone that we shall be mainly
concerned when we come to contrast the floras of the several
archipelagoes of the tropical Pacific. Here the agencies of dispersal
have, to a large extent, ceased to act; and the question will arise as
to the connection between the endemic character of the plants and the
endemic character of the birds. We shall have to ask why this island,
after receiving so many plants, ceased to be centres of dispersal to
other regions. It is possible that these seeds or fruits have lost their
capacity for dispersal; but only a few instances of this change present
themselves. Rather it may be supposed that the birds that originally
brought the seeds to the island came to stay; and this at once suggests
another query as to the cause of the change of habit. I am alluding here
not to the plants with minute seeds, such as Sagina and Orchis, which
Mr. Wallace, in his _Darwinism_, regards as capable of being transported
by strong winds over a thousand miles of sea; but to those numerous
plants found in the Fijian, Tahitian, and Hawaiian forests, where the
seeds and “stones” are large and heavy, measuring often as much as a
quarter of an inch (6 mm.), and sometimes nearly an inch (25 mm.) in
size. The reader will be surprised to learn how little “size” has
determined the distribution of seeds and fruits in the Pacific. He will
have to appeal to the habits of pebble-swallowing of the Dodo, the
Solitaire, the Goura pigeon, the Nicobar pigeon, &c., if he desires to
find a parallel in the habits of birds.

It is here assumed that the reader is already acquainted with the
principles involved in a discussion of island-floras, principles clearly
laid down in the writings of Hooker, Wallace, Hemsley, and others. As a
general rule in an island or in a group of islands where there are a
large number of plants not found elsewhere, there is also a large
endemic element in the avifauna, and where none of the plants are
peculiar, endemic birds are either few or wanting. As an example of the
first we may mention Hawaii, and Iceland affords an instance of the
second. But there is no hard and fast rule connecting the endemic
character of the plants and birds of an island with its distance from
other regions. Even the small group of Fernando Noronha, lying only some
200 miles off the coast of Brazil, possesses its peculiar birds and its
peculiar plants; and we can there witness the singular spectacle, as
described by Mr. Ridley, of an endemic bird, a frugivorous dove, engaged
in scattering the seeds of endemic plants over the little group. This is
the only fruit-eating bird in the islands, remarks the same botanist in
the _Journal of the Linnean Society_ (vol. 27, 1891); and “when one sees
the number of endemic species with edible fruits, one is tempted to
wonder if it were possible that they were all introduced by this single
species of dove, or whether other frugivorous birds may not at times
have wandered to these shores.” This inter-island dispersal in a
particular group of peculiar plants by peculiar birds is a common
spectacle in the Pacific. The contrast between the large number of
plant-genera possessing fruits that would be dispersed by frugivorous
birds and the poverty of fruit-eating birds in the avifauna is well
displayed in Hawaii.

The island of St. Helena would seem to offer an exception to the rule
that endemic birds and endemic plants go together, since, though its
flora possesses a very large endemic element, there are scarcely any
endemic or even indigenous birds recorded from the island. We can never
know, however, how much of the original fauna disappeared with the
destruction of the forests. It would nevertheless appear that but few of
the genera possessing peculiar species of plants were adapted for
dispersal by frugivorous birds. The lesson to be learned from this
island concerns the Compositæ, often arboreous, that constitute the
principal feature of its flora. St. Helena retains almost more than any
other island evidence of the age of Compositæ which has left its impress
on many insular floras; and when we discuss the original modes of
dispersal of the endemic Hawaiian genera of the same order we shall look
to the flora of this Atlantic island for assistance in the matter. To
the age of Compositæ belong the beginnings of several insular floras.

To return to the main line of our argument, it would seem that in a
Pacific island there is a constant relation between free means of
dispersal and the preservation of specific characters. The ocean-current
and the aquatic bird are in our own time actively engaged in dispersing
the seeds of shore-plants and water-plants, and we see the same species
ranging over the world. On the other hand on the mountain-top the
agencies of dispersal are beginning to fail, and as a result many a
mountain has some of its species restricted to its higher regions. In
the forest zone there has been a more or less complete suspension of the
activity of the dispersing agencies, and new genera are formed whilst
peculiar species abound. Free means of communication with other regions
restrains but does not arrest the differentiating process that is ever
in progress throughout the organic world. Isolation within certain
limits gives it play.

It is in this connection interesting to reflect that during the
differentiation of the inland flora the littoral plants have lagged
behind or have remained relatively unchanged. The currents have been
working without a break throughout the ages; and the cosmopolitan
Ipomœa, that now creeps over the sand of the beach, or the wide-ranging
Rhizophora, that forms the mangroves of the coast-swamp, must have
witnessed the arrival of the ancestors of several of the endemic inland
genera. The swamp-plants of the littoral flora are probably older,
however, than the beach-plants which have been recruited from time to
time in one region or another of the tropics from the inland flora. Yet
as a body the littoral plants have lagged far behind the inland flora.
We might thus expect that in a Pacific island, excluding the
wind-distributed plants, such as the ferns and the lycopods, the most
ancient types of the plants would be found at the coast, the most modern
in the forests, whilst the plants of the mountain-summit would represent
an intermediate age.

But true as this may be, the composition of a strand-flora is a very
complex one. Although, as Prof. Schimper remarks, the mangrove formation
is more isolated than the beach formation, and affords evidence of a
much earlier separation, the beach-plants as a body are anything but
homogeneous in their character. Their physiognomy varies to some extent
with the alteration in the characters of the inland flora, changes to
which the mangrove formation makes a very slow response. Yet amongst the
plants of the beach we find strangely assorted forms that are as ancient
denizens of the coast as the mangroves themselves. Take, for instance,
Salsola Kali, that thrives alike on a beach in Chile, on the sea-shore
in Devonshire, and in the salt-marshes of the interior of Tibet. Then,
again, there is a type of littoral plant, of which Armeria vulgaris and
Plantago maritima may be taken as examples, which is equally at home on
the beach and on the tops of inland mountains. We might in a sense apply
the wrecker’s motto,

“What the sea sends and the land lends,”

to the history of a littoral flora. Yet on the other hand the inland
flora in its turn receives a few recruits from the littoral flora; and
it is the relation between the inland and coast species of the same
genus that offers one of the most fascinating studies in the botany of
the Pacific Islands.

This introductory chapter may be concluded with a few remarks on what
may be termed “the ethics of plant-dispersal.” Not that this is in any
way a suitable phrase, but it best expresses my sense of the lack of
propriety in some things connected with this subject. It is odd, for
instance, that we speak of the dispersal of plants and animals in the
same breath, as if the process was in both cases identical. Seeing that
from this point of view we judge a plant only by its seeds and fruits,
it is apparent that we are following quite a different method than that
which we employ in the study of the dispersal of animals. Whilst the
zoologist classifies the units of dispersal, the botanist does nothing
of the kind; and the two systems of classification are at the outset
fundamentally distinct. The student of plant-dispersal thus often finds
himself placed in an awkward dilemma. For him a family is a collection
of allied genera having similar seeds or fruits and fitted often for the
same mode of dispersal. A family like Sterculiaceæ, possessing such a
variety of seeds and fruits suitable for very different modes of
dispersal, is from his standpoint a collection of dissimilar units.
Genera like Commersonia, Waltheria, Kleinhovia, Sterculia, and
Heritiera, that he so often meets with in the Pacific Islands, have in
these respects frequently very little in common; and yet one of the
earliest determining influences in plant-life must have lain in the
capacity for dispersal.

Yet chance seems to reign in the processes of plant-dispersal ever going
on around us. In the floating seed, in the achene with its light pappus
blown before the gale, in the prickly mericarp entangled in the plumage
of a bird, in the “stone” of the drupe disgorged or ejected by the
pigeon, in the small grain that becomes adhesive in the rain, in the
tiny rush-seed enclosed in the dried pond-mud on the legs of some
migratory bird, in all these we see the agencies of dispersal making use
of qualities and of structures that were developed in quite another
connection and for quite another purpose. That such characters have been
so to speak appropriated by these agencies is a pure accident in a
plant’s life-history. If the evolutionary force had been in operation
here, it would have selected some common ground to work on. There would
have been some uniformity in its methods, whereas the modes of dispersal
are infinite. The qualities and characters that happen to be connected
with dispersal belong to a plant’s development in a particular
environment. They can never have been adapted to another set of
conditions that lie quite outside that environment. There is a relation
of a kind between the specific weight of wood and the density of water,
and this, in a sense, sums up the connection between a seed and its
distributing agencies.

Evolution has never concerned itself directly with means of dispersal.
Evolution and Adaptation represent the dual forces that rule the organic
world, the first an intruding force, the last a passive power
representing the laws governing the inorganic world. To these laws the
intruding power has often been compelled to bend, and it has had to pay
its price, and sometimes it has succumbed, and sometimes it has turned
its defeat into a victory. Nature, so watchful over the young plant, as
represented by the seed, is finally compelled to let it go, and
dispersal begins where evolution ends, or rather when the evolutionary
power fails. The seed-stage itself is the price of adaptation. The death
of the individual may also be regarded from the same standpoint. It
represents a defeat of the evolutionary force, which, however, has been
retrieved by the gift of reproductive power.

CHAPTER II

THE FLORAS OF THE PACIFIC ISLANDS FROM THE STANDPOINT
OF DISPERSAL BY CURRENTS

IN the previous introductory chapter some of the numerous questions
affecting insular floras were briefly referred to. I will now ask my
reader, if he has had the patience to read it, to consign that chapter
for the time at least into oblivion, and to proceed with me to our
Pacific island with the intention of investigating its flora from the
standpoint of dispersal. We will together take up the subject _de novo_,
after banishing from our minds all preconceptions that we may have
possessed.

After having been over the island gathering specimens of all the seeds
and fruits, we return to our abode on the beach. But we are puzzled
where to begin. The problem presents itself as a tangled skein, and our
difficulty is to find an “end” that we can follow along with some
chances of success. In our trouble we look around us; and at that moment
we see a number of floating seeds and fruits carried by the current past
the beach. This presents us with a clue and our investigation begins.

[Illustration:

OCEANIA
John Bartholomew & Co., Edin^r.
]

We place all our seeds and fruits in a bucket of sea-water and notice
that many of them sink at once. In a few days we look again and observe
that many more are at the bottom of the bucket, only a small percentage
remaining afloat. We then remark to our surprise that nearly all of the
floating seeds and fruits belong to coast plants, those of the inland
plants, which indeed make up the great bulk of the flora, having, as a
rule, little or no buoyancy. After a lapse of weeks and months the seeds
and fruits of the coast plants are found to be still afloat. In the
results of this experiment we see the work of the ages. There has been,
in fact, a great sorting process, during which Nature has “located” the
plants with buoyant seeds or seed-vessels at the sea-coast, placing the
others inland. This is the clue that we shall follow up during many
chapters of this book; and having in this manner introduced the reader
to the subject, I will now refer to the general results of my
investigations in this direction in the Pacific Islands.

In Fiji there are about eighty littoral plants out of a total of at
least 900 species of indigenous flowering plants, that is to say about
nine or ten per cent. (Note 1), the littoral grasses and the sedges
being with one or two exceptions excluded. These shore plants belong to
the sandy beach and to the coast swamp, and most of them are distributed
over the tropical shores of the Indian and Pacific Oceans, whilst not a
few occur on the coasts of tropical America. They form the
characteristic plants of the coral atoll, and many of them have long
been known to be dispersed by the currents. From the list given in Note
2 it will be seen that these eighty species belong to about seventy
genera. Nearly all of them (95 per cent.) possess seeds or seed-vessels
that float at first in sea-water; whilst three-fourths of them (75 per
cent.) will float unharmed for two months and usually much more, and
several of them will be found afloat after a year or more, being still
capable of reproducing the plant (Note 3).

The prevalence in the Fijian strand-flora of Leguminosæ, which are
included in my list under the divisions Papilionaceæ, Cæsalpinieæ, and
Mimoseæ, is very significant. They make up about 29 per cent. of the
total. Excluding weeds and a few other introduced plants, there are some
fifty species known from the Fijian Islands, and of these almost half
belong to the littoral flora, which as we have seen constitutes only a
fraction (one-tenth) of the whole flora. If we regard the genera, we
find that out of some thirty Leguminous genera twenty are littoral and
in most cases exclusively so. This conspicuous feature in the
constitution of the strand-flora is of prime importance as concerns the
question of adaptation to dispersal by currents, since nearly all the
Leguminosæ with buoyant seeds offer themselves as defiant exceptions to
any such law.

I will now contrast the Fijian inland flora with that of the coast from
the point of view of the buoyancy of the seed or fruit, according as it
presented itself for possible dispersal by currents. Rather over a
hundred plants were experimented upon (Note 4). After excluding some
introduced plants there remain some ninety species belonging to about
sixty genera, and of these quite 75 per cent. sank at once or in a few
days. I may add that all kinds of fruits are here represented, the
capsule, the achene, the coccus, the berry, the drupe, &c. Of the
buoyant residue few possess seeds or fruits that will float uninjured
for any length of time. Not many gave indications directly in opposition
to the principle that whilst the seeds or fruits of shore-plants
generally float, those of inland plants usually sink, since as pointed
out in Note 5 most of the difficulties are removed during the subsequent
developments of the principle discussed in the later pages of this work
or are to be explained on other grounds stated in the note.

We pass now from Fiji as typical in its flora of the Western Pacific to
Tahiti as representing in its flora the more strictly oceanic groups of
Eastern Polynesia. In the Tahitian region, which is taken as including
in a general sense the Society Islands, the Marquesas, and the Paumotus,
there are only between 50 and 60 littoral plants, excluding the
occasional additions from the inland flora. As indicated by the letter T
preceding the species in the list of Fijian shore plants, nearly all are
to be found in Fiji, and the few not yet recorded from that group, which
I have referred to in the remarks following the list, will probably be
found there by some subsequent investigator. In Tahiti also between 75
and 80 per cent. of the strand plants have seeds or seedvessels that
float for months; and here also Leguminosæ predominate, forming about 30
per cent. of the total. A conspicuous negative feature in the Tahitian
strand-flora is concerned with the absence of the mangroves and their
numerous associated plants, which together form the mangrove formation
in Fiji. This remarkable character in the distribution of shore plants
in the Pacific is discussed in Chapter VI.

Not having visited Tahiti, I can only deal inferentially with the inland
plants, as in the case of the strand-flora. Here also the plants are in
the mass Fijian in a generic and often in a specific sense, and there is
no reason to believe that the principle involving the non-buoyancy of
the seeds or fruits of inland plants does not as a rule apply to Tahiti
as well as to Fiji.

The Hawaiian Islands, standing alone in the North Pacific, form a floral
region in themselves, a region that is the equivalent not of one group
in the South Pacific, such as that of Fiji or of Tahiti, but of the
whole area comprising all the groups extending from Fiji to the Paumotu
Archipelago. Lying as it does mainly outside the zone of influence of
the regular currents that would bring the seeds of tropical plants to
its shores, Hawaii possesses a strand-flora that is meagre in the
extreme. Not only does it lack the mangrove formation so characteristic
of Fiji, but it lacks also many of the plants of the beach formation
that are found both in Fiji and in Tahiti, plants that give a peculiar
beauty to the reef-girt beaches all over the South Pacific. Its poverty
is sufficiently indicated in the number of its species, thirty in all,
barely more than half of the number found in Tahiti, and not much over a
third of those occurring in Fiji. Though coral reefs with their
accompanying beaches of calcareous sand are relatively scanty, the
characteristic littoral plants have not been numerous enough to hold
their own against intruders from the inland flora, and endemic species
have taken a permanent place amongst the strand plants. The Hawaiian
strand-flora has thus quite a facies of its own, and it will be found
discussed in Chapter VII., whilst a list of the plants is given in Note
28. It will thus not be a matter for surprise that the littoral flora of
Hawaii follows the principle of buoyancy only in a modified degree. It
is true that about two-thirds of the species of the present beach flora
possess seeds or seed-vessels that float for months; but since there are
reasons for believing that several of them are of aboriginal
introduction, this proportion is reduced to a third. In the list of the
Fijian shore plants given in Note 2, those occurring also in Hawaii are
preceded by H.

When we look to the Hawaiian inland flora for indications respecting the
principle of the non-buoyancy of the seeds or seed-vessels of inland
plants, we find that so far as it has been there tested this principle
receives fresh support from the plants growing on the slopes of the
Hawaiian mountains. Although the author was only able to sample the
inland flora, we have in the list given in Note 6 all kinds of plants,
from the forest-tree to the herb, and most varieties of fruits.
Excluding a few introduced plants, there are in this list about fifty
species of indigenous plants belonging to about forty genera. Of these
plants quite 80 per cent. possess seeds or fruits that sink either at
once or in a week or two. Of the “buoyant” residue very few have seeds
or fruits that will float for months. These apparent exceptions to the
principle are in great part capable of being explained on the grounds
referred to in Note 5 in connection with the Fijian inland plants; and I
have alluded to them in Note 7.

The littoral flora of Fiji is essentially Malayan and Asiatic, and for
our purpose is eminently typical. Its plants are found far and wide on
the tropical coasts of the Old World, and sometimes also in the New
World. In more than half the species we are concerned with the dispersal
by currents of more or less dry indehiscent fruits that range usually in
size from a marble to a cricket-ball, as illustrated by those of
Hernandia peltata and Barringtonia speciosa, whilst with most of the
rest the currents distribute large seeds, several of which are
Leguminous, as in the case of Mucuna, Cæsalpinia, and Entada, with
others of the Convolvulus type, as in the instance of Ipomœa pes capræ.
It is remarkable that in selecting plants with buoyant seeds or
seed-vessels for a station at the coast Nature has generally ignored
those with very small seeds. When such small seeded plants, as Sesuvium
portulacastrum, occur on the beach, the seeds have as a rule no
buoyancy. Pemphis acidula is, however, an exception; but its case is a
very rare one. It will be established in the next chapter that the
non-buoyancy of small seeds is generally true also of plants growing by
the river or by the pond.

The point at which we have arrived in our inquiry concerning the general
collection of seeds and seed-vessels that we placed in sea-water is that
_the plants with buoyant seeds or seed-vessels have been for the most
part “located” at the coast_. But if we look a little more closely at
the sunken and floating seeds, we find that in the same genus there are
species with seeds or seed-vessels that sink and species with those that
float. We look again and then perceive that the same general principle
is true of different species of the same genus growing inland and at the
coast. We learn now that as a rule when a genus possesses both littoral
and inland species, the seeds or fruits of the former float in sea-water
for a long time, whilst those of the latter have little or no floating
power. But we have yet to examine the structure of the coverings of the
buoyant seed or fruit; and we shall then discover that the different
behaviour in water is often associated with corresponding structural
differences of a striking character. The structural causes of buoyancy
are dealt with in Chapter XII.; and we will now content ourselves with
enunciating the second principle that _in a genus comprising both coast
and inland species, only the coast species possess buoyant seeds or
seed-vessels_.

The important principle above indicated was not altogether new to me, as
is shown in the next chapter. But it was new in the case of the floras
of the Pacific Islands. When it first presented itself in Hawaii I was
engaged in trying to find a connection between the inland and littoral
species of Scævola; and its discovery led me to form a plan worthy
almost of Don Quixote, namely, to cultivate the beach species of Ipomœa,
Scævola, and Vitex in the interior with the hope of finding them
converted into inland species when I returned to Hawaii after a lapse of
years. Little matters often determine a career, and for a while my
future movements and probably the remainder of my life were largely
centred around my interests in the well-being of Scævola Kœnigii. The
scheme was actually undertaken, and I had fixed on a little plot of land
at the foot of the mountains rising behind Punaluu in Kau. The
transaction was on the point of completion when the owner changed his
mind and the plan fell through. Subsequent observation and reflection
have led me to believe that in most cases no connection exists between
the littoral and inland species of a genus; and I have dwelt on this
incident merely to show the importance that I rightly attached to this
distinction, whilst misinterpreting its meaning.

But to return to my own investigations. Had I indeed read more carefully
Professor Schimper’s work on the Malayan strand-flora, this subject
would have been found discussed by an observer far abler than myself,
though from a very different standpoint, that of Adaptation and Natural
Selection. He points out (pp. 179-182) that with a number of these
tropical genera possessing both littoral and inland species, such as
Barringtonia, Calophyllum, Clerodendron, Cordia, Guettarda, and
Terminalia, greater buoyancy of the fruits of the shore species is
associated with certain structural characters in the fruit-coverings,
whilst with the inland species, where the floating power of the fruits
is either much diminished or entirely absent, these structural
characters are either less developed or lacking altogether.

The question of structure and the debateable matters concerned with it
are treated at some length in Chapters XII. and XIII., and Professor
Schimper’s views are there given. I will content myself with remarking
that the genus Terminalia was especially studied by him in this respect.
He tested the buoyancy of the fruits of ten species, and found that the
flotation period varied from nothing to 126 days and more. By far the
best “floaters” were the fruits of Terminalia Katappa, the only littoral
species tested, all the others being inland species with less buoyant
fruits, and diminished ranges, some of the fruits sinking at once,
whilst the others sank usually in a few days or in a few weeks. It was
also ascertained that, although the buoyant tissue in the fruit-coats
varied in amount generally with the floating-powers, it was rarely
absent altogether in the inland species, a very significant conclusion,
as will subsequently be pointed out.

Several other striking examples of this principle came under my notice
in the Pacific, and perhaps the most significant is that of Scævola, a
genus of the Goodeniaceæ, confined mainly to Australia and the Pacific
islands, but possessing also a littoral species, S. Kœnigii, that is
found on tropical beaches all round the globe. It is associated in both
Hawaii and Fiji with inland species, none of which are common to the two
archipelagoes, and in the case of the Hawaiian species not found outside
the group. All the species have fleshy drupes, both coast and inland
plants, the “stone” in the littoral species possessing a thick covering
of buoyant tissue, which is absent or but slightly developed in the
inland species. The fruits of the shore species float for many months;
whilst those of the inland species experimented on by me (S.
Chamissoniana and S. Gaudichaudii in Hawaii, and S. floribunda in Fiji)
sank at once or within a few hours. Here we are only concerned with the
difference of buoyancy between inland and littoral species. The several
other questions involved concerning this genus will be dealt with later
on in this work.

The genus Morinda offers another good example of this principle. It
includes one widely-spread littoral species (M. citrifolia), found not
only in all the Pacific archipelagoes, but also over much of the
tropics. It is associated in all the large groups with one or more
inland species, some of which are endemic and others more generally
distributed. The littoral species displays in its pyrenes a singular
air-cavity, the nature of which is discussed in Chapter XII., which
endows them with great floating powers. This cavity is not found in
inland species, and the pyrenes have in consequence no floating power
(see Note 8).

Calophyllum Inophyllum, an Old-World littoral tree, spread far and wide
over the Pacific islands, has very buoyant fruits. In the groups of the
South Pacific it is associated with inland species that are commonly
found in the forests, namely, C. spectabile and C. Burmanni, the fruits
of both of which, according to my observations in Fiji, have limited
floating powers, sinking after periods varying from a few days to four
weeks, and lacking in great part the buoyant coverings of the littoral
species. Professor Schimper obtained similar results with inland species
from other regions (Note 9).

The fruits of the two Fijian coast trees, Barringtonia speciosa and B.
racemosa, possess great floating powers; whilst those of an undescribed
species that I found in the mountains of Vanua Levu sink at once.
Another Fijian inland species (B. edulis, Seem.) that is often planted,
has fruits that float heavily for about a month. This difference in
buoyant powers is also associated with characteristic differences in the
structure of the fruits. It would be interesting to learn what floating
capacity belongs to those of the Samoan endemic species (B. samoensis,
Gray). Professor Schimper’s observations on the genus in the Malayan
region point in the same direction, but more than one difficulty awaits
its solution in the re-examination of the genus. He says, however, that
B. excelsa, Bl., a Malayan species, sometimes cultivated and growing
both inland and at the coast, has fruits that floated for one hundred
days after drying (p. 173).

A striking instance of this principle is afforded in the case of the two
Fijian species of Tacca, the wide-ranging littoral species, T.
pinnatifida, where the seeds float for several months, and the inland
species, T. maculata, Seem., found also in Australia and Samoa, where
the seeds sink at once or in a few days. The seeds of the shore plant
owe their buoyancy to the spongy tissue in their coverings, which is
either absent or much less developed in those of the inland species.
This point might also be determined for the new Samoan inland species
described by Reinecke, the German botanist, as T. samoensis.

Another good illustration is afforded by the two species of Premna of
the South Pacific, though here the buoyancy of the “stone” of a drupe is
concerned. With P. taitensis or P. integrifolia, a small littoral tree
or shrub, these stones possess great floating-power, and are often found
in the floating seed-drift of the Fijian estuaries and in the stranded
drift of the beaches. In the case of Premna serratifolia, an inland tree
of moderate size, the stones have as a rule little or no buoyancy. As
shown in Note 32, where this genus is discussed in detail, the buoyancy
is mainly due to empty seed-cavities.

Other instances might be given in illustration of this principle; but it
will have been noticed that already many of the familiar trees and
shrubs of a tropical beach have been mentioned in this connection either
by Professor Schimper or by myself. There are other genera that afford
similar indications but in a less direct fashion.

For instance, there are three widely spread Leguminous beach plants of
the Pacific, Erythrina indica, Canavalia obtusifolia, and Sophora
tomentosa, none of which are found in Hawaii; but in that group the
genus is represented in each case by an inland species, Erythrina
monosperma, Canavalia galeata, and Sophora chrysophylla, the last two
species being peculiar to those islands. The seeds of the three littoral
species will float for a long time in sea-water, whilst those of the
three Hawaiian inland species have no buoyancy. I may say that some very
interesting questions relating to the origin of these inland species are
here raised. They will be discussed in a later chapter (Chap. XV.).

There are a number of plants belonging to the Convolvulaceæ in these
islands that behave in an irregular way in flotation experiments ; but
their inconstant behaviour can in most cases be explained in accordance
with the principle that in the same genus the shore species have buoyant
seeds and the inland species non-buoyant seeds. Thus, whilst the seeds
of the littoral species, Ipomœa pes capræ, I. grandiflora (Lam.), and I.
glaberrima (Boj.), can float for long periods, and those of the inland
species, I. pentaphylla, I. tuberculata, and I. Batatas (Sweet Potato),
have no buoyancy, the seeds of other inland species, I. insularis
(Steud.), I. bona nox (L.), and I. turpethum (R. Br.), are inconstant in
their behaviour. The three last-named species are, however, to be found
also flourishing at times at and near the coast, and the varying
floating powers of their seeds may probably be connected with their
varying stations. This is indeed suggested by the case of Argyreia
tiliæfolia in Hawaii, in which in my experiments the seeds of plants
growing at the coast floated, sometimes for months, whilst those from
inland plants sank.

This behaviour of the Convolvulaceæ becomes yet more intelligible, and
more in accordance with the principle, when we reflect that the cause of
buoyancy is not concerned with the seed-coats or with the nucleus,
neither of which are able to float, but with the air-spaces left by the
incomplete filling-up of the seed-cavity by the crumpled embryo. The
extent to which the seed-cavity is filled up varies not only between
different genera and between different species of the same genus, but
also amongst individuals of the same species. Even the seeds of Ipomœa
pes capræ, amongst the most typical of floating seeds, display this
variation, and they show it also in their floating power, since about a
third of the seeds usually sink during the first month or two of the
flotation experiments. We can thus explain also why in the case of
Ipomœa insularis seeds from Fiji floated for months, whilst those from
Hawaii had no floating power.

The seeds of the different species of Hibiscus also appear to behave
very irregularly; but even here most of the difficulties can be removed,
when we come to consider a further extension of the principle. Thus,
whilst the seeds of Hibiscus tiliaceus, a wide-ranging littoral tree
known to be dispersed by the currents, float for a long time, those of
H. Youngianus (Gaud.), an endemic Hawaiian species, and of two
wide-ranging species, H. diversifolius (Jacq.) and H. Abelmoschus (L.),
also float for some time. The Hawaiian plant, however, grows in wet
places; and this applies also to H. diversifolius which grows in swamps
at and near the coast. The extension of the principle to water-side
plants generally, which is discussed in the next chapter, will explain
the difficulties connected with these two species. But we have in H.
Abelmoschus a remarkable exception to any rule of buoyancy, since it
grows in dry situations, is often cultivated, and yet possesses a
special layer of buoyant tissue in the seed-coats to which the floating
power is due. The seeds of Hibiscus esculentus (L.), the widely spread
cultivated plant of the tropics, have no buoyancy.

Some curious indications are supplied by Cæsalpinia, a Leguminous genus,
containing two wide-ranging shore species. Speaking generally the rule
applies; and I found in Fiji that whilst the seeds of the two littoral
plants (C. Bonducella and C. Bonduc) were as a rule buoyant, those of an
inland mountain species sank. But it is very remarkable that although
the seeds of C. Bonducella have long been known to be transported by the
currents, and are often stranded by the Gulf Stream on the coast of
Scandinavia, when it grows in Hawaii, where it is as a rule an inland
plant, the seeds lose their buoyancy. This is quite in accordance with
the general principle; but I must refer the reader for a general
treatment of this genus to Chapter XVII. There also will be found the
instance of another Fijian littoral plant, Afzelia bijuga, a common
littoral tree with buoyant seeds which also lose their buoyancy when the
tree grows inland. A similar instance is afforded by Kleinhovia Hospita,
the seeds of which seem to lose their buoyancy in inland stations. Not
all littoral plants, however, lose the floating power of the seeds when
grown away from the coast. The seeds of Ipomœa pes capræ retain it in
spite of the change of station. This point is dealt with in Chapter XIII
and in Note 44.

In concluding this general sketch of the first results obtained by
testing the buoyancy in sea-water of a collection of seeds and fruits
from a mountainous Pacific island, such as we find in Fiji, I must
remind the reader that the subject has only been lightly treated.
Enough, however, has been said to illustrate the character of the
sorting-process by which in the course of ages the plants with buoyant
seeds or seedvessels have been gathered at the coast. This is
indicated:—

(1) By the far greater proportion of species with buoyant seeds and
seedvessels amongst the shore plants than among the inland plants.

(2) By the circumstance that almost all the seeds or fruits that float
unharmed for long periods belong to shore plants.

(3) By the fact that when a genus has both inland and littoral species,
the seeds or fruits of the coast species as a rule float for a long
time, whilst those of the inland species either sink at once or float
only for a short period.

These results, therefore, justify our dividing the flora of our island
into two groups, the one including the plants with buoyant seeds or
fruits and comprising most of the littoral plants, the other including
the plants with non-buoyant seeds or fruits, a group which contains
almost all the inland plants and indeed nine-tenths of the flora. This
classification is a very crude one; but it enables us at once to assign
a value to the agency of currents in stocking a Pacific island with its
plants. Yet this is but the initial step in an inquiry that branches off
in a thousand different ways, even if restricted to the littoral plants.
There are a host of difficulties connected with the history of the
strand-flora of such an island which can only be properly gauged when
viewed from various standpoints.

CHAPTER III

THE LESSON OF THE BRITISH FLORA

THE singular relation between station and seed-buoyancy that exists in
an island of the tropical Pacific, such for instance as Vanua Levu,
Tahiti, or Hawaii, would lose much of its significance if it stood alone
in the economy of plant-life. It must be true not only of tropical
floras generally, but of those of the temperate regions; and there can
be little doubt that it prevails all over the world. Displayed to us at
first in a Pacific island, it acquires a new significance when we study
it in the light of numerous observations made in Europe. It exhibits
itself then as part of a far wider method pursued by Nature in
determining the stations of plants. It is not only at the coast, but
also at the river-bank and at the lake-side that Nature “locates” the
plant with the buoyant seed or seedvessel. This relation is indeed as
well exhibited in inland districts as it is at the coast.

In this connection I have the results of my own investigations on the
buoyancy of the seeds and fruits of British plants and on the
composition of the seed-drift of ponds and rivers, which were carried on
in the years 1890-96. Some of them were published in a short paper on
the seed-drift of the Thames, read before the Linnean Society of London
in June, 1892, and in the columns of _Science Gossip_ for April, May,
and October, 1895; but the mass of the observations remain in my
notebooks. Nor do my observations of the period since elapsed lead me to
alter the position then adopted. I have since pursued the same line of
inquiry in Hawaii, Fiji, on the Pacific coast of South America, and in
Sicily, and with the same results.

Since the elaboration of my notes was begun in 1900, Dr. Sernander, the
Swedish botanist, has published (1901) his work in Swedish on the
Dispersal-biology of the Scandinavian plant-world, in which the
seed-drift of river, pond, and sea is exhaustively treated. Although
this author has dealt with plant-dispersal from a somewhat different
standpoint, I have perused his pages with the keenest interest and with
great profit, having gone over much of the same ground with respect to
the seed-drift of ponds and rivers. Yet the introductory remarks to my
paper in _Science Gossip_ in 1895 are as apposite now as they were then,
and the reader will, I trust, pardon my reproducing them.

“By following up the path of inquiry that is concerned with the
flotation of seeds and seedvessels, we are guided into other fields of
research that give promise of interesting discoveries in connection with
plant-life. We are led in the first place to consider the question of
utility, and to ask whether the buoyancy of the seed or fruit has been a
matter of moment in the history of the species. Nature is ever engaged
in telling off the plants to their various stations. She places the
yellow iris at the river’s side and assigns to the blue iris its home in
a shady wood. Under her direction the common alder thrives at the
water’s edge, whilst its fellow species live on the mountain slope.
These and similar operations are carried on daily around us, and we know
but little of the wherefore and the how. We are induced, therefore, to
inquire whether by pursuing the line of investigation above indicated we
may be able to get a glimpse at the methods adopted by Nature in
selecting stations for plants.”

I possess the results, which are given in Note 10, of buoyancy
experiments and observations on the seeds and seedvessels of about 320
British flowering plants belonging to about 65 families. Of these about
260 are included in my own results, the data for the rest being obtained
from the writings of Darwin, Martins, Thuret, Kolpin Ravn, and
Sernander. In the great proportion of cases, 240, or 75 per cent.,
sinking took place at once or within a week; whilst 80, or 25 per cent.,
floated for a longer period, usually a month or more; and about 60, or
nearly 20 per cent., floated for several months. It is to this last
small group that belong the seeds or seedvessels that float through the
winter in our ponds and rivers.

If the grasses had been properly represented, the grains of which
possess as a rule but little buoyancy, except through air-bubbles
temporarily entangled in the glumes, the proportion of seeds and fruits
that sink at once or in a few days would probably have been about 80 per
cent. Then again, since the plants from stations where buoyant seeds and
seedvessels are most frequently found—that is at the river-side, the
pond-margin, and the sea-coast—are much more completely represented in
these experiments than those from other stations, it would seem that
even 80 per cent. is too low a figure. Even if the 80 plants with the
buoyant seeds or seedvessels included all the species thus
characterised, which they certainly do not, we should obtain an estimate
for the British flora (rather over 1,200 species of flowering plants) of
about 93 per cent. with non-buoyant seeds or fruits. This is, of course,
too high. It is, however, very probable that the proportion of plants
with non-buoyant seeds or seedvessels for the whole British flora is
about 90 per cent.

This proportion of plants with non-buoyant seeds or seedvessels, that is
to say, of those that sink at once or within a week, is also
approximately correct for the flora of one of the larger islands of the
tropical Pacific. The data at my disposal only enable me in the cases of
Fiji and Hawaii to fix it at between 95 and 85 per cent., or on an
average 90 per cent. With the floras of continental regions the
proportion would doubtless be markedly higher. That seeds and
seedvessels as a rule possess but little buoyancy was a sound conclusion
of Darwin, and one, as he remarked, that is in accordance with the
common experience of gardeners. Thuret, after experimenting on the
buoyancy in sea-water of the seed or seedvessels of 251 species of
plants, belonging to 77 families and to various regions, found that
scarcely two per cent. had any powers of flotation, all the rest sinking
at once or in a few days, a result that led De Candolle in a note to
this memoir to reiterate his opinion regarding the inefficacy of
currents as plant distributors. Thuret, however, did not select many of
his plants from stations where buoyancy is most frequently exhibited,
and his estimate errs, therefore, in imputing too little buoyancy to
seeds in general. The power of seeds and fruits to germinate after
prolonged flotation in sea-water has long been well established, and it
is often illustrated in this work, so that there is no need to dwell
upon it here. (See Note 11.)

Of the 240 species of British plants where sinking took place at once or
within a week, in about 50 per cent. the plants had dry indehiscent
fruits, such as we find in the genus Ranunculus and in the Umbelliferæ,
the Compositæ, and the Labiatæ; whilst in about a third the plants had
dehiscent fruits with small seeds, such as are characteristic of the
Cruciferæ, the Caryophyllaceæ, and the Juncaceæ. Plants with large
seeds, such as those of Nuphar luteum and Convolvulus arvensis, make up
only six per cent. of those of the non-buoyant group, the remainder
comprising plants with berries, such as Solanum, and others with
miscellaneous fruits.

Of the 80 plants where the seeds or fruits floated more than a week,
usually for several weeks, and often for months, 70 per cent. possessed
dry, indehiscent fruits, such as those of Hydrocotyle vulgaris, Bidens
cernua, Lycopus europæus, Carex, &c., whilst only 6 or 7 per cent. had
dehiscent fruits with small seeds, such as we find in Lysimachia and
Menyanthes, the remainder being generally characterised by large seeds,
such as those of Convolvulus sepium, C. soldanella, Iris pseudacorus,
Calla palustris, &c. It would thus appear that, in so far as buoyancy is
concerned, Nature has for the most part ignored the small seed and has
confined herself mainly to the dry indehiscent fruit. We have already
seen that this is also true of the same great sorting-process in the
tropical islands of the Pacific, and it doubtless applies all over the
world.

We have now to learn the significance of this distinction amongst
British plants between those with and those without buoyant seeds or
seedvessels. When we regard the stations of these 80 plants of the
buoyant group we find that about 70 per cent. of them are placed by the
river, or the pond, or the sea, the fresh-water stations much
predominating. But if we include the plants of the moist meadows
adjoining the rivers, such as Ranunculus repens, Rhinanthus crista
galli, some Cyperaceæ, &c., the buoyant fruits or seeds of which are
regularly swept into the stream in the time of flood, we shall raise the
proportion possessing a water-side station to 80 per cent. On the other
hand, about two-thirds of the 240 plants of the non-buoyant group, which
are enumerated in Note 10, live away from the water-side; but the
proportion of plants with a relatively dry station would be considerably
higher than this figure for the whole flora, since my investigations
were especially directed towards plants frequenting wet stations, and
the number of them is excessive in the list.

Supposing, however, that our materials were restricted to the 260 plants
tested by myself, we should obtain highly instructive results, since in
a general sense the floating powers of their seeds or fruits were tested
to the finish. We place them, let us say, in a bucket of water, and
after six months we find that in not more than forty plants are the
seeds or seedvessels still afloat. These forty plants, excluding two or
three littoral plants, are nearly all plants of the borders and vicinity
of rivers and ponds. (They are indicated in the list given in Note 10 by
the numbers vi. and xii., the last being those where the flotation
experiment was prolonged to a year and over.)

It would thus seem—I am now quoting mainly from my paper in _Science
Gossip_ for May, 1895—that there are gathered at the margins of rivers
and ponds, as well as at the sea-border, most of the British plants that
could be assisted in the distribution of their seeds by the agency of
water. This great sifting experiment has been the work of the ages, and
we here get a glimpse at Nature in the act of selecting a station. But
the curious character of the sorting process becomes yet more apparent
when we discover that the buoyancy of the seeds or fruits of species of
the same genus may become a matter of station.

We will first take the four British species of Stachys (arvensis,
betonica, sylvatica, and palustris). Of these the fruits of S. palustris
alone possess any buoyancy, being able to float for weeks. It is the
only species that finds its characteristic home at the water-side; and
as observed by Sernander its reproductive shoots occur in the
Scandinavian fresh-water drift.

Galium illustrates the same principle. Whilst in my experiments the
fruits of G. aparine and of another species growing in a dry station
displayed little or no floating power, those of G. palustre, which alone
grows at the water-side and in wet situations, have great buoyancy. As
my observations show, they float unharmed through the winter in our
ponds and rivers, and, according to Sernander, are often found in the
Baltic sea-drift. (See Note 12.)

The achenes of Potentilla afford another example. Those of P.
tormentilla and of another species from dry situations have but little
floating power. On the other hand, those of P. comarum float
indefinitely. The last also came under my notice in the floating drift
of ponds in February; and we learn from Sernander that they occur in the
fresh-water and salt-water drift of Scandinavia.

As a further instance, I will take the two British species of Iris. The
familiar river-side Iris pseudacorus has seeds that float unharmed in
our ponds and rivers from the autumn to the spring, and often for a year
or more. On the other hand, the seeds of Iris fœtidissima, which has its
home in the shady wood, sink at once even after drying for months.

The nature of the sorting-process is especially well shown in some of
the families, as for instance with the Labiatæ. Let the reader put on
one side the four species with buoyant fruits, namely, Lycopus europæus,
Mentha aquatica, Scutellaria galericulata, and Stachys palustris, and on
the other side all the species with non-buoyant fruits, such as Salvia
verbenaca, Thymus sp., Calamintha officinalis, Nepeta glechoma, N.
cataria, Prunella vulgaris, Stachys arvensis, S. betonica, S. sylvatica,
Galeopsis tetrahit, Ballota nigra, Lamium purpureum, L. album, Teucrium
scorodonia, and Ajuga reptans, and he will at once perceive that he has
separated the regular water-side plants from those growing in drier
stations.

If he does the same with the Umbelliferæ he will find that when he is
separating Hydrocotyle vulgaris, Cicuta virosa, Œnanthe crocata, and
Angelica sylvestris from Æthusa cynapium, Pastinaca sativa, and
Chærophyllum sylvestre, on account of their buoyant fruits, he is also
distinguishing them on account of their stations. On the other hand,
there are apparently weighty exceptions to this rule in the non-buoyancy
of the fruits of the three British species of Apium (graveolens,
nodiflorum, inundatum), which grow in streams and marshes. Or, again, if
we look at the sea-coast representatives of the family, we find that
whilst the fruits of the Samphire (Crithmum maritimum) float buoyantly
for months, those of Eryngium maritimum seemingly set the law at
defiance, and all sink in less than a week or ten days, even after
months of drying. To regard these as exceptions, however, is to miss the
essential point of the principle concerned. It is not thereby implied
that all water-side plants, whether by the sea or by the river or by the
pond, have buoyant fruits or seeds, but that nearly all plants with such
fruits or seeds have been gathered at the water-side. It will be shown
in the next chapter that several other influences go to determine the
station of a plant on a beach or by a river. This is true of the
Compositæ, which, if we except our two species of Bidens (cernua and
tripartita), come under the play of other determining causes, as
indicated by the little or no buoyancy displayed by the fruits of Aster
tripolium, Senecio aquaticus, and Carduus palustris.

Within the limits of a genus we can, however, point to other examples of
this principle. Take, for instance, Convolvulus arvensis, the common
weed of our fields. Its seeds, whether fresh or dried for months, have
no buoyancy. On the other hand, those of Convolvulus soldanella float
unharmed in sea-water for half a year and more. Its seeds have come
frequently under my notice among the stranded drift of the Devonshire
beaches, and also on the coasts of Chile; whilst Sernander includes them
amongst the drift of the beaches on the Norwegian coasts. It is
remarkable that Convolvulus sepium, which accompanies C. soldanella over
much of its great range, has seeds that are sometimes able to float
unharmed for long periods, even for years (Notes 13, 41, 49). Though not
strictly a water-side plant, it grows commonly over other plants on the
banks of the Thames; and when it fruits its seeds occur typically in the
floating drift of that river. According to Gray, it is almost a
river-side plant in the United States, where it is found “especially on
the moist banks of streams.” Not all the seeds of C. sepium, however,
are buoyant; and in its varying behaviour in this respect it resembles
the inland species of Ipomœa, which are referred to in the previous
chapter.

The British species of Euphorbia also seem to behave in accordance with
the principle that when a genus has littoral and inland species, the
first-named alone possesses buoyant fruits or seeds. Thus, whilst the
sound fruits of E. helioscopia and of another species found commonly as
a garden weed are non-buoyant, those of E. paralias, the familiar
beach-plant, float for several weeks, and are to be noticed among the
stranded drift of the coasts frequented by this plant. (See Note 90 for
later results.)

The structural characters connected with the buoyancy of the seeds or
seedvessels of some of the British plants are dealt with in Chapter XII.
Here it may be remarked that this capacity is often associated, as with
the Pacific island plants, with a “buoyant” tissue, that is either
absent or less developed in the case of the non-buoyant group.

Enough has now been said to show in a general fashion how Nature through
the agency of buoyant seeds and fruits has affected the stations of
plants of the British flora. Allowing this line of inquiry to develop
itself as the work proceeds, we will here pause and close the chapter
with a reference to some of the principal points that have been brought
into prominence.

(_a_) The proportion of flowering plants of the British flora that
possess buoyant seeds or seedvessels is very small, probably not more
than 10 per cent.

(_b_) In so far as buoyancy is concerned, Nature has for the most part
ignored the dehiscent fruit with small seeds, such as we see in the
Cruciferæ and the Caryophyllaceæ, and has chiefly endowed with floating
power the dry indehiscent fruit, such as we see in the Umbelliferæ and
in the Labiatæ.

(_c_) In the great sorting-process that has been in operation through
the ages, nearly all the plants with buoyant seeds or seedvessels have
been located at the water-side, principally by ponds and rivers, but
also on the sea-beach. On the other hand, the great majority of the
plants with seeds or seedvessels that sink have found a home in drier
stations.

(_d_) The character of the operation is well displayed in certain genera
possessing species of the water-side and species of drier situations,
and in the case of genera having both coast and inland species. In both
instances the species by the water-side possesses buoyant seeds or
fruits, whilst that of the station in a drier locality or removed from
the coast has seeds or fruits that sink.

(_e_) Yet it is necessary to remember that the principle involved is not
that all water-side plants have buoyant seeds or fruits, but merely that
plants thus endowed gather at the water-side. There are many plants with
non-buoyant seeds or fruits on our beaches and beside our ponds and
rivers.

(_f_) We have now learned from the British flora that the “locating” of
plants with buoyant fruits or seeds on the beaches of the tropical
islands of the Pacific, and indeed of tropical regions generally, is but
a part of a much wider principle by which plants thus endowed are placed
at the water-side, whether by a river or a pond or by the sea.

(_g_) It is with this distinction between a fresh-water and a salt-water
station that we shall be occupied in the next chapter; and it is of
great interest, since it leads us to discover that the wider principle
is in its turn part of a far larger scheme.

* * * * *

_Note._—It must be clearly understood that by water-side plants the true
aquatic plants, such as the Water-lilies, the Myriophylls, the
Potamogetons, &c., are not implied. It will be seen from the list in
Note 10 that in most cases the seeds or fruits of aquatic plants have
little or no floating power. This is true, for instance, of Ranunculus
aquatilis, Nymphæa, Nuphar, Myriophyllum, Ceratophyllum, Callitriche,
Naias, Zannichellia, Ruppia, and half the Potamogetons.

CHAPTER IV

THE LESSON OF THE BRITISH FLORA (_continued_)

BY following up the clue supplied by the floating seed, we have arrived
at the conclusion with respect to the British flora that plants with
buoyant seeds or fruits gather at the water-side. But we have yet to
inquire why some of these plants are “located” at the sea-coast and
others on the borders of ponds and rivers. Mere buoyancy aided by chance
has not determined the choice. There are definite principles at work in
the economy of plant-life that make the selection for each plant.

Rivers in all parts of the world carry to the sea in great abundance the
seeds and fruits of the plants that are stationed at their borders; and
such seed-drift is found in quantity washed up on the beaches in the
vicinity of the estuary. One finds, for instance, on such beaches in the
South of England the stranded fruits and seeds of Bidens cernua, Alnus
glutinosa, Sparganium ramosum, Iris pseudacorus, &c., mingled with those
of true beach plants like Cakile maritima, Convolvulus soldanella,
Euphorbia paralias, &c. Yet we would be much surprised if either the
Bidens or the Alder or the Sparganium were to establish itself on the
sandy beach, even though they have had through the ages innumerable
opportunities of doing so. We thus see that mere buoyancy of fruit or
seed cannot determine a station on a sea-beach, and that some other
factor makes the choice. The nature of this factor I will now endeavour
to explain; but in so doing it will be necessary to employ a few
technical terms, which it is not easy to dispense with altogether.

It may be doubted whether Professor Schimper could have conferred a
greater benefit on the student of plant-distribution than in his clear
delineation of the connection between the habit or organisation of a
plant and its station. Nature has imposed an important structural
distinction between plants that have been endowed with the means of
checking excessive transpiration or water-loss in stations where there
is risk of drought, as in deserts and in similar arid localities, and
those that live in stations where such safeguards are not needed. Hence
arises the distinction between Xerophytes on the one hand, and
Hygrophytes on the other. This contrast is shown not only in minute
structural features, but also, as my readers are aware, in the external
characters, as in hairiness, succulency, a leathery cuticle, the
occurrence of thorns, and in several other characters of the plants of
the steppe and the desert. This important subject is dealt with by
Professor Schimper in his recent work on Plant-Geography; but it was
from his earlier work on the Indo-Malayan strand-flora that I learned
this valuable lesson in plant-distribution.

It has been ascertained, however, that a safeguard against excessive
water-loss by transpiration is not only needed by plants living in arid
localities, but also by those placed at the coast. Both the shore plant
and the plant of the steppe and the desert present the same xerophilous
organisation, provision against excessive transpiration being also
required by the beach plant to prevent the injury of the green cells
from the accumulation of salt in the tissues. It would thus appear that
plants of the Hygrophytes that possess buoyant seeds or fruits are
gathered at the borders of ponds and rivers, whilst those of the
Xerophytes that are similarly endowed find their station on the
sea-shore. This important distinction penetrates very deeply into the
conditions defining the stations of plants. The connection between the
plant of the coast and the plant of the steppe or the desert is
strikingly shown on those occasions when the beach plants extend inland
over parched and arid plains, such as occurs for instance in North
Africa, and in the larger islands of Fiji, as described in Chapter V.

The causes of the buoyancy of fruits and seeds, as pointed out in
Chapter XII, are so various, that it appears at first sight impossible
to connect them with the xerophilous or hygrophilous organisation of a
plant, or, in other words, with any structural characters associated
with particular stations; yet behind all lies the general principle
that, given a plant of the buoyant group, if it is a Xerophyte it finds
its way to the coast, and if a Hygrophyte it makes its home by ponds and
rivers. In the case of a tropical littoral flora, such as we find in a
Pacific island, the large proportion of plants with buoyant fruits or
seeds gives so much prominence to the subject of their distribution by
currents that the question of “station” is often masked. On the other
hand, in the shore-flora of a temperate region like that of Great
Britain, the plants with buoyant seeds or fruits are in the minority,
and the question of “station” is the first to obtrude itself.

In establishing the principle that most of the plants with buoyant seeds
or fruits have been gathered at the water-side, it was never implied
that all the plants by the river or by the pond or at the coast are thus
characterised. There is much to learn from the circumstance that whilst
nearly all plants with buoyant seeds or fruits are placed at the
water-side, not all water-side plants have buoyant seeds or fruits. In
the first place, it is to be inferred in the light of what has been said
above that the first determining principle in the selection of a station
is concerned not with the buoyancy of the seeds or fruits, but with the
xerophytic or hygrophytic organisation of a plant. In other words, it is
the fitness or the unfitness of a plant for living in situations where
the loss of water by transpiration requires to be checked that primarily
determines the station at the coast. We thus see in the internal
organisation of the plant the primary determining influence on station.
Buoyancy of seed or fruit comes subsequently into play, the Xerophyte
and the Hygrophyte, thus endowed, ultimately finding their way, the
first to the beach, the second to the bank of the river or to the margin
of the lake or pond.

In the next place, when we regard the composition of the British
coast-flora, and examine the distribution of the plants in other
situations than on the beach, we obtain some interesting results. There
is first a group of plants, including such as Armeria vulgaris,
Artemisia maritima, Cochlearia officinalis, Erodium maritimum,
Matricaria inodora, Plantago coronopus, Polycarpon tetraphyllum,
Raphanus maritimus, Spergularia rubra, Silene maritima (see Note 15),
and others, all of which occur not only at the coast and on the adjacent
hill-slopes, but also often far inland, and sometimes at considerable
elevations in mountainous districts, as in Central Europe. It is on this
occurrence of certain shore-plants in alpine regions that Prof. Schimper
lays much stress in his memoir on the Indo-Malayan strand-flora (p. 28),
and in his later work on Plant Geography (Engl. edit., p. 716), when
pointing out that here temperature does not play a determining part, and
that in both stations, whether on the sandy beach or on the
mountain-top, the same xerophilous organisation is needed to obviate the
risk of impeded water-supply. He quotes in this connection the
observation of Battandier that many alpine species from the Atlas
Mountains occur on the Algerian beaches, but not in intervening regions.
Mr. Druce, in his discussion of the British species of Sea-Thrifts and
Sea-Lavenders (Armeria, Statice), brought the subject of the occurrence
of maritime plants on mountain summits again to the front; but he did
not advance any general explanation, and seems to regard it as the
result, as it doubtless is, of the recurrence of suitable stations
(_Jour. Linn. Soc. Bot._, Dec. 1900).

Very few of these plants have any capacity for dispersal by currents, a
subject dealt with in Note 16. Several of them have dehiscent,
small-seeded fruits which, as pointed out in the previous chapter,
hardly ever come into the buoyant category. I have experimented on the
greater number of them, and in only one species, Matricaria inodora
(var. maritima), do the results indicate a capacity for dispersal over
wide tracts of sea.

If we look again at a list of British shore-plants, we find another
group of plants frequenting salt marshes and muddy shores, and found
also often far inland, as in the saline plains of Central Asia. Here we
have such plants as Aster tripolium, Glaux maritima, Plantago maritima,
Salicornia herbacea, Salsola kali, Samolus valerandi, Scirpus maritimus,
Suæda fruticosa, S. maritima, Triglochin maritimum, T. palustre, &c. It
becomes in this connection a subject of peculiar interest to the student
of plant-distribution when he reads in Mr. Hemsley’s paper on the flora
of Tibet (_Jour. Linn. Soc. Bot._, vol. 35) that amongst the British
shore-plants above-named the two species of Triglochin and the same
species of Glaux and Salsola occur in the salt marshes of the Tibetan
uplands at elevations of 15,000 to 16,000 feet, Scirpus maritimus also
being found in the swamps of the lower levels. We have the same thing,
affecting much the same plants, illustrated in America. Thus we learn
from Asa Gray that Salicornia herbacea, Scirpus maritimus, Triglochin
maritimum, &c., which are common in salt marshes on the coast of the
United States, occur also in the interior of the continent in the
vicinity of salt-springs.

Facts of this sort are well known, and I merely refer to them here in
order to emphasise the importance of this little group of British
littoral plants, those of the salt marsh. Their very wide distribution
is connected with the frequent recurrence of suitable conditions, not
only in space, but what seems of greater import, also in time. One can
scarcely doubt when the Saltwort (Salsola kali) is seen on the
Devonshire coast, on a beach in Chile, and in the elevated regions of
Central Asia that here a very ancient type of plant finds its still more
ancient conditions of existence. In the capacity which most of the
plants of the salt marsh possess of germinating in sea-water, this group
of littoral plants is sharply distinguished, as far as my observations
show, from the other groups of British shore-plants. For instance, in my
experiments the seeds of Aster tripolium, Salicornia herbacea, and
Triglochin maritimum germinated freely in sea-water, whilst those of
Spergularia rubra, Cakile maritima, Convolvulus soldanella and others
failed to do so (see Note 19). It will also be noticed with respect to
this group of littoral plants that, except in the case of Scirpus
maritimus, the seeds or fruits have little or no floating power, the
exception offered by Salsola kali being not very striking. This feature
is brought out in the Table given in Note 10; but some of the details of
my observations are given in Note 17.

There yet remains a third group of the British shore-plants, namely,
that comprising the plants that rarely stray far from the beach and
often possess seeds or seedvessels that will float for months. Here we
have such species as Arenaria (Honckeneya) peploides, Beta maritima,
Cakile maritima, Crambe maritima, Crithmum maritimum, Convolvulus
soldanella, Eryngium maritimum, Euphorbia paralias, Glaucium luteum,
Lathyrus maritimus, Polygonum maritimum, &c. The seeds or seedvessels of
quite half of these species will float for months unharmed in sea-water,
but in a few, as with Cakile maritima and Eryngium maritimum, they float
for only a week or two, whilst in others again like Glaucium luteum they
have no buoyancy. (Some details of the buoyancy experiments on these
plants are given in Note 18; and the long list in Note 10 may be first
consulted.)

It is not necessary to enter here into more detail with respect to
British shore-plants. Enough has been said to disclose cleavage-lines in
what might have appeared as a homogeneous plant-formation. We can thus
discern the elements of at least three groups amongst the plants of our
beaches, each group bearing the impress of an independent history:—

(_a_) The plants of the beach and of the inland plain or of the distant
mountain peak, excluding those of the salt marshes. Armeria vulgaris,
Silene maritima, and Spergularia rubra may be taken as examples. The
currents here as a rule take little or no part in their dispersal.

(_b_) The “saline” group, including the plants of the saline plains and
the salt marshes of the interior of continents. Of these Glaux maritima,
Salsola kali, and Triglochin maritimum are examples. The capacity of
germinating in sea-water is a distinguishing character of most of the
plants; and but few of them possess seeds or seedvessels that are
markedly buoyant.

(_c_) The true beach plants that rarely stray far from the beach, of
which Arenaria peploides, Cakile maritima, and Convolvulus soldanella
are examples. Many of them have buoyant seeds or fruits capable of
dispersion over wide areas through the agency of the currents.

The reader will be able to extend this subject for himself if he is so
inclined, but we have gone far enough together to learn that the plants
with buoyant seeds or fruits are in the minority on our beaches,
scarcely a third of the total being fitted for dispersal by the currents
over broad tracts of sea. The British strand-flora thus differs
strikingly from the littoral flora of a Pacific island, or indeed of any
ordinary tropical coast, and in this respect it is to be regarded as
typical of the temperate regions. It has been remarked before that on a
beach in the tropics we would expect to find that quite three-fourths of
the plants are provided with buoyant fruits or seeds distributed far and
wide over the tropical seas by the currents.

We pass on now to briefly discuss from the same standpoint the British
plants that find their homes on the borders of rivers and ponds. It is
here that the hygrophytes with buoyant seeds or fruits gather together,
just as the xerophytes with similar seeds or fruits collect on the
beaches. We have seen before that only a portion of the beach plants
belong to the buoyant group, and the same applies to the plants at the
edges of rivers and ponds. The plant-formation is no more homogeneous
there than it is in the case of the strand-flora. Let us see if we can
discern some lines of division there also, or in other words let us
endeavour to connect the absence or presence of floating power in the
fruits and seeds with some variations in the placing of the plants. We
still pursue the clue to the study of the complicated problems connected
with plant-stations by taking the floating seed as our guide.

We will carry ourselves in thought to the Thames-side between Teddington
and Twickenham at the end of August, 1892. The river is at the
high-water level, and we see flourishing at the margins, sometimes a
little above the water and sometimes a little within its reach,
Ranunculus repens, R. sceleratus, Spiræa ulmaria, Lycopus europæus,
Scutellaria galericulata, different species of Rumex, Alnus glutinosa,
Iris pseudacorus, Sparganium ramosum, and different species of Carex,
with several other plants, all contributing their seeds or fruits to the
drift that floats in the river from the autumn to the spring.

But besides these plants there are a number more or less submerged in
the stream, including Nasturtium amphibium, N. sylvestre, Stellaria
aquatica, Myosotis palustris, and Veronica beccabunga; and as the water
falls other plants still more submerged come into view on the exposed
flats, such as Nasturtium officinale, Apium nodiflorum, and Polygonum
hydropiper. None of these plants are represented by their seeds or
fruits in the floating river-drift. Several of them possess dry
dehiscent fruits with small seeds, such as Nature ignores in the matter
of buoyancy, and the small fruits of Myosotis, Apium, and Polygonum have
little or no floating power.

We have thus here a clear dividing line between the plants with buoyant
seeds or fruits that were more or less exposed above the high-water
level, and those that were more or less submerged at that state of the
tide. That which occurs in the Lower Thames twice in the day within the
reach of the tide represents what happens in the higher part of the
river during the seasonal floods, but in the last case the effects
cannot be so readily distinguished. We thus perceive that the buoyant
seed or fruit is as a rule only characteristic of the plants of the
river-side that grow more or less exposed above the water, whilst those
plants liable to periodic submergence have seeds or fruits that sink.

In this connection it is of especial interest to observe that as a
general rule the truly aquatic plants of English rivers contribute
little or nothing to the floating seed-drift. I pointed this out several
years ago, in my paper on the Thames, as an agent in plant-dispersal,
and it has been already noticed in this work (page 30). We look in vain
amongst the floating winter drift of our rivers for the seeds or fruits
of Ranunculus aquatilis, Nuphar luteum, Nymphæa alba, and of the species
of Myriophyllum, Limnanthemum, Callitriche, Ceratophyllum, Zannichellia,
and of several of the Potamogetons, all of which give character in
summer to the aquatic vegetation of the river. In their place we find
only the seeds and fruits of the plants growing on the banks.

There is, however, another small group of river plants, which in their
structure and habits and in the behaviour of their floating fruits come
between the true aquatics and the plants of the river-banks. They belong
mostly to the Alisma family, and Alisma plantago and Sagittaria
sagittifolia may here be specially mentioned. Their fruits display great
variation in their floating power; and on this point M. Kolpin-Ravn,
writing to me in 1895, made the following interesting suggestion, that
since these plants approach true aquatics in structure they may be also
regarded as approaching them in the inconstancy of the buoyant capacity
of their fruits, those of aquatics having typically little or no
floating power.

Seed-buoyancy, however, does not play quite such an important part in
the plant-economy of a river as the examination of the floating drift
would lead one to expect. Only a portion of the bank-plants have buoyant
seeds or fruits, whilst amongst the true aquatics, the semi-aquatics,
and the plants periodically submerged, the rule of non-buoyancy
prevails. And, indeed, when we look at all the possible stations for the
plants of the British flora, we discover that seed-buoyancy can rarely
be connected with station. It is, however, in those few stations that
plants with buoyant seeds have mainly gathered. There it is, probably,
that the remnants of a past floral age find a refuge, since it would
seem likely that the tendency has been in the course of geological time
for the development of dry stations for plants at the expense of the wet
stations.

The following is a summary of some of the points discussed in this
chapter:—

(1) In the case of the strand-flora of a Pacific island, and indeed in
that of an ordinary tropical region, the large proportion of plants with
buoyant seeds or fruits tends to mask all other issues, and we are
seemingly only concerned with dispersal by currents.

(2) But in the British strand-flora where plants with buoyant seeds and
fruits are in a minority, constituting less than a third of the total,
it is seen that the issue is primarily an affair of station, an
inference that may be applied generally to temperate regions.

(3) All British shore-plants may be regarded as owning certain
characters in common which may be collectively designated the
xerophilous habit, and we may extend this view to other temperate
strand-floras.

(4) But this xerophilous habit is also characteristic of inland plants
in certain localities, as of those of the steppe, the desert, the rocky
mountain-top, and of other exposed situations, in all of which checks to
the loss of water by transpiration are required. Whilst the risks of
drought are thus guarded against in the case of plants stationed in arid
localities, the risk of injury to the plant from the accumulation of
salt in the tissues is obviated in the instance of the plants of the
coast.

(5) On the other side we have the hygrophilous habit characteristic of
plants living under conditions where checks to transpiration are
relatively little needed. All the plants of the margins of rivers and
ponds belong here, and indeed all plants living under moist conditions.

(6) This distinction between the xerophilous and hygrophilous habits
penetrates deeply into all questions connected with stations, and lies
behind all matters relating to the buoyancy of seeds or fruits. It is
the fitness or unfitness of a plant for living in dry situations that
primarily determines the station. If a xerophilous plant has a buoyant
seed or seedvessel it finds its way ultimately to the coast; if it is
hygrophilous and its seeds or fruits can float, then it is finally
established on the side of a pond or river.

(7) The composite character of the British strand-flora is to be
explained on the above principles. We have in the first place the plants
confined to the sandy beach, many of which possessing buoyant seeds or
fruits are dispersed by the currents. Next come the plants of the sandy
beach which are found also far inland in open plains and on
mountain-tops; and afterwards come the plants of the salt-marsh and
mud-flats of the coast, which appear again in the saline plains and
swamps in the interior of the continents.

(8) The plant-formation of the river’s border displays also lines of
division, and is by no means homogeneous; and indeed other factors
besides those connected with seed-buoyancy have here been in operation.

(9) In only a few of the possible stations of British plants can a
direct connection be traced with seed-buoyancy. Yet it is at these few
stations, such as at the coast and by the pond or river, that the plants
with buoyant seeds and fruits have mainly gathered.

(10) The plants now frequenting wet stations may often be regarded as
the remains of an age when moist conditions for plant-life prevailed.

CHAPTER V

THE FIJIAN STRAND-FLORA

The inland extension of the beach plants.—The grouping of the coast
plants.—Their modes of dispersal.—The zone of change.—Summary.

HAVING learned from the British flora the real significance of the
buoyant seed or fruit in a littoral flora, we will now return to the
Pacific and proceed to deal with the composition and general character
of the strand-plants.

Speaking of the Malayan strand-plants, Professor Schimper remarks (pp.
11, 12) that both in outward appearance and in anatomical structure they
are xerophilous in character, whether in the case of those of the
mangrove-swamp or in those of the beach. Since the tropical shore-flora
of the Pacific islands is essentially Malayan, the identity usually
extending to the species, the same conclusion may be applied to its
character. The xerophilous habit may show itself externally in a variety
of ways, as in hairiness, leaf-structure, a leathery cuticle,
succulency, &c.

From this xerophilous habit of the Pacific strand-flora we should expect
to find that many of the plants stray far from the coast, wherever the
suitable conditions for their type of organisation occur, whether in the
inland plain or on the mountain-top. This is indeed the case; but in
dealing with this subject it will be necessary to discuss in some
general detail the littoral floras of the Fijian, Hawaiian, and Tahitian
groups in succession.

THE FIJIAN STRAND-FLORA

THE INLAND EXTENSION OF THE BEACH PLANTS

Viewed from the old standpoint of “station,” where one would distinguish
sharply between the coast and the inland plants, the Fijian strand-flora
exhibits a number of inconsistencies, all at first sight extremely
puzzling. When, however, we regard their xerophilous character and
reflect that this habit, and not mere fitness for growing at the coast,
is the primary determining factor of their station, much that is strange
appears normal and plain.

Let me refer in this connection to the impression that the distribution
of the Fijian shore-plants made on Mr. Horne, the director of the
Botanic Gardens of Mauritius, who spent a year in the botanical
investigation of the group about a quarter of a century ago. In his
account of the group (pp. 59, 60) he says that several of “what are
known as sea-shore plants” are found far in the interior of the larger
islands; and amongst others he names such characteristic beach plants as
Cerbera Odollam, Hibiscus tiliaceus, Ipomœa pes capræ, and Pandanus
odoratissimus. On the other hand, he remarks that several species of
inland plants occur at the coast, and that several plants growing on the
mountain-tops are found near the sea. This apparent confusion of station
he seems to attribute to the circumstance that the mountains of Fiji are
not high enough for the development of an alpine flora. But such a view
could not be held now, since the effect of an alpine flora would be the
introduction of further elements of confusion in the occasional
occurrence of some of the alpine plants on the sea-coast, as we find in
Hawaii.

Yet this apparent mingling of the littoral and inland floras in Fiji
becomes intelligible when we perceive that the seeming confusion of
station is mainly restricted to the xerophilous plants of the arid
inland plains and of the bare mountain-tops. The rank humid forests that
cover so much of the interior of the islands, and the luxuriant
vegetation of the mountain-gorges, are not here concerned. Such a
mingling occurs it is true under certain conditions; but in the general
physiognomy of the flora the distinction between the shore and inland
plants holds good. The same shore plants that are distributed far and
wide over the Pacific here present themselves; and although some of them
extend far inland, where the scantily-vegetated plains descend to the
coast, this does not deprive them of the right of being still regarded
as littoral plants.

Still, when we look at a fairly complete list of the shore-plants of
Fiji, numbering in all about eighty, we perceive that about two-thirds
of them also occur inland, either in Fiji or in some other tropical
region; and if we reflect that many of the residue are plants of the
mangroves that would not be found inland except under estuarine
conditions, it becomes evident that with this reservation there are very
few littoral plants in Fiji that do not at times leave the coast.

Cæsalpinia Bonducella may be taken as a type of those shore-plants that
stray far away from the coast, even into the interior of continents,
since in India it reaches the Himalayas. Although Terminalia Katappa and
Calophyllum Inophyllum often owe their existence inland in different
parts of the tropics to man’s agency, this cannot be said of most
others, as Cassytha filiformis, Casuarina equisetifolia, Cycas
circinalis, Ipomœa pes capræ, Pandanus odoratissimus, Premna tahitensis,
Tacca pinnatifida, Tephrosia piscatoria, Vitex trifolia, &c., when they
occupy the extensive inland plains that slope to the coasts on the lee
sides of the large islands of Fiji. Plants, like Hibiscus tiliaceus, are
found in a Pacific island almost as frequently away from the beach as on
the beach itself; and this is true of most other regions of the tropics
where it occurs.

Other plants that appear to be altogether confined to the sandy beach in
Fiji, break away on rare occasions from their usual station and appear
on the bare rocky summits of hills near the coast, even though the
hill-slopes are densely wooded. On such bare hilltops in Vanua Levu,
varying from 500 to 1,100 feet in elevation, one is surprised at times
to find shore creepers and climbers like Canavalia obtusifolia and
Derris uliginosa associated with other beach-plants more frequently
found inland, such as Tephrosia piscatoria and Vitex trifolia, and in
the company of climbing species of Morinda and of small trees of Fagræa
Berteriana. When the “talasinga” (sun-burnt) districts, as the Fijians
term the plains on the north sides of the islands, extend a long
distance from the coast into the heart of the island, they carry with
them their peculiar vegetation and the intruding beach-plants up to
considerable elevations above the sea. We then find familiar
beach-plants like Cerbera Odollam and Ipomœa pes capræ growing far
inland at heights of 1,000 feet and over above the sea. (See Notes 20
and 21.)

One is never quite sure of the behaviour of shore-plants in Fiji when
the “talasinga” plains lie behind the beach, since even Scævola Kœnigii,
usually a steadfast beach-plant, occurs at times some miles inland. (See
Notes 20 and 55.) There are, however, a few that never came under my
notice inland, such as Pemphis acidula, Triumfetta procumbens, and
Tournefortia argentea. The extension of sea-coast plants for any
distance inland depends a good deal on the occurrence of
scantily-vegetated plains, or of scrub-covered, rolling country at the
back of the beaches; and doubtless that which I have described in the
case of Fiji is to be found in other tropical coast-regions. Professor
Schimper informed me by letter that he had noticed a similar inland
extension of the shore-plants in the Seychelles.... I have only here
touched on this subject. In Notes 20 and 21 the reader will find further
details of the inland extension of the beach-plants, and in Note 22 is
given a general account of the “talasinga” plains, in which the
wandering beach-plants mingle with the peculiar vegetation of the plains
themselves. Covered with reeds and bracken, and dotted over with clumps
of Casuarinas and Acacias, with the Cycad and Pandanus distributed
irregularly over their surfaces, such level districts possess, as
remarked by Seemann, a South Australian look.

THE GROUPING OF THE FIJIAN LITTORAL PLANTS.

The littoral plants readily divide themselves into three principal
groups as concerning their station, namely:

(_a_) The “beach-formation,” typically exhibited on the whitish
calcareous beaches of reef-bound coasts.

(_b_) The “mangrove-formation,” found at intervals all along the coasts,
but most fully developed at the estuaries, and for the most part
occupying flats regularly overflown by the tide.

(_c_) The “intermediate formation,” comprising the plants of the tracts
between the beach and the mangrove-swamp and at the borders of the
swamps.

This grouping does not differ materially from that adopted by Professor
Schimper in the instance of the Indo-Malayan strand-flora. (See Note
23.)

To the beach-formation belong, amongst the trees and shrubs,
Barringtonia speciosa, Calophyllum Inophyllum, Guettarda speciosa,
Pemphis acidula, Scævola Kœnigii, Tournefortia argentea, &c., and
amongst the creepers and procumbent plants, Canavalia obtusifolia,
Ipomœa pes capræ, Triumfetta procumbens, &c. To the mangrove-formation
belong the Asiatic and the American species of Rhizophora, and species
of Bruguiera, Carapa, Lumnitzera, &c. Amongst the trees that gather
around the borders of the mangrove-swamp, constituting the intermediate
formation, occur Barringtonia racemosa, Excæcaria Agallocha, Heritiera
littoralis, Hibiscus tiliaceus, and several other species, all of them
being equally at home on the sandy beach, at the border of a
mangrove-swamp, and on the banks of an estuary. The climbers, such as
Entada scandens, Mucuna gigantea, Derris uliginosa, &c., belong more to
the mangrove and to the intermediate formations than they do to that of
the beach. Referring the reader to the more complete lists given in Note
24, I may remark that it is not always possible to distinguish sharply
between the three formations, since some of the plants belong to two,
and a tree like Cerbera Odollam may, in different localities, be
referred to all three formations. The general distinction, however,
prevails in the physiognomy of the coast-flora.

The mangrove-formation comprises, it may be pointed out, many plants
other than mangroves, plants that find a home in the mangrove-swamps of
Fiji, either within their limits or at their borders. It presents,
indeed, a world in itself. When the mangroves establish themselves in a
new locality they carry along with them a host of hangers-on, both
plants and animals, that only find a home under the favourable
conditions of a mangrove-swamp. Thus, the absence of the
mangrove-formation from a Pacific island deprives its littoral flora of
many very striking features. For this reason the Tahitian shore-flora
must seem to a botanist coming from Fiji comparatively tame and
monotonous; whilst that of Hawaii, for this and for other reasons to be
subsequently mentioned, is still less interesting, and scarcely even
gives a character to the coasts.

We are now, therefore, prepared to learn that a large number of the
plants other than true mangroves, that thrive in or around the Fijian
mangrove-swamp, are not to be found in those Polynesian islands where
true mangroves do not exist; and that a law of association here exists.
Many of the plants of the intermediate formation are so closely bound up
with the mangroves in their life-conditions that they are not to be
found where the mangroves are absent, even though their seeds or fruits
are pre-eminently fitted for dispersal by the currents. The influence of
“station” here rules supreme. This matter will be treated more in detail
when discussing the Tahitian and Hawaiian strand-floras in Chapters VI.
and VII.

THE MODES OF DISPERSAL OF THE FIJIAN STRAND-PLANTS.

The predominant influence of the currents having been already
established, there remains for consideration the distribution of the
floating capacity of the seeds or fruits among the different formations.
One can say that almost without exception the seeds or fruits or
seedlings of the mangrove and intermediate formations float for long
periods. In the case of some of the true mangroves, as in Rhizophora and
in Bruguiera, where germination takes place on the tree, it is the
seedling that floats, whilst in others, as in Carapa and Lumnitzera, it
is the seedvessel that floats. The plants with non-buoyant seeds or
fruits that belong to the littoral flora are all confined to the beach
formation, but they do not form more than a sixth of the total. Almost
all the “good floaters” of the beach-plants are widely spread over the
shores of the Pacific and of much of the tropics, and include such
familiar species as Barringtonia speciosa, Cæsalpinia Bonducella,
Terminalia Katappa, and many others mentioned in the lists of Notes 2
and 24.

When, however, we come to the dozen or so of beach-plants that possess
seeds or fruits with little or no floating power, we find that several
of them have a limited distribution in the Pacific, such as Acacia
laurifolia, Drymispermum Burnettianum, Eugenia Richii, &c., whilst
others, such as Casuarina equisetifolia, Tephrosia piscatoria,
Triumfetta procumbens, and Wikstrœmia fœtida, are widely spread. This
small non-buoyant group of the beach-plants has a nondescript
appearance, and it is here that the inland flora is most likely to make
its influence felt by additions to the number. It is here indeed that
the littoral floras of the tropics mostly differ, the accessions from
the inland flora varying in each region. It is in fact the zone of
change.

A number of these plants, such as the species of Drymispermum, Eugenia,
and Wikstrœmia, have probably been dispersed by frugivorous birds;
whilst others, like Triumfetta procumbens, possess fruits that might
have been transported in birds’ plumage. From the frequency with which
Tephrosia piscatoria is associated on hilltops in Fiji with Fagræa
Berteriana and climbing species of Morinda that are well suited for
dispersal by frugivorous birds, it seems likely that it is also
distributed by birds fond of a drier diet. It is possible that the
Polynesians, who much value the wood of Casuarina equisetifolia, have
often assisted in dispersing the tree.

The following is a summary of the contents of the chapter.

(1) The extension inland of the Fijian strand-flora is to be attributed
to the xerophilous organisation of the plants, and to the exceptionally
favourable conditions that are offered to such plants on the plains, and
in other scantily vegetated localities, lying usually on the drier sides
of the larger islands.

(2) Excluding the mangroves and the plants associated with them in the
coast-swamps, there are few littoral plants of the islands of the
tropical Pacific that do not extend inland in one region or another.

(3) The Fijian shore-plants can be rudely arranged in three groups,
those of the mangrove-swamp, those of the sandy beach, and those of the
intermediate districts, the last including those plants that occur
typically at the borders of a mangrove-swamp, though some of them can
thrive equally well on a beach.

(4) There is a law of association connecting many plants with a
mangrove-swamp in such a manner that when the true mangroves are not
represented in a Polynesian group, as in Tahiti or in Hawaii, the plants
in question are also absent, notwithstanding that in many cases, such as
those of Clerodendron inerme and Heritiera littoralis, they possess
seeds or seedvessels of great floating power.

(5) The fruits or seeds or seedlings, as the case may be, of the plants
of the mangrove-swamp and of the bordering districts float almost
without exception for long periods. This is true also of five-sixths of
the beach-plants, whilst the remainder owe their dispersal chiefly to
birds.

(6) The small non-buoyant group of the beach-plants represents that
portion of the strand-flora that is most likely to be recruited from the
inland flora. It is here that exists the zone of change; and it is in
this respect that the littoral floras of the tropics differ principally
amongst each other, the recruits from inland varying naturally with the
floras of different regions.

* * * * *

Though it does not come within my plan to discuss the littoral floras
of the adjacent smaller groups of Tonga and Samoa, it may be remarked
that they reflect most of the principal features of the strand-flora
of Fiji. In particular it may be observed that they possess the
mangrove-formation, but to a more limited extent. Both own the
mangrove genera Rhizophora and Bruguiera, whilst Carapa is also found
in Tonga. The intermediate formation is represented in Tonga by
Clerodendron inerme, Excæcaria Agallocha, and Heritiera littoralis;
whilst in Samoa we find, besides the first-named species, Barringtonia
racemosa and Scirpodendron costatum. In both the beach-formation is
well represented.

CHAPTER VI

THE TAHITIAN STRAND-FLORA

(_From materials supplied mainly by the work of Drake del Castillo_)

Lacks the mangroves and their associated plants.—Possesses mainly the
plants of the coral beach.—Predominant agency of the currents.—Inland
extension of shore-plants.—Summary.

JUST as the littoral plants of Fiji may be regarded as typical of
Western Polynesia, so the strand-flora of Tahiti, or, rather, of the
Tahitian Islands, may be considered as representing Eastern Polynesia.
We have thus the Tahitian area, comprising generally the Cook and
Austral Groups, the Society Islands, the Paumotus, and also the
Marquesas, as contrasted with the Fijian area, including the
neighbouring Samoan and Tongan groups. For the sake of brevity the terms
Fiji and Tahiti are often used as equivalents of the entire areas (see
Note 25).

The littoral flora of this part of the Pacific lacks the mangroves and
most of the plants that are associated in the Fijian region with a
mangrove-swamp, either at its borders or within its interior. Thus we
miss here the true mangroves of the genera Rhizophora, Bruguiera,
Carapa, and Lumnitzera, as well as the accompanying trees and shrubs,
such as Barringtonia racemosa, Excæcaria Agallocha, and Heritiera
littoralis. The climbers and straggling plants that are so
characteristic of the borders of the mangrove-creeks in Fiji proper are
also wanting, such as Clerodendron inerme, Derris uliginosa, and Smythea
pacifica; and we do not find in the Tahitian region the Giant-Sedge
(Scirpodendron costatum) that is so common in the mangrove-swamps of
Fiji, and occurs also in Samoa.

It is not at first sight easy to account for the absence from Tahiti of
the mangrove-formation and of so many of the plants that grow at the
borders of a mangrove-swamp in Fiji. Their absence can scarcely be due
to the want of suitable stations, as is indicated by the common
occurrence in the Tahitian coast-marshes of Chrysodium aureum, the Great
Swamp-fern, that not only abounds in the mangrove belts of Fiji, Tonga,
and Samoa, but is associated with mangrove-swamps over much of the
tropical zone. Nor can it be said that the currents are ineffective, or
that the seeds or fruits of the missing plants possess, as a rule,
insufficient floating powers. Most of the plants of the Tahitian beaches
hail, like those of Fiji, from Malaya, and have been brought through the
agency of the currents; and many of the absent littoral plants that have
the same home, such as Heritiera littoralis and Clerodendron inerme,
have fruits or seeds just as capable of floating unharmed over the same
extent of ocean. It is not any defect in floating-power that has
prevented the establishment of two such plants in the Tahitian area.
Entada scandens, which in some parts of the world is a typical climber
of the mangrove-formation, and in other places thrives well in the
absence of mangrove-swamps, has only been recorded from Rarotonga in
this region by botanists, but I believe Wyatt Gill refers to its
occurrence in Mangaia in one of his books.

On the other hand, it is likely that the floating seedlings of
Rhizophora and Bruguiera, which represent the only means of dispersal by
the currents at the service of these mangroves, would not arrive at
Tahiti in a condition favourable for the establishment of the plants. My
observations, which are described in Chapter XXX., go to show that,
though the seedlings will float uninjured in still sea-water for months,
they will not withstand prolonged sea-buffeting. These two genera of
mangroves, it is most important to remember, supply the pioneers and the
principal components of a mangrove-swamp in the Western Pacific. Where
they fail to establish themselves, the requisite conditions for the
large number of plants and animals that find their home in and around a
mangrove-swamp would not be provided. We thus perceive that the absence
from the Tahitian coast flora of several plants that are associated in
Fiji with the mangrove-swamps depends on a law of association, which has
already been referred to in the preceding chapter, and is not concerned
with incapacity for dispersal by currents (see Note 26).

Whilst the Tahitian coast flora does not, therefore, possess the plants
of the mangrove-swamp and its vicinity, it includes most of the typical
beach-trees of the coral islands and reef-fronted coasts of other parts
of the South Pacific. Thus we find here on the sandy beaches
Barringtonia speciosa, Calophyllum Inophyllum, Cerbera Odollam,
Hernandia peltata, Guettarda speciosa, and numerous other plants that
are indicated by the letter T in the list of Fijian littoral plants
given in Note 2. The total number of Tahitian shore-plants is thus
considerably less than that of Fiji (there are about 55 in Tahiti and
about 80 in Fiji); but in its turn, as will subsequently be shown, it is
much larger than that of Hawaii, where the number is about 30.

Quite three-fourths of the strand-flora of this region have buoyant
seeds or seedvessels capable of floating for long periods; and there is
no difficulty in assigning by far the greater share in the stocking of
the beaches with their plants to the agency of the currents. The
currents in their operations have indeed carried the fruits or seeds of
many of these plants across the South Pacific as far as the islands
extend, namely, to Ducie Island and to Easter Island. There are few more
significant proofs of the efficacy of the currents in distributing
plants over the Pacific than the discovery, by Mr. Arundel, of
Barringtonia speciosa in Ducie Island in association with Tournefortia
argentea (Challenger, Botany, III. 116).

The residue of the Tahitian coast flora possessing fruits or seeds that
are unsuited for dispersal by currents includes such plants as
Heliotropium anomalum, Triumfetta procumbens, Tephrosia piscatoria,
Wikstrœmia fœtida, &c. The small nucules of the first-named are perhaps
dispersed by granivorous birds; the fruits of Triumfetta are probably
transported in birds’ plumage; those of Wikstrœmia are distributed by
frugivorous birds; and the seeds of Tephrosia may be dispersed like
those of Heliotropium.

The recruits or intruders from the inland flora do not appear to be
numerous or to give any special character to the shore flora. (See Note
27.)

From not having a personal acquaintance with this region it is not
possible for me to discuss the extension of the shore-plants inland
except in a general way. From the pages of the work of Drake del
Castillo we can, however, infer that several plants such as Cassytha
filiformis, Cerbera Odollam, Colubrina asiatica, Hernandia peltata,
Morinda citrifolia, and Pandanus odoratissimus have extended inland to
the mouths of the Tahitian valleys, and have ascended the lower slopes
of the hills that lie near the coast. Others, like Cæsalpinia Bonduc,
Gyrocarpus Jacquini, and Ochrosia parviflora, have climbed far up the
mountain-sides to elevations of from 2,000 to 2,400 feet above the sea.
It is also evident from Mr. Cheeseman’s memoir on the Rarotongan flora
that coast plants also stray inland in that island. In an island like
Rarotonga, where a sorry substitute for a mangrove-swamp exists in the
form of a few coastal muddy places occupied by Vitex trifolia and
Sesuvium Portulacastrum, Entada scandens takes to the hills; and thus it
is that in this island it is most abundant in the interior, climbing to
the tops of the highest trees and “covering acres of the forest with a
dense canopy of green.”

_Summary of the Chapter._

(1) The Tahitian region possesses most of the plants that frequent the
sandy beaches of the Pacific islands.

(2) But it lacks the mangroves and the associated plants of the
mangrove-swamp.

(3) It also wants many of the plants that grow in the vicinity of such
swamps.

(4) But since the plants last-mentioned often possess fruits or seeds
capable of being carried great distances by the currents, their absence
is to be attributed to the necessary conditions being lacking on account
of the failure of the mangroves.

(5) Most of the beach plants, however, owe their existence in this
region to the transport of their buoyant fruits or seeds by the
currents.

(6) The negative features of the Tahitian strand-flora are mostly to be
connected with the absence of Rhizophora and Bruguiera, the pioneers of
the mangrove-swamp; and their absence is, in turn, to be attributed to
the inability of their floating seedlings to reach this region in a fit
condition for establishing themselves.

CHAPTER VII

THE HAWAIIAN STRAND-FLORA

Its poverty.—Its negative features.—Their explanation.—The subordinate
part taken by the currents.—The Oregon drift.—The inland extension of
the beach plants.—Summary.

COMPARED with the rich strand-flora of Fiji, that of Hawaii presents but
a sorry aspect. In the number of species (30) it does not amount to
half; whilst it lacks the great mangrove-formation and the luxuriant
vegetation accompanying it that gives so much character to the shores
and estuaries of Fiji. Strangely enough, it is also deprived of most of
the familiar trees that, whether in foliage, in flower, or in fruit,
form the chief attraction of the sandy beaches of the Pacific islands.

Neither the mangroves, therefore, nor the plants of the intermediate
formation, are to be found in Hawaii; and when we reflect that the
absentees from the beach formation include most of the trees, under the
shade of which the visitor to the Pacific islands can nearly always find
protection from the fierce rays of a tropical sun, it cannot be a matter
of surprise that this littoral flora has such a poverty-stricken
appearance. We look in vain for such shady beach trees as Barringtonia
speciosa, Terminalia Katappa, and Hernandia peltata; and we are lucky if
we find some small trees under which we can obtain a scanty shade.

I have been speaking, of course, of the indigenous shore-plants, those
that have arrived at these islands without the assistance of man. Yet it
must be added that the existing littoral flora does include some of the
missing indigenous trees, though rarely in any number. There is,
however, scarcely one of them that is regarded by Dr. Hillebrand as
having formed part of the original flora. That botanist would indeed rob
the present beach flora, scanty as it is, of most of its conspicuous
plants, as far as their claims to be considered indigenous are
concerned. Dr. Hillebrand indeed includes Calophyllum Inophyllum,
Hibiscus tiliaceus, Thespesia populnea, Morinda citrifolia, Cordia
subcordata, and Pandanus odoratissimus in the present Hawaiian flora,
and nearly all of them are to be found at times at the coast as well as
inland; but he regards all, excepting the last-named, as having been
introduced by the aborigines. I was not inclined at first to go quite so
far as Dr. Hillebrand in this direction; but he carefully considered the
case of each individual plant, and, remembering his sojourn of twenty
years in the islands, his authority cannot be lightly put aside. In the
list of Hawaiian strand-plants given in Note 28 there are several
species not always littoral in the group, but typically littoral in
other tropical regions. One species, Ipomœa glaberrima, Boj., has not
been recorded before from these islands.

A strong reason in favour of the contention of this botanist is that all
the trees above-named are useful in some way to the natives; and,
indeed, when we look at the works dealing with the floras of the islands
of the South Pacific, we observe that in almost all the groups one or
other of these six trees bears the reputation of having been introduced
by the aborigines. All of them in their turn lose their fame as truly
indigenous plants in some group or other. The occurrence of two or three
useless South Pacific beach trees, that are known to be dispersed by the
currents, in the indigenous strand-flora of Hawaii, would go far to
invalidate Dr. Hillebrand’s argument, since the six trees in dispute are
also known to be dispersed by the currents. But such trees are not to be
found; and we look in vain for trees like Cerbera Odollam, Guettarda
speciosa, Gyrocarpus Jacquini, and Hernandia peltata, that are spread
far and wide over the beaches of the South Pacific.

It is also of interest to notice how trees like Morinda citrifolia and
Terminalia Katappa, concerning the non-indigenous character of which
there can be but little doubt, are in our own day acquiring a littoral
station. The second is not even regarded by Dr. Hillebrand as having
been introduced by the natives, but is referred by him to the European
epoch. After having been extensively planted, it is now, as I found,
becoming a littoral tree on the coast of Oahu, and supplies its buoyant
fruits in a regular way to the beach drift. Its native name of Kamani is
merely that of Calophyllum Inophyllum. All the six trees in dispute are
known in Hawaii by the names by which they are distinguished far and
wide over the South Pacific, a fact of which the reader may satisfy
himself by referring to my paper on Polynesian plant-names. The
Hawaiians, when their ancestors abode in the South Pacific, must have
been well acquainted with one or other of the prevailing names of
Terminalia Katappa (Talie, Tara, &c.); but it had lapsed in the memory
of the race when the Europeans introduced the tree into Hawaii.

It may be added in this connection that Dr. Hillebrand weakens his
argument by regarding Pandanus odoratissimus as of pre-aboriginal origin
or as truly indigenous. Like the other six trees in question, its fruits
are known to be capable of dispersal far and wide by the currents; and
if this species of Pandanus is indigenous, we are obliged to assume that
its fruits were first brought by the currents. That being so, we cannot
exclude the probability of the currents having been also effective with
several of the other plants regarded by Hillebrand as of aboriginal
introduction, more especially those with large fruits like Calophyllum
Inophyllum, and Cordia subcordata, where the alternative agency of
frugivorous birds would be impracticable, at least over a wide extent of
ocean. Pandanus odoratissimus is, as I venture to think, a tree that was
introduced ages since by the aborigines. Next to the Coco palm, few
trees have been more utilised by island-peoples, more particularly
perhaps in the ruder stages of their history.

This point has been discussed at some length, because on the correctness
of Dr. Hillebrand’s view depends the explanation to be subsequently
given of the origin of the shore-flora of Hawaii. Though differing in
some details, my observations on the Hawaiian coast plants, which are
given in Note 29, tend to strengthen his contention.

I now return to the consideration of some of the negative features of
the Hawaiian strand-flora, and will allude first to the absence of the
mangroves and of the numerous other plants that live in and around a
mangrove-swamp. This cannot be connected with a total absence of
suitable stations. Although it is true that there are but few large
rivers and but few suitable localities, yet such localities exist. The
shores of Hilo Bay might readily have been the home of a mangrove-swamp;
and one can point to different places on the coast of Oahu, such, for
instance, as Pearl Harbour, which in Fiji would have been occupied by a
luxuriant growth of mangroves. The same argument applies to the missing
beach trees, such as Barringtonia speciosa, Hernandia peltata, Guettarda
speciosa, &c., that adorn the beaches of many a coral island or of many
a coral-bound coast in the South Pacific. Although in a large island
like Hawaii with its lava-bound coasts but few white calcareous beaches
exist where we might expect to find such a flora, yet such beaches occur
wherever the scanty coral reefs are found off the coast; and it is just
in those localities, as is pointed out in the account of my observations
in Note 29, that the “plantes madréporiques” of the French botanists,
the plants of the coral atoll and of the reef-girt coast, make their
best endeavours to establish themselves. In other islands like Oahu,
where coral reefs are more developed, calcareous beaches are more
frequent, and there the few “madreporic” plants of Hawaii make a home.

Nor can the deficiencies in the Hawaiian strand-flora be connected with
climatic conditions. That its meagre character cannot be so explained is
indicated by the manner in which the Indo-Malayan shore-plants have
pushed their way northward on the western side of the Pacific to the
Liukiu and Bonin Islands. Here in latitude 26-27° N. we find several
Fijian littoral trees and shrubs, such as Hernandia peltata, Pemphis
acidula, Pongamia glabra, Sophora tomentosa, Terminalia Katappa,
Tournefortia argentea, &c., that do not occur in Hawaii, although this
group is some degrees nearer the equator, namely, in latitude 19-22° N.
They are accompanied by the mangroves (Rhizophora, Bruguiera, &c.) in
strength as far as South Liukiu in latitude 25° N.; but we learn from
Dr. Warburg that the mangroves thin off further north, though they reach
to South Japan, where Döderlein found in latitude 32° N. solitary
examples of Rhizophora mucronata. These interesting facts of
distribution, which are taken from Schimper’s work on the Indo-Malayan
shore-plants (pp. 85, 90), show us that we can scarcely look to climatic
conditions for the explanation of the absence of mangroves and of many
other tropical littoral plants from Hawaii. We form the same opinion
when we regard the extension northward of the mangrove-formation on the
American coasts of the North Pacific Ocean. According to the account of
Dr. Seemann given in the “Botany of the Voyage of H.M.S. _Herald_,” the
mangroves with the coco-nut palm, and many other littoral plants common
on the western shores of tropical America, reach their northern limit a
little north of Mazatlan within the mouth of the Gulf of California in
latitude 24° 38ʹ N. The parallel of 25° N. latitude, as indicated in
Drude’s Atlas, probably represents the extreme northern limit, which is
thus five or six degrees north of the latitude of the large island of
Hawaii.

Neither can the explanation be found in the deficient floating powers of
the seeds or seedvessels of many of the “absentees.” Those of
Barringtonia speciosa, Guettarda speciosa, Heritiera littoralis, the two
species of Terminalia, &c., possess great buoyant powers equal to, and
probably often exceeding, those of the plants that, like Ipomœa pes
capræ, have succeeded in establishing themselves in Hawaii. One has only
to look at the lists giving the results of flotation experiments in
Notes 2 and 3, in order to realise that there are very few of the
“absentee” littoral plants, the non-existence of which in Hawaii could
be attributed to deficient floating powers of the fruit or seed. Being
able to float unharmed for months, and in several cases even for years,
the seeds or fruits of the shore-plants unrepresented on the Hawaiian
beaches have been carried far and wide by the currents over the tropical
Pacific even to Ducie and Easter Islands, that is, as far as the islands
extend.

The only plants about which one could express a doubt concerning their
ability to reach Hawaii through the agency of the currents, and to
establish themselves there, are the true mangroves of the genera
Rhizophora and Bruguiera. Since germination takes place on the tree, it
is only through the floating seedlings that they could reach these
islands; but, as shown in Chapter XXX., it is doubtful whether the
seedlings would be in a fit condition for reproducing the plant after
such a long oceanic voyage. If they had been as successful in
establishing themselves in Hawaii as they have been in the Liukiu
Islands, which lie in latitude a few degrees farther north, these two
species through their reclaiming agency would alone have prepared the
way for the whole mangrove formation. We have seen in the preceding
chapter that the absence of the mangrove formation from Tahiti appears
to be mainly due to the failure of the pioneer species of Rhizophora and
Bruguiera to establish themselves there. This evidently also applies to
Hawaii, the cause of their exclusion being connected neither with
climate nor with station, but as in Tahiti with the general unfitness of
the floating mangrove seedlings for crossing broad tracts of ocean
without injury to the growing plantlet.

With regard, however, to the bulk of the “absentee” littoral plants,
those of the beach-formation, no such incapacity on the part of the
buoyant seed or fruit can be accepted. These plants, which have reached
Tahiti in numbers, have in the mass failed to reach Hawaii. It will,
therefore, be of interest to glance at the character of the fruits of
the “absentee” trees, which a traveller fresh from a visit to the coral
islands and reef-girt coasts of the South Pacific sadly misses on the
Hawaiian beaches. We notice in the first place that the absent trees,
such as Barringtonia speciosa, Cerbera Odollam, Guettarda speciosa,
Heritiera littoralis, Terminalia Katappa, &c., have large fruits which
could only have been carried to Hawaii by the currents, the agency of
birds being quite out of the question. On the other hand, almost all the
littoral plants of Hawaii, whether trees, shrubs, or herbs, which are
regarded as truly indigenous by Mann, Hillebrand, and other Hawaiian
botanists, have only small fruits or seeds available for dispersal, from
which the agency of birds cannot, on the point of size, be excluded.
Amongst these shore plants possessing buoyant seeds or fruits are
Cassytha filiformis, Colubrina asiatica, Ipomœa pes capræ, Scævola
Kœnigii, Vigna lutea, and Vitex trifolia; whilst amongst the plants with
non-buoyant fruits or seeds are to be reckoned Heliotropium anomalum, H.
curassavicum, Tephrosia piscatoria, Tribulus cistoides, &c. The seeds or
seedvessels of the plants of the buoyant group possess great floating
powers; and it seems at first sight scarcely credible that the currents
which have failed to establish Barringtonia speciosa, Guettarda
speciosa, and the other trees that through this agency have often found
a home on the remotest islands of the Pacific, should have succeeded in
the instances of plants like Scævola Kœnigii and Vitex trifolia.

It would indeed almost seem that in nearly all cases where it would be
impossible in point of size for a bird to transport the fruit or seed of
a shore-plant to Hawaii, such a plant is not to be found in the
strand-flora of that group, even though it is well adapted for dispersal
by the currents. Many of the littoral trees missing from the Hawaiian
coast-flora, having large buoyant fruits, come into this category; and
grave suspicion is thus apparently cast on the agency of the currents in
the case of the plants with small fruits and seeds that really compose
the strand-flora, even when their capacity for sea-transport has been
well established by observation and experiment. The efficacy of the
currents would thus seem to be called into question for the whole
littoral flora of Hawaii.

If, however, we were to adopt such a sweeping conclusion we should be
led into an error. It is pointed out in the following chapter that
nearly all these large-fruited beach trees that are found far and wide
over the South Pacific, but are absent from Hawaii, do not occur as
indigenous plants in America. If, therefore, the fruits of such Old
World littoral trees as Barringtonia speciosa, Cerbera Odollam,
Guettarda speciosa, Ochrosia parviflora, Terminalia Katappa, &c., that
could be dispersed only by the currents, have failed to reach Hawaii, it
is essential to remember that they have also failed to reach America.
This suggests that Hawaii may have received some of its littoral plants
from America through the agency of the currents; and it is shown in the
following chapter that, as a rule, when a South Pacific plant with
buoyant fruits or seeds is not found in America, it is equally absent
from Hawaii. The question thus acquires quite a different aspect, and we
shall accordingly have to regard tropical America in the next chapter as
a possible centre of diffusion of littoral plants over the globe, a
centre possibly as important as that connected with the tropics of the
Old World.

Although, however, the currents have played a part in stocking the
Hawaiian beaches with their plants, their share in the work has been
unimportant, and the number of plants concerned is limited. If we take
away the seven or eight littoral plants introduced by the aborigines, as
well as the three endemic species as indicated in the list in Note 28,
and then remove from the residue the plants with small fruits or seeds
possessing little or no buoyancy, there remain only the following eight
species, the presence of which in Hawaii might be attributed to the
currents, namely, Cæsalpinia Bonducella, Cassytha filiformis, Colubrina
asiatica, Ipomœa glaberrima, Ipomœa pes capræ, Scævola Kœnigii, Vigna
lutea, and Vitex trifolia. Of these plants, three species, those of
Cassytha, Scævola, and Vitex, possess fruits that would be likely to
attract frugivorous birds, and are in some cases known to be dispersed
by them (see Chapter XIII.); so that we are not in these instances
restricted to the agency of the currents. With the other five the
currents offer the readiest explanation, but, as is indicated in the
cases of Cæsalpinia Bonducella and Ipomœa glaberrima (Chapter XVII.), it
is quite possible that birds have occasionally intervened. Altogether we
may infer that in stocking the Hawaiian beaches with their littoral
plants the currents have taken a subordinate part.

Coming to the Hawaiian littoral plants having seeds or fruits that have
no floating power, we find that they present a motley group. It has been
already remarked that this is the group of shore plants that derives
most recruits from the inland flora, and that it is in this group that
the differences between the shore-floras of tropical regions find their
expression. Yet a very odd collection of plants is here exhibited.
Sometimes the beach-flora is composed in great part of these plants; and
a sorry spectacle is presented by a beach possessing such plants as
Gossypium tomentosum, Heliotropium anomalum and H. curassavicum,
Lipochæta integrifolia, Tephrosia piscatoria, Tribulus cistoides, &c.
Yet to the student of plant-distribution such a motley collection would
be full of suggestiveness. From the circumstance that species of
Cuscuta, Jacquemontia, and Lipochæta, that are peculiar to the Hawaiian
Islands, have made their homes on the beach, he would infer that since
Nature has been compelled to borrow from the endemic inland flora, there
has been some difficulty in stocking the beaches with their plants. The
occurrence of endemic species amongst the strand-plants would be viewed
by him as especially indicating incapacity on the part of the ocean
currents.

Yet in the quantities of drift timber, showing evidence of many months
and probably even of years of ocean-transport, to be seen stranded on
the weather coasts of these islands, the observer discerns undoubted
evidence of the efficacy of the ocean currents. But what he finds are
huge stranded pine logs of “red-cedar” and “white-cedar” from the
north-west coasts of America. He may search the drift for days together,
as I have done, and discover no tropical fruits or seeds except such as
could be supplied by the present Hawaiian flora. The subject of this
drift is especially discussed in Note 30; and it need only be mentioned
here that it is not improbable that, as shown in the next chapter, some
drift may reach Hawaii from tropical America under exceptional
conditions, and that its presence is masked by the Oregon drift.

The agency of the drifting log in carrying small seeds in its crevices
would be effectual in the instance of plants from the temperate coasts
of North America. For example, the nutlets of Heliotropium curassavicum,
which have no buoyancy, might easily be washed, together with sand, into
the cracks of a pine log stranded temporarily on the Oregon coast where
this plant occurs. The _modus operandi_ was brought home to me when
examining the drift brought down by the Chancay River on the coast of
Peru. Here I found this species of Heliotropium growing on the margin of
a swamp near some stranded logs, that would probably be carried out to
sea when the river was next in flood.

It is probable, I may add, that the seeds or fruits of some of the
plants of the non-buoyant group of the Hawaiian littoral flora may be
dispersed in birds’ plumage. For instance, the spiny fruits of Tribulus
cistoides sink in sea-water; but they are well suited for entangling
themselves in birds’ feathers.

It is possible that the hairy seeds of Gossypium tomentosum may have
been thus distributed; but there is much that is enigmatical about this
plant (see Chapter XXVI).

THE INLAND EXTENSION OF THE BEACH PLANTS OF HAWAII.—When we regard the
inland extension of littoral plants in Hawaii, we get fresh indications
of the meagreness of the strand-flora. Several of the species, as
Cæsalpinia Bonducella, Cassytha filiformis, Tephrosia piscatoria, &c.,
show themselves only occasionally on the sandy beaches, though they are
common enough on the old scantily vegetated lava-flows near the coast
and are often found miles inland. Indeed, Dr. Hillebrand not
infrequently in describing the station only gives prominence to the
situation of the plants away from the beaches, and places most of them
on the old lava plains that extend inland from the coast. It is only by
a detailed examination of extensive coast lines in these islands that I
have succeeded in preserving to a small degree their reputation as beach
plants. A few of them behave somewhat strangely in their inland station.
Thus, the seeds of Cæsalpinia Bonducella obtained from various
localities showed no buoyancy in my experiments; and had I not found a
solitary buoyant seed in the stranded drift I should have inferred that
this was a rule without exception.

It is to be remarked that whilst some plants like Scævola Koenigii
occasionally stray a few hundred yards inland on the surface of the old
lava-flows, others like Ipomœa pes capræ and Vitex trifolia, that are
spread far and wide over the inland plains of Fiji, are confined in
Hawaii to the beaches and their immediate vicinity. Some of the plants
like Hibiscus tiliaceus, Morinda citrifolia, and Pandanus odoratissimus,
that are regarded as having been introduced by the aborigines, behave
exactly like indigenous plants in the inland plains; but this is not
necessarily an indication of an indigenous plant in this group, since
the Cactus (Opuntia Tuna) and the Castor-Oil Plant (Ricinus communis)
have spread all over the drier lower regions of the islands, whilst
Aleurites moluccana, the Candle-Nut Tree, which has no means of reaching
these islands without man’s agency, now forms entire woods on the
mountain slopes, usurping the place often of the original forests....
Further details relating to this subject are given in Note 31.

The principal points in the foregoing discussion of the strand-flora of
Hawaii may be thus summed up:—

(1) The indigenous, that is, the pre-aboriginal, strand-flora of this
group lacks not only the mangroves and their associated plants, but also
most of the characteristic beach-trees of the South Pacific, which are
known to owe their wide distribution in tropical regions to the
currents.

(2) The meagreness of the littoral flora is intensified by the tendency
of some of the plants to extend inland and to desert the coasts, and by
the occurrence on the beaches of peculiar species not found outside the
Hawaiian Islands.

(3) The absence of the mangrove formation and of so many of the typical
beach trees of the Pacific cannot be attributed either to the lack of
suitable stations, or to climatic conditions, or to deficient floating
power of the seed or fruit.

(4) As in the case of Tahiti, the mangroves and their associated plants
are lacking because the floating seedlings of Rhizophora and Bruguiera,
the pioneer plants of a mangrove-swamp, have failed to reach Hawaii in a
fit condition for establishing themselves. The numerous plants that
accompany a mangrove-swamp have thus been unable to find a home, though
the buoyant powers of their fruits or seeds are often great.

(5) With the missing beach-trees, however, which possess fruits that can
float for years unharmed in sea-water, no such incapacity is suggested.
Most of them have large fruits, which could only reach Hawaii through
the currents. This absence from the Hawaiian indigenous strand-plants of
most, if not all, of the large-fruited species, where on account of size
the agency of birds is absolutely excluded, is very remarkable; and it
at first seems to throw grave suspicion on the efficacy of the currents
for the whole strand-flora.

(6) It is, however, to be noticed that these large-fruited beach trees
have not only failed to reach Hawaii but have also failed to reach
America. The question thus acquires quite a different aspect, and
America becomes the possible source of most of the Hawaiian plants with
buoyant seeds or fruits.

(7) This subject is discussed in the next chapter; but it is here shown
that at their best the currents have taken but a secondary part in
stocking the Hawaiian beaches with their plants, since many of the
plants have non-buoyant seeds or fruits.

(8) The drift stranded on the shores of the Hawaiian Islands is composed
of logs from the north-west coast of North America. No drift from the
south has been discovered; but it is not unlikely that future
investigators will find some seed-drift from tropical America.

[Illustration:

THE WORLD
SHOWING
OCEAN CURRENTS

John Bartholomew & Co., Edin^r.
]

CHAPTER VIII

THE LITTORAL PLANTS AND THE CURRENTS OF THE PACIFIC

ACTIVE as the currents are in dispersing seeds and fruits over the
Pacific, it should be remembered that those plants that owe their
distribution to this agency are only shore-plants, and not, indeed, all
the shore-plants, but only those with buoyant seeds or fruits. Even the
coral atoll owes a great deal to the agency of the fruit-pigeon and of
other birds; for instance, their species of Ficus, Eugenia, and Pisonia.
In order, therefore, not to form an exaggerated notion of the efficacy
of the currents, it will be necessary to obtain some numerical idea of
what they have really accomplished in transporting seeds and seedvessels
over the oceans in a state fit for successful germination on the shores
upon which they are stranded. It is requisite to make this proviso,
because in some cases the currents work to no purpose. Thus, the empty
nuts of Aleurites moluccana are carried far and wide over the Indian and
Pacific Oceans, and are stranded on the beaches of the various islands,
as I have found myself in the cases of Keeling Atoll, Java, and Fiji.
The Coco-de-Mer, or the Double Coco-nut Palm, is another apt instance.
Though its fruits have been carried far and wide over the Indian Ocean,
the species is restricted to the Seychelles. So also the acorns of
various species of Quercus are widely but ineffectually distributed by
the currents both in temperate and tropical regions. (This subject of
useless dispersal is dealt with in Chapter XIII.)

It is essential to bear in mind at the outset that for their inland
plants the Pacific islands can draw on the floras of a relatively large
portion of the globe. Such plants, having as a rule fruits or seeds that
sink in sea-water, or are incapable of floating for long periods, could
only have arrived at these islands, where man’s interference is
excluded, through the agencies of winds and birds, assisted by other
lesser agencies, as those of bats, insects, &c. On the other hand, for
their littoral plants, which are for the most part dispersed by the
currents, the source of supply is very restricted. The shore-plants with
buoyant seeds or fruits of the islands of the tropical Pacific, that are
here dealt with, number only about seventy, and it is not likely that
this number will be greatly increased, since, whatever may be the
deficiencies in our acquaintance with the inland floras of these
islands, we have a fairly complete knowledge of the strictly littoral
plants.

I do not suppose, indeed, that the number of such plants with seeds or
fruits capable of being transported unharmed over wide tracts of sea
would much exceed 100 for the whole Indo-Pacific region from India to
Tahiti. Professor Schimper gives a list containing 117 tropical plants
distributed far and wide over the shores of this region, and made up of
species dispersed by currents, birds, and man. Taking a liberal
estimate, not over two-thirds of the plants mentioned in this list are
dispersed by currents. Then, again, if the flora of a coral atoll, like
that of Diego Garcia or of the Keeling Islands, is taken as affording an
index of the work of the currents, the number of plants dispersed by the
currents would appear to be indeed restricted, since in either case
their indigenous flowering plants, including those of both the buoyant
and non-buoyant groups, do not exceed fifty.

About twenty years ago, Mr. Hemsley, who, in his work on the botany of
the _Challenger_ Expedition, prepared the way for the investigation of
this subject, made a list of not less than 120 plants, almost all
tropical, that are “certainly or probably dispersed” by the currents
(Introd. _Chall._ Bot., p. 42). This is admittedly only a preliminary
list, and as the result of recent investigations some plants have to be
omitted and others to be added; but I doubt whether, numerically, it is
far below the mark. The relative efficacy of the currents seems to have
been first systematically discussed by De Candolle in his _Géographie
Botanique_, which was published in 1855. Data were then very scanty, and
out of a list of nearly 100 inter-tropical species (Old World plants
found in the New World and New World plants found in the Old World) he
designates nine only as exclusively dispersed by the currents. Even this
list, in one respect, needs correction (see Note 33); but it is of
interest to note that this eminent botanist from the first never looked
upon the agency of the currents as a very important factor in
plant-dispersal; and, finding in the specially directed and carefully
performed experiments of Thuret confirmation of his views, he reiterated
his opinion in a note to that author’s paper in 1873 (cited in Chapter
III.).

However, De Candolle was quite right in minimising the effect of
currents on the distribution of plants. His extensive survey of the
plant-world from the standpoint of dispersal gave him that sense of
proportion in assigning values to dispersing agents which enabled him to
feel his way almost intuitively, even where exact data were often
lacking. It is, however, a little disappointing to find such a slight
treatment of the subject in Kerner’s great work on the Natural History
of Plants, though one can scarcely controvert his opinion that the
dispersion of plants, as a whole, is not appreciably affected by this
process. Numerically speaking, this is in the main correct; yet it is
here that the genius of Schimper led him to recognise and to mark out a
line of investigation, fruitful in important results, in connection with
the weighty question of “Adaptation.” If the author of this work has
been able to add a little to our acquaintance with this subject, he owes
much to the inspiration he received from Schimper’s memoir on the
Indo-Malayan Strand-Flora.

Still, it must be admitted that the effectual operations of the currents
as plant-dispersers are limited to the shore-plants with buoyant seeds
or fruits. If we were to include in our list the shore-plants of
temperate regions that possess seeds or fruits capable of floating in
sea-water for long periods, and of afterwards germinating, the total for
the whole world would not, I imagine, reach 200. We cannot here concern
ourselves with those purely river-side plants that contribute their
buoyant seeds and seed-vessels to river-drift, since there is no
evidence indicating that river-side plants are effectively dispersed by
the currents unless they also frequent the estuary and the coast-swamp;
and in that case they come under the head of littoral plants. The total
for the whole British flora would probably not far exceed a dozen, and
nearly all of them are very widely dispersed.

The working value of the currents as plant-dispersers in the Pacific can
be rudely estimated by the number of littoral plants with buoyant seeds
or fruits that occur in the various groups. Most of these plants hail
from the Indo-Malayan region. Speaking generally of the extension
eastward of the Indo-Malayan strand-plants over the Pacific, Prof.
Schimper (page 195) remarks that they become fewer and fewer in number
as they extend farther from their original home, their number shrinking
to a very few in the most remote groups of the Marquesas and the
Hawaiian Islands. This is well illustrated in the following numerical
results that I have prepared. Of the whole number, some seventy in all,
of the littoral plants of the tropical Pacific with buoyant seeds or
fruits, Fiji possesses about sixty-five, Tahiti about forty, and Hawaii
only about sixteen. As shown, however, in Chapter VII., some of the
Hawaiian littoral trees that are useful to the aborigines were probably
introduced by them. The number actually introduced through the currents
into Hawaii in all likelihood therefore does not exceed ten. There is a
method in this diminution in numbers, as the plants migrate eastward and
northward over the Pacific, which has been described in detail in the
preceding chapter. The efficacy of the currents as plant-dispersers in
the tropical Pacific therefore diminishes as we proceed eastward.

In the South Pacific the littoral plants preserve their Old World origin
as far as the Polynesian archipelagoes extend eastward across to
Pitcairn, Elizabeth, and Ducie Islands, where we find in one or other of
them such characteristic Indo-Malayan beach trees as Barringtonia
speciosa, Cerbera Odollam, Guettarda speciosa, Hernandia peltata, and
Tournefortia argentea (see Note 34). In the more distant Easter Island
there is a suspicion, for the first time, of immigration from South
America in the presence of Sophora tetraptera. In the islands relatively
close to the American continent, as in Juan Fernandez and in the
Galapagos group, the Indo-Malayan strand-plants are no longer
represented.

We come now to consider the relation between the distribution of the
shore-plants and the currents. It is quite legitimate to discuss the
currents of the Pacific from the botanist’s point of view, that is to
say, from the standpoint of the distribution of littoral plants with
buoyant seeds or fruits. For ages the buoyant seeds and fruits of the
strand-plants of the tropical Pacific have been drifting over that
ocean, and we have the results now before us in the dispersal of the
species to which they belong. There is no necessity to endeavour to make
the distribution of such littoral plants square with the arrangement of
the currents as shown in a chart. The usual result of such a comparison
has been to lead the investigator, whether an anthropologist, a
zoologist, or a botanist, to find his facts at variance with the course
of the prevailing currents. Man, animals, and plants have entered the
Pacific from the west, whilst the most available currents are from the
east; and one may be perhaps permitted the solecism that the Pacific
islands have apparently been stocked with their shore-plants, with their
aborigines, and with much of their fauna by currents running in the
wrong direction. These Pacific islands could only have had a direct
communication with the Old World, from which they have mainly derived
their shore-plants, by the currents; but since both the aborigines and
the plants have forced their way across the ocean to the Tahitian region
in the teeth of the regular currents, indicated as such in the chart, we
are compelled to assume that they have availed themselves either of the
Equatorial Counter-Current or of the occasional easterly drift currents
that mark the prevalence of westerly winds during the short season of
the year when the easterly trade-winds do not prevail.

The Equatorial Counter-Current hypothesis would involve a preliminary
crossing of the whole breadth of the Pacific Ocean, that is to say, a
voyage of some 8,000 miles, before the drifting seed doubled back to the
Polynesian Islands. The other view is a much more probable one, as is
sufficiently indicated by the following extract from the “Admiralty
Sailing Directions for the Pacific Islands” (II., p. 25, 1900).... “In
the western part of the Pacific these trades ... are frequently
interrupted by winds which blow from west or north-west, especially
during the months of January, February, and March, when the north-west
monsoon of the Indian Ocean extends out in the Pacific as far as the
Samoa Islands.” In various works on this region one may find reference
to canoes blown off the shore during this season and carried some
hundreds of miles to the eastward. A ship can then sometimes sail with a
fair wind from the southern end of the Solomon Group to the Fijis; and
as we learn from Mariner, the crocodile may be at such times carried
away from the Solomon Islands and stranded in Fiji. Mr. Hedley, in his
exceedingly interesting paper on a zoogeographic scheme for the
mid-Pacific (_Proc. Linn. Soc. N.S.W._, 1899), gives many details of
this nature; but there is no space to deal further with the matter here.

After all, the botanist must take his cue from the drifting seed and the
distribution of the plant. He finds the seed floating in the open sea as
well as stranded on the beach. He then discovers the plant growing on
the beaches, and by experiment he tests the floating capacity of the
fruit or seed. Finally he ascertains the home of the plant. He does this
for all the littoral plants with buoyant seeds or fruits, and he forms
his own conclusions of the efficacy of the currents independently of the
current-chart, remembering that he has in Time an important factor that
the geographer does not possess in dealing with the currents. The effect
of time has often been to obscure the differential results of the
operations of the currents in the case of those species that, like
Barringtonia speciosa, are almost universally distributed in the islands
of the Pacific. It is obvious that such plants cannot aid us much in the
matter of ascertaining the track followed by the drifting seed in
entering this ocean. But if we find a littoral plant with buoyant seed
or fruit that has only partially performed the traverse we shall possess
in the interrupted operation an important piece of evidence.

Several years ago, in my paper on Polynesian plant-names, read before
the Victoria Institute, I developed this argument when endeavouring to
find in the floating seed a clue to the route pursued by the Polynesians
in entering the Pacific. Since that time my acquaintance with these
islands and their plants has been considerably extended; but no
important modification of the principal argument is now needed. It was
then pointed out that in Nipa fruticans, the swamp-palm of the Malayan
Islands and of tropical south-eastern Asia, we have a plant well fitted
for the purpose and one well known to be dispersed by the currents over
small tracts of ocean. The Nipa Palm has attempted to enter Polynesia
from the Malayan region by two routes, namely, by Melanesia and by
Micronesia. Along the first route it has in the course of ages reached
the Solomon Islands, where I found it in 1884. Along the second route it
has extended its range to Ualan at the eastern end of the Caroline
Group, where it was observed by Kittlitz many years ago, as indicated in
the narrative of his voyage (_Reise nach russische America, nach
Mikronesien_, etc., 1858, ii. 35), and in Dr. Seemann’s English edition
of the same author’s _Vierundzwanzig Vegetationsansichten ... des
stillen Oceans_.

The question now arises as to which of these two routes was taken by the
drifting seed. In my paper I adopted the view that the shore plants
reached Fiji and Samoa by Micronesia, that is to say, by the Caroline,
Marshall, and Gilbert Groups. This is the route which, as mentioned by
Mr. Hedley in the paper above quoted, Mr. Woodford prefers for some of
the Lepidoptera; and it is the one that is favoured by Mr. Wiglesworth
for the birds, since in his memoir entitled _Aves Polynesiæ_ he remarks
that certain indications tend to show that the Pelew Islands have served
as a sort of bridge for the spread of species from Indo-Austro-Malaya
right across the Pacific. Though I still think that the beach trees,
most of which would find a home on the numerous coral atolls of the
Marshall, Gilbert, and Ellice Groups, often followed that track, yet I
am now inclined to consider that the mangroves and their associates,
plants which find their most suitable home in the estuaries of large
elevated islands, like those of the Solomon Group, in all probability
reached Fiji in the mass by the Melanesian route.

[Illustration:

TRADE ROUTES

OF THE

PACIFIC OCEAN

(_On Mercator's projection_)

John Bartholomew & Co., Edin^r.
]

Although the Old World has supplied to the Pacific islands most of their
littoral plants that are dispersed by the currents, that is to say, the
plants with buoyant seeds or seed vessels, yet there is an appreciable
American element, and it is with the plants occurring in the New World
that we are now concerned. The total number of the littoral plants of
these islands that possess buoyant seeds or fruits is, according to the
lists given under Note 35, about seventy. Of these about forty-five are
exclusively Old World species, sixteen occur in both the Old and New
Worlds, three are exclusively American, and six are Polynesian.

The question we have now to ask ourselves is whether the shore plants
common to both the Old World and America have their homes in America, or
whether they have been derived from the other hemisphere. With one or
two exceptions, as in the cases of the Australian genera Dodonæa,
Scævola, and Cassytha, which, as shown in a later page in this chapter,
present no great difficulty, there does not seem to be any serious
objection, as far as the numerical distribution of the species is
concerned, in regarding America as a possible home of the genus. It is
not often we shall come upon such a striking instance of the principle
that where the species are most numerous there is the home of the genus,
as in the instance of Cocos. The Coco-nut palm has been carried around
the world through the agencies of man and the currents, whilst the home
of the genus is in America.

Now assuming that in having to choose between the Old World and the New
World as the home of most of the genera in the list we selected the
latter, we have to ask ourselves in what degree this would be consistent
with the place America holds with regard to the distribution of tropical
shore-plants dispersed by the currents and with reference to the
arrangement of the currents. If we except the African continent, there
is no part of the world that bears such a definite relation to the
currents as America, and with an ordinary chart of these regions their
arrangement is to be understood at a glance. Yet strange to say, as far
as the distribution of tropical littoral plants is concerned, America
holds a position that the present system of the currents on its coasts
will not altogether explain. Within the lifetime of the species of
mangroves and other plants of the coast swamps that are found on both
the Pacific and Atlantic coasts of tropical America the two continents
of this name have been united by the emergence of the Isthmus of Panama.

Few things are more significant in plant-distribution than the
arrangement of the tropical littoral plants with buoyant seeds or
fruits, a subject that is discussed with some detail by Professor
Schimper in his work on the Indo-Malayan strand-flora (page 190). These
plants group themselves into four sections:—

(_a_) Those of the Pacific and Atlantic coasts of tropical America
(including the West Indies) and of the West Coast of Africa. They
include mostly plants of the mangrove-swamps and their vicinity, such as
Anona paludosa, Avicennia tomentosa, A. nitida, Conocarpus erectus,
Laguncularia racemosa, Rhizophora mangle, etc.

(_b_) Those of the Old World excluding the African West Coast and
extending from the East Coast of Africa eastward to the Pacific islands.
This is much the largest group and comprises many of the plants named in
the list given in Note 35 under Old World species. One may cite as
examples of plants ranging almost all over this area, Barringtonia
speciosa, B. racemosa, Bruguiera gymnorhiza (in its most comprehensive
sense), Carapa moluccensis, Derris uliginosa, Guettarda speciosa,
Hernandia peltata, Heritiera littoralis, Pemphis acidula, Rhizophora
mucronata, etc. Plants of the mangrove-swamp and of the beach are,
therefore, here included.

(_c_) Those occurring all around the tropics and including many of the
plants mentioned under Note 35 as Pacific island shore-plants found also
in America. Most of them belong to the Leguminosæ, and there may here be
mentioned Canavalia obtusifolia, Cæsalpinia Bonducella, Entada scandens,
Gyrocarpus jacquini, Ipomœa pes capræ, Sophora tomentosa, and Vigna
lutea.

(_d_) Those confined to a portion of the two great regions, such as Nipa
fruticans in the Old World, and the Manchineel (Hippomane mancinella) to
tropical America.

It is to be noted that the ubiquitous species do not include any of the
mangroves. Each of the two regions has its own species, none being
common to both the American and Asiatic regions, although, as is shown
in Chapter XXX., the American species of Rhizophora is now seemingly
breaking its bounds and intruding into the Pacific islands. On the other
hand, some of the mangrove genera, Avicennia, Carapa, and Rhizophora,
are found all round the globe, whilst others are restricted to one or
other of the two regions, Bruguiera, Lumnitzera, and Sonneratia, for
instance, to the Old World region, and Laguncularia to the American and
West African region.

For convenience we may designate the two great regions of tropical
strand-plants, with buoyant seeds or fruits, the American and the
Asiatic regions, remembering that the first includes both coasts of
America as well as the African West Coast, whilst the second extends
from the East Coast of Africa to Polynesia. Excluding the ubiquitous
species, these two regions are well distinguished from each other. If we
look at the chart of the currents we perceive the reason of the American
region including the West African Coast, and we see why none of the
indigenous plants of this region occur on the African East Coast. So
also with the Asiatic region, a glance at the chart will show that all
the portions of its area are in connection with each other directly or
indirectly through the currents, and that only time is required for the
transport of buoyant seeds over most of the region.

Hitherto I have mainly followed Professor Schimper in this matter; but
since my visit to Ecuador and the Panama Isthmus some further
considerations have presented themselves to me. If the reader will look
again at the map of the currents, he will observe that there is little
reason for supposing that the Asiatic region can lend its littoral
plants to the American region. On the other hand there are greater
facilities, as far as currents are concerned, for America supplying the
Asiatic region, namely by means of the great equatorial currents that
course westward across the Pacific to the tropics of the Old World.

It would therefore seem that the American region can receive nothing by
the currents from the Asiatic region. If accordingly it gives but gets
nothing back, we are compelled to assign an origin in the American
region to all littoral plants dispersed by the currents that are found
in the tropics around the globe. This is what we have already regarded
on other grounds as possible for nearly all the littoral plants of the
tropical Pacific with buoyant seeds or seedvessels that are found in
America. These plants are practically the same as those distributed
around the tropical zone which are enumerated in the list given under
Note 35, _b_. With their home in America, by crossing the Pacific they
would ultimately arrive at the East African coast, where their course
westward would terminate; whilst commencing their journey from the east
side of the American continent they would reach the West African coast;
and their distribution around the tropics of the world would be
explained. There follow from these considerations the corollaries that a
tropical strand-plant dispersed by the currents which has its birthplace
in Asia could never reach the American region, and that American
strand-plants are for the most part native-born, excepting those, if
there are any, that hail originally from the African West Coast.

It is necessary in passing to explain the similarity of shore plants on
the Pacific and Atlantic coasts of Tropical America. For the mangroves
and their accompanying plants inter-communication between the two coasts
is now impossible; and a communication between the two oceans must be
postulated within the lives of the existing species. For the plants like
Entada scandens and Ipomœa pes capræ, which occur inland as well as at
the coast, it is easy to show that in the case of the Panama Isthmus,
their seeds could be readily carried into the Atlantic and Pacific
Oceans by rivers draining the opposite slopes of the same “divide,” so
that the dispersal of the same species from a common centre into two
oceans may be seen in operation in our own day. My observations on this
subject are given in Chapter XXXII., to which the reader is referred.

I have now gone far enough to indicate the place that America holds with
regard to the distribution of tropical shore-plants dispersed by the
currents and with regard to the currents. There is every probability, as
I venture to think I have shown, that the Pacific islands have derived
most of their ubiquitous shore-plants with buoyant seeds or fruits from
America. But one of the results of our discussion of America in this
double aspect was that excepting in the case of the African West Coast
it gives but does not receive plants from the Old World. We apply this
test, with perhaps a little hesitation, to the shore-plants of the
Pacific islands that are dispersed by the currents; and we find, as will
be seen below, that it is responded to in a remarkable manner.

It has been observed in the previous chapter that scarcely any of the
large-fruited beach-plants of the South Pacific islands, that could only
have been dispersed by the currents, have reached Hawaii. We do not find
amongst the truly indigenous coast flora of this group any of the
following trees: Barringtonia speciosa, Calophyllum Inophyllum, Cerbera
Odollam, Guettarda speciosa, Hernandia peltata, Ochrosia parviflora,
Pongamia glabra, Terminalia Katappa, Terminalia littoralis, &c. It was
also noted that the currents had not only failed to establish these
plants in Hawaii, but that they had also failed to establish them in
America, the suggestion being that the Hawaiian Islands had been, in
part at least, stocked by the currents from America. That the
Indo-Malayan strand-plants in their extension eastward over the Pacific
should have failed to reach America, is a result we might have expected
from the arrangement of the currents. Yet mingled with them we have
plants like Ipomœa pes capræ, Canavalia obtusifolia, and Sophora
tomentosa, that also occur in America. Since, however, their seeds are
not better adapted for accomplishing the passage across the Pacific from
the Old World to America than the equally buoyant fruits of the
above-named littoral trees that have failed, the presumption arises that
their home is in America, and that they have performed the easier
passage across the Pacific westward from America to the Old World.

The exclusion of so many characteristic shore-trees from America that
range often over the whole tropical region from the African East Coast
to the islands of the Central Pacific, is not a matter of seed or
fruit-buoyancy, but a matter concerned with the home of the species, and
with the arrangement of the currents. Those shore-plants of this region
that occur also in America have their home in that continent, and have
subsequently been carried across the Pacific by the currents westward to
the Asiatic shores.

The only exceptions, that I can recall, to the rule that America does
not receive shore-plants dispersed by the currents from the Old World,
are presented by the three Australian genera, Dodonæa, Scævola, and
Cassytha, of which widely spread littoral species occur in America,
namely, Scævola Lobelia, Dodonæa viscosa, and Cassytha filiformis. They
offer, however, but little difficulty, since, as pointed out in other
parts of this work, Dodonæa viscosa has probably been in part dispersed
by man, whilst the other two species are as well fitted for dispersal by
birds as by currents. The occurrence therefore of these species in
America does not necessarily raise the question of the currents.

The same exclusive principle is illustrated in the scanty littoral flora
of Hawaii. Deprived, like America, of the characteristic large-fruited
beach-trees of the South Pacific, species that could only have reached
it through the agency of the currents, it is scarcely to be expected
that it would have received its few littoral plants with buoyant seeds
from the source which has failed it in the cases of the numerous
absentees. It is to America therefore that we look for the source of its
littoral plants as far as the agency of the currents is concerned.

The Hawaiian Islands contain about twelve plants, named in the list
given in Note 36, that possess seeds or fruits known to be dispersed by
the currents, and capable, as experiments indicate, of floating in
sea-water for prolonged periods. Not all of them are at present littoral
in their station in this group; but their claim to be considered such in
other regions is established in the Note above mentioned. Of these
plants, seven at least are found in America, five in the Old World also,
and two exclusively in America. This proportion of American plants is
far greater than that characterising the whole littoral flora of the
Pacific islands dispersed by currents, where out of some seventy species
only nineteen are found in America (see Note 35). As far as the
distribution of the plants is concerned, it is therefore quite possible
that Hawaii has received most of its plants that are dispersed by the
currents from tropical America.

We will now consider how such a possibility is in accordance with the
arrangement of the currents in the North Pacific. If we look at the
Quarterly Current Charts for this ocean published by the British
Admiralty we notice that all through the year the Hawaiian Group lies
more or less within the area of currents flowing from the West Coast of
America, the Northern Equatorial Currents as they are collectively
named. Except in the winter months these currents come from the N.E. and
E.N.E., and bring drift from the coasts of British Columbia, Oregon, and
Northern California. It is then that they pile up huge pine logs on the
shores of the Hawaiian Islands, as I have described in Chapter VII. and
in Note 30; and, according to Dr. Hillebrand, they transport this drift
timber much farther south to the shores of the Marshall and Caroline
Groups. One might cite other facts illustrative of the working of these
currents, such as one finds in the pages of Fornander and other authors;
but this would scarcely come within the province of this work. I may
here remark that when in Honolulu I was informed that a bell-buoy which
had got adrift on the Californian coast was subsequently washed up on
the coasts of Kauai. It is stated in Findlay’s “North Pacific Directory”
(1886, p. 1068), that a junk carrying nine hands that had been blown off
the south coast of Japan in a typhoon, anchored, after ten or eleven
months at sea, in December, 1832, near Waialea in Oahu, the view taken
of its course being that after drifting along in the Japan Current it
came within the range of the south-west current that carries pine timber
to Hawaii from the West Coast of America.

The portion of the Northern Equatorial Current that strikes the Hawaiian
Group during the greater part of the year is no doubt a south-westerly
deflection of the Japan Current from the American West Coast; and it
would be impossible to find any tropical drift mingled with the pine
logs stranded on the islands during that period. However, in the winter
months, centering in January, the Japan Current flows down the West
Coast of America to about the latitude of Cape Corrientes on the coast
of Mexico, before being deflected westward. Here it meets with a portion
of the Peruvian Current, and both flow westward, the united stream
striking probably only the southernmost islands of the Hawaiian Group.
It is at this season alone that there would be any likelihood of drift
from tropical America being stranded on the Hawaiian beaches, and it is
quite possible that at such a time the Northern Equatorial Current may
carry intermingled in its stream pine logs from Oregon and seed-drift
from Panama.

I am not inclined to attach any value except in the Western Pacific to
the agency of the Equatorial Counter-Current in transporting seeds and
fruits over the Pacific. It presents seemingly the only opportunity of
the transportal of the seeds and fruits of Asiatic littoral plants to
America; but if at all effective in this way, it would have endowed the
littoral flora of the western shores of tropical America with many of
the trees so characteristic of the coral islands of the Pacific. In this
sense, it has failed completely as an effective agency in
plant-dispersal; and judging by results we may, I think, dismiss it from
our consideration. However, Dr. Hillebrand (p. xv.) assumes that during
the prevalence of south-westerly gales in winter in the Hawaiian
Islands, the Equatorial Counter-Current would be pushed northward so as
to mingle to the east of the group with the North Equatorial Current. In
this manner it is supposed that seed-drift brought direct from the
Asiatic side of the Pacific would be stranded on these islands. This
appears to me to be most improbable, since some ten or twelve degrees of
latitude usually intervene between the Hawaiian Group and the Equatorial
Counter-Current (_see_ Admiralty Sailing Directions, Pacific Islands,
1900, II., 31, and the Quarterly Current Charts; also Encyclopædia
Britannica, vol. 18, p. 118).

The most serious objection from the botanist’s standpoint against such a
view as that of Dr. Hillebrand is the absence from Hawaii of most of the
shore-plants that we should expect the currents to have brought from the
Old World. It is also evident that as far as the currents are concerned
the Hawaiian Islands are far more likely to receive littoral plants from
America than from the Old World. Though no tropical drift has yet been
found stranded on the coasts of these islands, yet it is not unlikely
that future investigators may find some seed-drift from Central America
on the most southerly coasts of the group, as on the south-east shores
of the large island of Hawaii. It would only be stranded in the winter
months and then probably in small quantities.

_Summary of the Chapter._

(_a_) Since the effective operations of the currents are limited to the
shore-plants with buoyant seeds or fruits, such plants forming but a
small proportion of any flora, it must be acknowledged that, numerically
speaking, the results of the dispersing-agency of the currents on
plant-distribution in general are but slight.

(_b_) Yet the importance of the subject is by no means to be measured by
a numerical scale of results, a line of inquiry being here opened up
leading to fields of investigation full of promise for the student of
plant-distribution.

(_c_) Whilst dealing with the relation between the distribution of
shore-plants and the arrangement of the currents, it is quite legitimate
to discuss the currents of the Pacific from the point of view of the
botanist, who, after all, must take his cue from the drifting seed and
the resulting distribution of the plant.

(_d_) The shore-plants of the Pacific islands that are dispersed by the
currents being mainly Indo-Malayan in origin, it follows that they have
extended eastward over the Pacific to the Tahitian islands against the
stream of the South Equatorial Current and against the trade-wind. It
is, however, shown that they could have availed themselves of the
interval between January and March when the North-west Monsoon reaches
the Pacific.

(_e_) It is claimed that whilst the mangroves and their associated
plants have for the most part entered the Pacific by the Melanesian
route through the Solomon Islands, the beach-plants have also followed
the route through Micronesia by the Caroline, Marshall, and Ellice
Groups.

(_f_) A small number of the strand-plants of the Pacific islands that
are dispersed by currents occur in America as well as in the Old World;
and questions of prime importance arise when we have to decide whether
their home is in the Old World or in the New World.

(_g_) Good reasons are given for regarding them as chiefly of American
origin; and it is shown that America with regard to the arrangement of
the currents stands in the singular relation of being a disperser but
not a recipient of shore-plants.

(_h_) It is pointed out that the tropical shore-plants that are
distributed by currents belong to two great regions which are the effect
of the present arrangement of the currents, viz., the American including
the West Coast of Africa, and the Asiatic comprising the remainder of
the tropical zone. Each region has its own plants, and those that occur
in both, being in fact distributed all round the tropics, are regarded,
according to the principle above stated, as having their home in the
American region.

(_i_) The occurrence of the same strand species on the Pacific and
Atlantic coasts of tropical America is regarded as indicating that the
arrangement of the existing species of its shore-plants, more
particularly of the mangroves, antedates the emergence of the Panama
Isthmus. This hypothesis is not needed for the coast plants like Entada
scandens that occur inland, since we can now observe their seeds being
carried down into the Atlantic and Pacific Oceans by rivers draining the
opposite slopes of the same “divide” in the Panama Isthmus.

(_j_) It is shown that the currents of the Pacific have failed to
establish the numerous beach-trees (possessing buoyant fruits) of the
Pacific islands, not only in the Hawaiian Group, but also on the coast
of America; and it is therefore argued that we should expect the
Hawaiian Group to have received through the currents its shore-plants
with buoyant seeds or fruits from the tropical west coasts of America.

(_k_) In support of this contention it is pointed out that most of the
Hawaiian strand-plants that are dispersed by the currents are found in
America, and some indeed in America to the exclusion of the Old World.

(_l_) The arrangement of the currents in the North Pacific also favours
the view that the Hawaiian Islands are more likely to receive plants by
the agency of the currents from America than from the Asiatic side of
the Pacific.

CHAPTER IX

THE GERMINATION OF FLOATING SEEDS

THE tendency of the floating seed or fruit to germinate in the estuaries
of tropical rivers is especially characteristic of the plants of the
mangrove-swamps and of their borders. In the Fijian rivers, and
particularly in the estuary of the Rewa, where the river-water is
usually mixed with that of the sea, there are frequently to be found in
a state of germination floating fruits of Barringtonia racemosa, Carapa
obovata, Clerodendron inerme, Derris uliginosa, Smythea pacifica, &c.;
whilst the floating fruits of more characteristic beach-trees like
Barringtonia speciosa and Cerbera Odollam, that grow also on the sides
of the estuaries, were never noticed in this condition. That this
tendency should be restricted to the plants of the mangrove-formation
and is not to be observed in the beach-trees is a singular fact. There
is, however, an intermediate group of littoral plants mostly belonging
to genera of the Leguminosæ and Convolvulaceæ, such as Mucuna and
Ipomœa, where germination of the floating seed is apt to begin but ends
abortively, and results in the sinking and death of the seed. The
subject of the germination of seeds in the floating drift of tropical
estuaries presents itself, therefore, in three aspects:—

(1) As concerning the plants of the mangrove-formation, where, excluding
the viviparous species (when germination takes place on the plant),
germination is frequent in the water:

(2) As concerning the beach-trees where it is rare or absent altogether:

(3) As concerning certain Leguminous and Convolvulaceous littoral plants
where germination is not infrequent but always abortive.

Dealing first with the plants of the mangrove-formation, it may be
remarked that the same tendency of the floating fruits or seeds to
germinate, which is above noticed in the case of the estuaries of Fiji,
came under my observation in the floating drift of the estuary of the
Guayaquil River in Ecuador, the germinating fruits and seeds being
carried far out to sea. The seeds of Anona paludosa, which float in
quantities in the river-drift, were often found germinating; and the
same may be said of the fruits of Laguncularia racemosa and of the
“joints” of Salicornia peruviana which abound in the creeks of the
mangrove-delta and are carried out to sea in the germinating condition.

It might be expected that this readiness to germinate in the brackish
water of estuaries would prove to be a formidable obstacle to the
dispersal of these plants over wide tracts of ocean. The exposed
portions of the seedling might be deemed ill-suited to withstand,
without injury, the “wear-and-tear” of transport by currents over long
distances, even when not affected by the sea-water; and it might be
thought that they would be often nibbled off by fish or destroyed by
other aquatic animals. Only the specially organised seedlings produced
by a viviparous process on the tree, such as those of Rhizophora and
Bruguiera, might be regarded as able to survive the effects of prolonged
immersion in the oceanic currents.

Observation, indeed, shows that such seedlings are exposed to and suffer
from these perils; yet it is evident from the distribution of the
species that, whether in the germinating condition or not, the seeds and
fruits of Anona paludosa and Laguncularia racemosa have been carried by
the currents from America to the West Coast of Africa. The seedlings of
Avicennia and of Rhizophora mangle have also performed the same
trans-Atlantic voyage. Those of both these mangroves are to be observed
floating off the coasts and in the estuaries of both coasts of America.
The seedlings of Avicennia are particularly abundant in the
mangrove-creeks of the delta of the Guayaquil River; and I observed them
in a healthy condition, ten to twenty miles out at sea, floating
together with those of the Rhizophora. Since, as in the case of
Rhizophora, germination occurs normally on the plant, Avicennia can only
be dispersed by its floating seedlings. Yet it is noteworthy that
although Avicennia seedlings appear, to a marked degree, less fitted for
ocean transport than those of Rhizophora and Bruguiera, the species have
a much wider distribution. Avicennia officinalis has a cosmopolitan
distribution in the tropics and beyond, occurring as it does on the
Atlantic and Pacific coasts of America, on both coasts of Africa, over
Asia and Australia, as well as in New Caledonia and New Zealand, but not
in Polynesia (_Bot. Chall. Exped._, III., 178).... I have now gone far
enough to show that the tendency displayed by the seeds and fruits of
several of the plants of the mangrove-formation to germinate either on
the tree or in the floating drift of estuaries has not affected the
general distribution of the species in its main outlines. Few fruits are
found more often in a germinating condition in the floating drift of the
Rewa River in Fiji than those of Barringtonia racemosa, yet the species
ranges from the African East Coast eastward to Polynesia. Seedlings as
well as seeds or fruits, whether or not in a germinating condition, are,
therefore, able in such cases to disperse the species.

This readiness of the floating fruits of plants of the mangrove
formation (excluding the viviparous species) to germinate in the
estuaries is, I am inclined to think, due in the main to the strain of
vivipary that runs through nearly all the plants of the mangrove-swamp
and of its borders. It would, indeed, appear that the viviparous habit
(the capacity of germinating on the plant) which finds its extreme
development in Rhizophora and Bruguiera of the Fijian swamps is
represented in its earliest stage in the readiness of the floating
fruits of Barringtonia racemosa, Carapa obovata, &c., to germinate in
the Fijian estuaries, and as remarked in Note 37 there is a suspicion of
vivipary in the instances of both the species just named. Intermediate
cases, as that of Laguncularia in the Ecuador swamps, occur in other
regions with species where germination only takes place at times on the
plant. This subject is, however, generally discussed in Chapter XXX. and
need not be further dealt with here.

A predisposing cause of the germination of floating seeds and fruits in
tropical estuaries would seem to be afforded by the super-heating of the
water of the estuary. This came under my notice both in the Rewa River
in Fiji and in the Guayaquil River in Ecuador, where the water of the
estuary is often noticed to be some degrees warmer than that of the sea
outside, and of the water from the river above the estuary. (See Note
38.)

We come now to the subject of the non-germination in tropical estuaries
of the floating fruits of the beach-trees, such as Barringtonia speciosa
and Cerbera Odollam, that in the Pacific islands may contribute to
river-drift. Such trees may grow on the banks of the estuary, and their
fruits would thus readily fall into the water; but in the Rewa estuary
in Fiji it was evident that the fruits and seeds of beach-plants, such
as Scævola Koenigii, are also brought in by the tide. The seeds of
Morinda citrifolia were often noticed in the Rewa drift together with
the fruits of Heritiera littoralis, which is both a beach and a swamp
plant, but never in a germinating condition. The same remark applies
also to the fruits of beach trees found afloat in the sea between the
islands, such as Cordia subcordata, Guettarda speciosa, and Terminalia.
It is possible that a few of these plants, as in the case of
Barringtonia speciosa, display traces in the structure of their fruits
of a lost viviparous habit. (See Note 50.) It is pointed out in
discussing Guettarda that germination is much more easily induced than
one would expect in the case of fruits with such a hard ligneous
putamen.

An interesting subject is presented in the abortive germination of the
floating seeds of many plants of the Leguminosæ and Convolvulaceæ both
at sea and in a tropical estuary. My conclusions on this matter are
based partly on observations made in Fiji, but mainly on the results of
numbers of experiments, this being unavoidable, since the abortive
germination causes the sinking of the seed. The principal determining
cause of the germination in water of one of these floating seeds is
evidently to be sought in the temperature of the water, it being
immaterial for the earliest stage of germination, as many of my
experiments indicate, whether the seed or fruit is afloat in the sea or
in the river. In these flotation experiments, when conducted under warm
conditions with sea-water, the earliest signs of germination were
frequently displayed in the softening, swelling, and sinking of the
seed. If the swelling seed is taken out in time and planted after a
preliminary soaking in fresh water, the germinating process is at once
resumed and is often successfully and rapidly completed; but if the seed
is allowed to remain in the vessel after it has absorbed sea-water the
vitality of the embryo is destroyed and the seed decays.

That many seeds would fail from this cause to cross an ocean my
experiments repeatedly demonstrated. Nor does the appearance of a seed
afford any indication of its probable failure to cross an ocean. Some
seeds of Mucuna, as far as their hard coverings could guide one, would
seem to be quite secure from such a risk. The stony seeds, for instance,
of M. urens D.C. look as if they might safely be transported by the
currents round and round the globe; and De Candolle very rightly placed
this species in his scanty list of plants dispersed by currents. Yet few
seeds are more treacherous when their buoyancy in sea-water is tested in
a warm place, as in a hot-house. They may take up water, swell, and sink
in a week, or they may float unharmed for a year.

The seeds most exposed to this risk are those of the Leguminous giant
climbers, the lianes of the coast and inland forests of the islands of
the tropical Pacific. They belong to the genera Mucuna, Strongylodon,
&c.; and thus several of the plants that constitute for the student of
plant-dispersal the enigmas of the Pacific are here included. The seeds
of Mucuna are especially liable when afloat in sea-water under warm
conditions to display the early signs of germination, swelling up and
sinking to the bottom of the vessel, a process, however, soon arrested
and followed by the death of the embryo unless the seed is removed in
time. Yet the seeds of this genus are notably long “floaters.” Those of
an American species, variously designated as Mucuna pruriens D.C. and M.
urens D.C., have long been known to be washed ashore together with the
seeds of Entada scandens on the western shores of Europe, and
particularly on the Scandinavian coast, where they form regular
constituents of what the Scandinavian botanists correctly term the
Gulf-stream Drift.

Mucuna urens D.C. occurs with other American shore-plants that are
dispersed by the currents on the African West Coast; and there is no
reason to doubt that its seeds perform the trans-Atlantic voyage. It is
found in Polynesia, in Hawaii, in the Marquesas, and according to
Reinecke also in Samoa; and probably it occurs in other groups. The
specific determinations of the genus, however, need thorough
overhauling, so that it is not possible to deal more than in general
terms with the distribution of a species. The distribution of Mucuna
urens in the Pacific is, however, irregular, and no doubt this is to be
connected with the uncertain behaviour of its seeds when transported by
tropical currents. The seeds would, I venture to think, often sink
through abortive germination in the warm areas of equatorial seas.

When in Hawaii I kept ten of the seeds of this species (M. urens D.C.)
in sea-water for four and a half months, none of them sinking in that
period, the temperature of the water rarely reaching over 80°F., the
average daily temperature being 76-77°. However, when four years
afterwards in England I placed five of the seeds obtained at the same
time in sea-water under conditions where the water-temperature ranged
for the first few weeks between 75° and 90°, three of them began to
swell within ten days, and on removal at once germinated healthily. The
remaining two were afloat at the end of twelve months, and when planted
one of them germinated a month afterwards.

Having experimented on the seeds of about half a dozen different species
of Mucuna in sea-water, all with buoyant qualities, it is possible for
me to lay down the general rule for the buoyant seeds of the genus that
sinking is the result of an attempt at germination, which, as before
observed, proves abortive unless the seed is removed in time. It is
obvious that the gardener wishing to raise plants of this genus without
delay might profitably adopt the method of keeping them afloat in water
at a temperature of 80-90° F. until they begin to swell, which may
happen in some cases in a few days. Sea-water seems to produce the most
rapid results.

When on Keeling Atoll in the Indian Ocean I collected, amongst the
stranded seed-drift brought by the currents to those islands, the seeds
of five or six species of Mucuna, two of which were identified at Kew as
M. macrocarpa, Wall., and M. gigantea D.C. (see my paper on the
dispersal of plants at Keeling Atoll). No plant of this genus appears up
to that time to have been recorded from the Keeling Islands, so that at
all events most if not all of the seeds had been brought by the currents
from the Indian Archipelago, some 700 miles away. It may be added that
amongst the drift gathered by me on the south coast of Java the seeds of
three species of Mucuna were identified at Kew, including the two
above-named species from Keeling Atoll.

These current-borne seeds of the Keeling beaches had probably performed
an ocean journey of a thousand miles, since the route could scarcely
have been direct. Yet their behaviour when placed eighteen months after
in sea-water in a hothouse in England was most erratic. Of three seeds
of Mucuna gigantea all swelled and sank within eight days. Two seeds of
M. macrocarpa sank after floating from sixty to a hundred days; whilst
of two seeds of another species both remained afloat after a year. In a
sea-water experiment in England on five Hawaiian seeds of M. gigantea,
under the conditions referred to in the Mucuna urens experiment, one
sank within ten days, whilst three of them were afloat after twelve
months, one of them subsequently germinating. This species, it may be
remarked, is widely distributed as a coast plant over tropical Asia,
Australia, and in Polynesia. It seems to take the place in the Old World
which Mucuna urens takes in America, and it is curious that they meet in
Polynesia, being sometimes associated as in Hawaii. In the chapter on my
observations in Ecuador and in Panama it is remarked that Mucuna seeds
are frequent constituents of river, sea, and stranded drift. I,
therefore, have enjoyed the opportunity of observing the behaviour of
the seeds of this genus in a variety of localities, namely, in the
Keeling Islands, in West Java, in Fiji, Hawaii, and tropical America;
and this may be pleaded as an excuse for entering into so much detail
respecting them.

The large seeds of Strongylodon lucidum (S. ruber), a Leguminous liane
that ranks with the species of Mucuna amongst the huge climbers of the
forest of the Pacific islands, behaved in a similar way in my flotation
experiments in sea-water. Though, as shown in Note 3, these seeds can
float for a year and retain their germinating power, some of them
brought their buoyant capacity prematurely to an end by an abortive
attempt at germination. These black rounded seeds form a common object
amongst the river seed-drift stranded on some of the Fijian beaches in
the vicinity of estuaries. They are so hard and durable that they are
mounted in brooches in Honolulu. Yet these pebble-like seeds will
sometimes begin to swell in a few days in sea-water. Out of five seeds
placed in sea-water in England under warm conditions (the water
temperature for the first few weeks ranging between 75° and 90° F.), one
swelled and sank within ten days, another did so after two months,
whilst the other three were afloat after twelve months, and one of them
subsequently germinated. There is some disagreement amongst botanists as
to the limits of the specific characters of the plants of this genus
(see Note 39); but the plan seemingly most in accord with the
fundamental principles regulating plant-distribution in this region of
the Pacific is to regard the forms found in Hawaii, Tahiti, and Fiji, as
referable to one species. In addition to the Polynesian forms there are
only two or three species, found in the Philippines, Madagascar, and
Ceylon, and it is with the species from the last-named locality that the
Polynesian species is by some identified.

The seeds of several other Leguminous climbers would probably act in a
similar way, for instance, those of Entada scandens; but the seeds of
this plant experimented on by me were too few to enable an opinion to be
formed. Of four seeds of Dioclea violacea from Fiji that were subjected
to the same experiment as those of Strongylodon lucidum, all floated in
sea-water after a year, with the exception of one that did not swell and
sink until after ten months. On the other hand, in my experiment in Fiji
on the fresh seeds of Canavalia obtusifolia, a plant found on tropical
beaches all round the globe, seventy per cent. sank in the first six or
seven weeks, swelling and displaying the first signs of germination, but
quite ten per cent. were afloat after three months.

My experiments on the foregoing and other littoral species of the
Leguminosæ merely indicate that under the ordinary temperature of
tropical currents a portion of the seeds will probably sink owing to
abortive attempts at germination. It is likely that if in the
experiments in England a constant temperature of 85° to 90° F. had been
sustained throughout, most if not all of the seeds would have swelled
and sunk within a month or two. The temperature of the experiments in
Fiji and Hawaii did not exceed that of many tropical currents; but there
are areas of superheating in equatorial seas, which I think would prove
insurmountable barriers in the path of most drifting Leguminous seeds, a
subject to which further reference will be made.

Coming to the Convolvulaceæ, my experiments show that the buoyant seeds
often lose their floating powers from the same cause. Those of Ipomœa
pes capræ may be taken as an example. I was surprised to find when
experimenting on the buoyancy in sea-water of these seeds in Fiji and
Hawaii that a considerable proportion, about a third, sank in the first
two months, swelling and sinking to the bottom. That this swelling
represented the early stage of germination was well brought out in
parallel experiments in fresh water and sea-water made in England on the
buoyant seeds of the British littoral species, Convolvulus soldanella. A
good proportion of the seeds in the first part of the experiment
absorbed water, swelled, and sank, those in fresh water proceeding at
once to germinate healthily at the bottom, whilst those that sank in
sea-water merely decayed. Of the survivors about fifty per cent. in
either case floated after six months. It may be added that the seeds of
other tropical littoral species, such as those of Ipomœa glaberrima and
I. grandiflora, behaved in the same way.

It would appear from my experiments, and it is a result that we should
expect, that buoyant seeds of the Leguminosæ and Convolvulaceæ would
often float for much longer periods under cool than under warm
conditions. There must be areas of high temperature in mid-ocean that
would prove much more fatal to the chances of a drifting tropical seed
than the icy waters of a Polar current. In my paper on Keeling Atoll I
have described how I procured the germination of a seed of Ipomœa
grandiflora, Lam., after a year’s flotation in sea-water in London,
which included a period of three weeks when the water temperature was at
or about 32° F. These seeds from this point of view would be exposed to
much more risk of sinking through abortive attempts at germination when
drifting across some parts of the Pacific Ocean. It would appear from
the Admiralty Chart of Surface-Temperatures, published in 1884, that
such an area with a surface-temperature of 83° to 86° throughout the
year extends north and east of New Guinea well into the Pacific,
reaching in the first half of the year as far east as the Tahitian
region. It would seem highly probable that the immersion of Leguminous
or Convolvulaceous seeds for many months in these tepid waters would in
most if not in all cases induce incipient germination which would lead
to the sinking of the seed. There are, however, exceptional cases, as
that of Cæsalpinia bonducella, which, as my experiments recorded in
Chapter XVII. indicate, appear to be quite proof against any conditions
of temperature such as are likely to be found in tropical seas in the
present day.

There are a few general considerations arising out of the foregoing
observations to which reference may now be made. The study of the
behaviour of the floating seed or fruit often carries us, as I have
before implied, to the borderland of vivipary. When from a canoe on a
Fijian river we lift up the germinating fruit of Barringtonia racemosa
from amongst the drift floating past in the stream and pull down from
the branches overhead the seedling a foot in length of Rhizophora, we
hold in our hands the two extremes of the series of vivipary. With many
of the plants of the mangrove-formation there is a fine adjustment with
respect to the germinating capacity of the seed, or in other words a
delicate balancing of organisation on one side and of physical
conditions on the other. A slight disturbance of the equilibrium would
produce great results in plant distribution. Thus, an elevation of the
temperature of the sea-water in the tropics to 90° F. would, I
apprehend, produce the abortive germination of nearly every floating
seed and fruit in equatorial seas, even of those of the beach-trees like
Barringtonia speciosa and Terminalia littoralis that are regarded as
proof against such risks under existing conditions where the
surface-temperatures would average 78° to 80°.

There would thus be a barrier to the dispersal of plants by currents as
effective as that of a frozen ocean. In the warm, humid climates of the
early geological ages, seed-transport by currents may have been often
impossible, since the seeds that did not begin to germinate on the
plants of the swamps would probably do so in the tepid water of the sea.
Viviparous plants would, however, be placed at no greater disadvantage
than they are at present, since the genera Rhizophora, Avicennia, and
others are now only dispersed by the floating seedlings. But such an
increase of temperature at the present time would mean the death in the
current of the floating seeds and fruits of nearly all non-viviparous
shore-plants. As a rule every Leguminous and Convolvulaceous seed would
swell up and go to the bottom; whilst fruits like those of Barringtonia
racemosa and Carapa obovata, that often germinate afloat in tropical
estuaries, would invariably do so under the changed conditions, and the
seedlings not being adapted for ocean transport would perish.

Yet we know that with the seeds of many inland plants temperature has
seemingly very little to do with starting the process of germination. We
are familiar with the fact that the seeds of many plants that fail to
germinate in the summer of their production habitually germinate under
apparently less favourable conditions of temperature in the following
spring. This is attributed by botanists to the immaturity of the seed on
first falling from the plant, a further period of maturation being
necessary before, under any conditions, germination is possible.

We see this also well illustrated in the floating seeds and fruits of
the Thames drift. Most of them fail to germinate in the drift at the end
of the summer and the beginning of autumn, and defer the process until
the following spring, when they germinate freely in the water under much
cooler conditions than those which they experienced in the early part of
their flotation in the drift. There are, however, exceptions to this
rule. Plants like Caltha palustris, for instance, are rarely represented
in the spring seed-drift of ponds and rivers, because most of the fruits
or seeds germinated soon after falling into the water in the previous
summer.

In most of my sea-water experiments in England the immersion had a very
marked influence, not in causing premature germination and destroying
the germinating capacity, as often happens with the floating seeds of
Convolvulaceæ and Leguminosæ, especially in the tropics, but in
postponing without injury to the seed the process of reproducing the
plant. Such seeds or fruits when placed in fresh water after many months
of flotation in sea-water germinated very freely in a few days, whilst
those left in the sea-water under precisely the same conditions remained
unchanged. This is true of many of the seeds and fruits found in the
Thames drift, such as those of Ranunculus repens, Lycopus europæus,
Rumex, &c. A striking instance was also afforded by the seeds of
Arenaria (Honckeneya) peploides, where seeds transferred directly to
fresh water, after many months flotation in sea-water, germinated in a
few days; whilst those left in the sea-water remained unchanged. This
subject is discussed at length in Note 19, and needs no further mention
here.

If the seeds of many plants in Great Britain postpone through immaturity
their germination to the following or even to the second spring, it goes
without saying that this does not exclude temperature as the ultimate
determining factor in germination. The immaturity of seeds adds another
link to the series of the germination-range in plants. This range begins
with the plants where germination takes place on the tree and the
seedlings hang suspended from the branches, as in the typical mangroves
Rhizophora and Bruguiera. Here, as is shown in Chapter XXX., there is
evidently no period of repose between the completion of the maturation
of the seed and the commencement of germination. The range ends with the
detachment of immature seeds which ripen apart from the parent plant,
and may postpone the germinating process for months and often for years.
All intermediate stages exist between these two extremes. Thus the
seedling may at once detach itself from the parent as in Avicennia, or
the germinating process on the plant may be limited to the protrusion of
the radicle as in Laguncularia, or the seeds may be quite mature and
ready to germinate as soon as they fall to the ground, as we find with
many small seeded plants. All the stages, of which only a few are here
indicated, are full of suggestiveness for the student of plant-life.

This subject is dealt with from other standpoints in Chapter XXX., but
the reader will now see more clearly what was meant when I said that the
study of the behaviour of the floating seed leads us to the borderland
of vivipary. In this range of the germinating process we may possess an
epitome of the history of the climatic conditions of plant-life from an
early era in the world’s story, beginning with those ages when perhaps
under the uniform conditions that then prevailed, all plants were more
or less coast-plants and more or less viviparous, and coming down to the
present era when with an extensive and varied land-surface there is
great variety both in climate and in the range of germination. The
mangrove-swamp and its viviparous trees would thus represent from this
point of view a condition of things once more or less universal on the
globe.

_Summary of the Chapter._

(_a_) The tendency of the floating seed or fruit to germinate in the
brackish water of tropical estuaries is especially characteristic of the
plants of the mangrove-swamp and their vicinity; but with those of the
beach trees that occur in the river-drift it is rarely if at all to be
observed.

(_b_) From the wide distribution of plants of the mangrove-formation it
is evident that this readiness of the floating seed or fruit to
germinate is not prejudicial to the dispersal of the species.

(_c_) It may perhaps be in the main attributed to a strain of vivipary
running through all the plants of the mangrove-formation, which finds
its extreme development in the viviparous species, where germination
takes place on the tree. But it is probably favoured by the superheating
of the waters of tropical estuaries.

(_d_) In the case of the buoyant seeds of several climbers and creepers
of the Leguminosæ and Convolvulaceæ, more or less littoral in their
station, it is shown that in warm water, whether fresh or salt, a good
proportion are apt to sink through incipient germination, which results
when the experiment is made in sea-water in the death of the embryo.

(_e_) Though in tropical currents of ordinary temperature a good number
of such floating seeds would escape this risk, it is argued that there
are certain warm areas in the tropical seas that would prove much more
fatal to the chances of these drifting Leguminous and Convolvulaceous
seeds than the icy waters of a polar current. It is thus held that these
seeds often sink in mid-ocean in tropical latitudes through abortive
germination.

(_f_) The study of the behaviour of the floating seed or fruit leads us
to the borderland of vivipary. In the scale of the germinative capacity
of plants it is possible to arrange a continuous series that commencing
with the mangroves, where germination takes place on the tree, ends with
those numerous inland plants where seeds are liberated in an immature
condition.

(_g_) It is suggested that the viviparous habit may have been the rule
under the uniform climatic conditions of early geological periods and
that with the differentiation of climates that marked the emergence and
extension of the continental areas the viviparous habit has been lost,
except in those regions of the mangrove-swamps which to some extent
retain the climatic conditions once general over the globe. With
differentiation of climate the true seed-stage with its varying
rest-periods has been developed.

CHAPTER X

THE RELATION OF THE BUOYANCY OF SEEDS AND SEEDVESSELS TO THE DENSITY OF
SEA-WATER

TO find amongst the results of my numerous experiments examples
illustrating the influence of density on flotation has not been so easy
as I at first imagined. Excluding all adventitious causes of buoyancy, a
matter discussed in Note 40, it may be inferred that the great majority
of seeds and fruits sink both in fresh water and sea-water. Of those
that are buoyant many float indefinitely in both waters, whilst in a
very few cases, where the floating power is derived from an outer fleshy
covering, as with the fruits of Potamogeton natans, the fruits float a
much shorter time in sea-water than in fresh water, on account of the
injurious effect of the salt upon their coats.

Experiments have to be specially directed towards this subject. It would
be useless to experiment in fresh water at one time and in sea-water a
month later. Nor would it answer to employ seeds and fruits from
different localities, since variations in this way sometimes occur. It
is necessary that the experiments should be made on seeds or fruits
collected at the same time and place, and that they should be
simultaneous and carried on under the same conditions. As the discussion
proceeds, the reader will perceive that many interesting points are
opened up, and that such an investigation, instead of being, as the
title of this chapter might suggest, an abstruse and disconnected
inquiry, is of considerable importance in relation to the dispersal of
plants through the agency of currents.

Guided by the results of my experiments in this direction I will proceed
to lay down certain general principles:—

(A) In the first place it may be accepted as a general rule that _seeds
or seedvessels that sink in fresh water sink also in sea-water_, the
difference in density between the two being rarely a factor of any
importance in determining buoyancy. The great majority of seeds and
fruits come under this category, since, as is pointed out in Chapter
VIII., only a small proportion of the whole, say a tenth, possess
floating power. We might cite, as illustrative of this principle in
temperate regions, almost all the 240 species included in the
non-buoyant group of the British plants experimented on (see Chapter
III. and Note 10). As a general rule this is true alike of the small
seeds of the Cruciferæ and Scrophulariaceæ, of the nutlets of the
Labiatæ and Boragineæ, of the genus Scirpus, and of the dust-like seeds
of Juncus. The results of my experiments on the plants of the tropical
Pacific are no doubt typical of other tropical regions; and if I wished
to quote instances, I should have to enumerate not only most of the
plants without buoyant seeds or fruits that are mentioned in the Fijian
and Hawaiian lists given under Notes 2, 4, and 6, but also to appeal to
tropical regions generally.

(B) One can carry the principle above-named yet further and say that not
only as a rule do seeds or fruits that sink in fresh water sink also in
sea-water, but that so far as tested _many of them sink in water of much
greater density than that of ordinary sea-water_ (1·026). Thus, for
instance, the seeds of Nuphar luteum, Scrophularia aquatica, and
Stellaria aquatica, the nutlets of Polygonum persicaria, and the achenes
of Aster tripolium sank in sea-water the density of which had been
raised to 1·050, the limit of the experiment. The minute seeds of Juncus
communis and J. glaucus and the larger seeds of Luzula campestris, even
after drying for six months, sank in salt water having a density of
1·075. It would, however, seem probable that for most of these small
seeds and seedvessels a density of 1·100 would prove to be the critical
point. If this is so, then most of those that sink in sea-water would
float in the dense water (1·160) of the Dead Sea.

However, my investigations have only gone a small way in this direction;
and perhaps some of my readers will pursue the inquiry. I will take the
case of the nutlets of Scirpus palustris. They sink in fresh water and
in sea-water, or may float in rare cases for a day or two. Out of 100 of
these seed-like fruits, 25 floated in salt water of a density of 1·075,
13 in water of 1·050, 7 in sea-water (1·025), and 3 in fresh water,
(1·000). It would thus appear that the proportion of buoyant nutlets is
doubled with every increase of ·025 of the density scale. At this rate
of increase they would all float in salt water of a density of 1·125,
which may be regarded as the suitable medium for the flotation of the
fruits of this Scirpus.... The seeds of Glaucium luteum, the Sea-Poppy,
have no buoyancy either in fresh water or in sea-water even after
prolonged drying. They all sank in water of a density of 1·050, but 18
per cent. floated when the density was raised to 1·075. At the rate of
increase noticed in the case of Scirpus palustris, all the seeds would
float in water of a density of 1·130-1·140.... The acorns of the Common
Oak (Quercus robur) have usually but little buoyancy unless they have
been long drying. After soaking in fresh water for half an hour 100
mature fruits, without the cupule, that had been kept a fortnight, I
found that only 2 floated in fresh water, 6 in sea-water (1·025), and 18
in water of 1·050. At this rate of increase all would float in water
having a density of 1·080-1·090.

(C) There is also another general rule, and it is this:—_Seeds or fruits
that float for a long time in sea-water usually float almost as long in
fresh water_. Here belong the greater number of buoyant seeds and
fruits, those only able to float for a few weeks being comparatively
few. Now with the long-floating seeds and fruits, those for instance
that float in the drift of English rivers from the autumn to the spring,
or those that are transported by currents over the tropical zone, there
is, as a rule, but a slight difference between their flotation periods
in fresh water and sea-water. If one of them sinks after floating for
several months in fresh water, it will sink in sea-water a few days
after. Fruits of Scævola Kœnigii, pyrenes of Morinda citrifolia, and
seeds of Thespesia populnea, Ipomœa grandiflora, Cæsalpinia bonducella,
and of different species of Mucuna, that had been kept afloat for a year
in sea-water, floated just as buoyantly in fresh water at the close; and
in those cases where any sank during the course of the experiment, it
was ascertained that they were able to float in fresh water almost to
the end.

That many of the seeds and fruits of tropical littoral plants that are
known to be dispersed by the ocean-currents will float well in fresh
water is shown in the constant occurrence in the floating drift of
Fijian estuaries, where the water may be quite fresh or brackish, of the
seeds and fruits of plants like Cerbera odollam, Clerodendron inerme,
Entada scandens, Heritiera littoralis, Ipomœa pes capræ, Morinda
citrifolia, Mucuna, Vigna lutea, &c. In the same way I noticed afloat in
the Guayaquil River in Ecuador, when the water was quite fresh, seeds
and fruits characteristic of the sea-drift, such as those of Anona
paludosa (seeds), Entada scandens, Ipomœa, Mucuna, Vigna, &c.; and when
we supplement observation with experiment, as for instance in the case
of Anona paludosa, we find that they will float equally long in fresh
and sea-water.

The same rule prevails with most of the buoyant seeds and seedvessels of
plants of the British flora—seeds and fruits, as I may remind the
reader, that are mostly to be found in river and pond drift. I am not
able to distinguish any difference of importance in the results of the
separate fresh-water and sea-water experiments. Thus with the seeds or
seedvessels of Bidens cernua, several species of Carex, Galium palustre,
Iris pseudacorus, Lycopus europæus, Ranunculus repens, and numerous
others, the difference after a flotation of many months was but slight.
If the results of the separate experiments were to be compared, there
would be at least ninety afloat in fresh water for every hundred afloat
in sea-water; and if at the end of a sea-water experiment, whether
occupying three, six, or twelve months, the seed or fruits were to be
placed in fresh water, quite nine-tenths and sometimes more would remain
afloat. A striking illustration of the principle that the excess in
density of sea-water, as compared with fresh water, adds but little to
the floating capacity of seeds is to be found in the results given in
Note 41 of simultaneous experiments made some years since by Mr. Millett
and myself at Marazion and in London on the seeds of Convolvulus
soldanella.

(D) In their relation, therefore, to the density of fresh water and
sea-water, most seeds and seedvessels may be placed in _two principal
classes, the first including quite four-fifths of the total, where they
are much heavier than sea-water, and the second comprising most of the
remainder, where they are much lighter than fresh water_.

(E) It would be surprising, however, if there were not some seeds or
seedvessels that come between these two extreme groups; some, indeed,
that have a specific weight approximating to that of fresh water, or to
that of sea-water, or fluctuating between them, and presenting such
evidence of a fine adjustment that the observer, forgetting that they
are members of a series, might be apt to regard them as specially
adaptive in their origin. It will thus be seen that this subject is
gradually assuming a statistical character; and in truth we shall
ultimately recognise here the play of the laws of numbers.

As an example of the plants where the specific weight of the seeds or
fruits is near that of fresh water, Alisma plantago may be taken. In the
course of an experiment, by lowering the density of the water from 1·025
to 1·020, I sent a shower of floating carpels to the bottom. The results
vary considerably, as one might expect; but, generally, during the first
few days of an experiment about twice as many (sometimes in all as much
as 80 per cent.) sank in fresh water as in sea-water, a few only
floating in either water for long periods.... The seeds of Arenaria
peploides present an example where the specific weight is between that
of fresh water and of sea-water. For the purposes of dispersal they may
be considered as heavier than fresh water and lighter than sea-water.
The details are given in Note 18; but it may be remarked here that
plants possessing seeds or fruits that sink in fresh water and float in
sea-water are very rare. As indicated below, this is what we might look
for on statistical grounds.

Plants whose seeds or fruits are not much lighter than sea-water are
exceptional. In such cases the effect of increased density of the water
is to extend the period of flotation. Thus, in my experiments on the
nutlets of Scirpus maritimus, the majority of the fruits floated in
fresh water only eight to ten days; whilst in ordinary sea-water they
floated in most cases two to three weeks; but when the density was
raised to 1·050, the greater number of them were afloat after two
months. In a few plants, as with Spiræa ulmaria, the effect of the
difference in density between fresh and sea-water was not to extend the
period of flotation, but to increase the number that floated for a given
period, the extreme limit of the buoyancy of the carpels in either water
with this species being about three weeks.

Amongst tropical plants, as illustrated by those of the Pacific islands,
cases also came under my notice where the mean specific weight of the
seed is somewhere between those of fresh water and sea-water. The seeds
of Afzelia bijuga, an inland as well as a littoral tree in Fiji, offer
an interesting example. If we place 100 seeds of a littoral tree in
sea-water, we find that on the average about 70 float. If then we lower
the density gradually, some of the seeds begin to sink at once; and on
the removal of the survivors to fresh water, about 47 will remain
afloat. The results may thus be stated:—Out of 100 littoral seeds, 30
are specifically heavier than sea-water (1·025); 23 are between
sea-water and fresh water in specific weight; whilst 47 are lighter than
fresh water (1·000). When, however, we take 100 seeds of inland trees,
we find that on the average 87 are heavier than sea-water, 5 are in
weight between sea-water and fresh water, and 8 are lighter than fresh
water. The significance of these figures becomes evident when we arrange
them in curves. The combined result for littoral and inland seeds is
given in the diagram below; and we see there, what is also indicated
with the separate curves that we are dealing with a double series, one
concerned with seeds lighter than fresh water, and the other with seeds
heavier than sea-water. The reader can himself supply the separate
curves for the littoral and inland seeds. The point, however, to notice
is that if a botanist with a statistical bent were to make a
miscellaneous collection of the seeds of the Vesi (Afzelia bijuga) in
one of the Fijian islands, in order to test their buoyancy, he would
obtain such a result as is given in this diagram. Two varieties of the
tree would be at once indicated, and further research would indicate
that these varieties were connected with littoral and inland stations.
This subject is further dealt with in Chapter XVII.

+------------------------------------------------------------------------------+
|Combined results for 200 seeds of Afzelia bijuga (100 littoral; 100 inland). |
+--------------------------------+------------------------+--------------------+
| | | | |
|Percentage.| Heavier than sea- | Between sea-water and | Lighter than fresh |
| | water, or +1·025. | fresh water in weight. | water, or -1·000. |
+-----------+--------------------+------------------------+--------------------+
| | | | |
| 100 | | | |
| | | | |
| +--------------------+------------------------+--------------------+
| | | | |
| 80 | | | |
| | | | |
| +--------------------+------------------------+--------------------+
| | | | |
| 60 | | | |
| | | | |
| +--------------------+------------------------+--------------------+
| | . | | |
| 40 | . | | |
| | . | | |
| +--------------------+------------------------+--------------------+
| | . | | |
| 20 | .| | . |
| | |. | . |
| +--------------------+------------------------+--------------------+
| | | . . | |
| 0 | | . | |
| +--------------------+------------------------+--------------------+
| | | | |
+-----------+--------------------+------------------------+--------------------+

It might seem strange that the seeds of Entada scandens should come into
the category of seeds with a specific weight near that of fresh water;
yet my observations in Fiji indicate that such is the case. In the
discussion of this plant in Chapter XVII. it is pointed out that, as a
rule, not more than a fourth will float in a river when they are first
freed from the pod, and not more than fifty per cent. will float in the
sea. Those that float, however, in either water will usually float
indefinitely. The seeds also of Mucuna gigantea D.C. are not very much
lighter than fresh water. Out of six seeds that floated in sea-water
buoyantly, five floated in fresh water, but heavily.

It is of interest to notice in this connection that the
mangrove-seedlings produced by germination on the tree, as in the case
of Rhizophora and Bruguiera, have a mean specific weight somewhere
between fresh water and sea-water. This is often illustrated in a
curious way, when the seedling has not been prematurely detached from
the tree. Thus in the sea off the coast of tropical America, as well as
amongst the Fijian Islands, the seedlings of Rhizophora mangle are as a
rule to be observed floating horizontally; whilst in the fresh or
brackish water of the estuaries of these regions they assume a more or
less vertical position, only the plumular portion protruding above the
water. This is also true of the seedlings of Rhizophora mucronata, the
Asiatic mangrove, and of Bruguiera rheedii. This subject is discussed in
detail in Chapter XXX.; but it may be here remarked that a good
proportion of Rhizophora seedlings, when detached in the mature
condition from the tree, have no buoyancy, between 20 and 50 per cent.
going to the bottom when they fall into a river, and between 5 and 10
per cent. when they drop into the sea. The navigator might often obtain
an indication of the density of the sea-surface when approaching the
mouth of a large river by observing the floating Rhizophora seedlings (a
foot long) which are carried out to sea in numbers. If he sees them from
the deck of his ship floating horizontally he will infer that the
surface-water is mainly sea-water. In ordinary fresh water when they
float vertically he would not be able to distinguish them from floating
seeds or fruits.

It has only been possible to treat this subject in an illustrative
manner. More details might have been given; but I have gone far enough
to bring the following points into relief and to justify one in drawing
the conclusions to be now stated.

_Most seeds and seedvessels in respect of their floating powers tend to
gather around two centres or means and to form two groups, the sinking
group and the buoyant group._

_In the sinking or non-buoyant group, which includes 80 per cent. of the
whole, the mean specific weight is considerably greater than that of
sea-water (1·026), which would require its density to be raised to 1·100
in order to serve as a floating medium for many of them._

_In the buoyant group the mean specific weight is much lighter than that
of fresh water (1·000); and from this it is to be inferred that in
oceans of fresh water the same fruits and seeds in the mass would be
distributed by the currents that are transported by them at the present
day. Even though it arose from an ocean of fresh water, the coral island
would receive the same littoral plants through the agency of the
currents that it receives under its existing conditions._

The number of plants with seeds or fruits between fresh water and
sea-water in specific weight is very small, probably not over 2 per
cent. of the total. Most seeds or fruits that sink in fresh water sink
also in sea-water, and most that float in sea-water float also in fresh
water. _Nature has thus created a wide gap between the sinking and the
floating seed; and nearly all of the work of the present currents in
plant-dispersal might have been effected, so far as the density is
concerned, in fresh water._ She has not arranged seeds and seedvessels
in what the statistician would term “a good series.” As indicated in the
diagram below, there are two series that meet in the neutral region
where the density is between fresh water and sea-water, but with
culminating points placed on the one side far above the density of
sea-water and on the other far below that of fresh water.

+-------------------------------------------------------------------------------+
| Relation of the specific weight of seeds and fruits to the density of |
| fresh and sea-water. |
+--------------+--------------------+-----------------------+-------------------+
| Percentage. | Heavier than sea- |Between fresh and sea- | Lighter than fresh |
| | water, or +1·026. | water, 1·000-1·026. | water, or -1·000. |
+-------------+--------------------+-----------------------+--------------------+
| | | | |
| 100 | | | |
| | | | |
| +--------------------+-----------------------+--------------------+
| | | | |
| 80 | | | |
| | | | |
| +--------------------+-----------------------+--------------------+
| | . | | |
| 60 | . | | |
| | . | | |
| +--------------------+-----------------------+--------------------+
| | . | | |
| 40 | . | | |
| | .| | |
| +--------------------+-----------------------+--------------------+
| | | . | |
| 20 | | . | |
| | | . | |
| +--------------------+-----------------------+--------------------+
| | | . | . |
| 0 | | . .| |
| | | . | |
| +--------------------+-----------------------+--------------------+
| | | | |
+-------------+--------------------+-----------------------+--------------------+

I do not, therefore, think that the buoyancy of seeds and fruits has had
any relation either in the present or in the past to the density of the
sea. Nor is it to be supposed that any slight variations in density in
the course of ages would have materially affected the dispersal of
plants by currents. It is to be inferred that the physicist and the
geologist would be prepared to grant only small variations, such as a
change from 1·020 to 1·025. It will be gathered from what has been said
before that changes of this nature would have a very slight influence on
the buoyancy of seeds and fruits, since the plants they would affect
would be very few. The change that the student of plant-dispersal would
require to produce any marked alteration in distribution would be in
amount alarming to the physicist.

Whether or not the oceans have been getting fresher or salter in the
course of ages (see Note 42), we will be moderate in our demands, and
will listen to the physicist when he argues that a diminishing density,
for instance, from 1·035 to 1·025, in the course of ages might explain
some of the peculiar features in the present isolation of insular
floras. Many seeds, he would contend, that could float across an ocean
having a density of 1·035 would be unable to accomplish it when the
density fell to 1·025. It has, however, been remarked that the critical
point of density for the flotation of seeds or fruits that sink under
present conditions is probably about 1·100. Cases of such a fine
adjustment to the density of sea-water are too few to endow this
argument with any weight. Or it might be suggested that with a gradual
increase in density in the lapse of ages seeds might float now that sank
before, or they might float for a longer period. Such a change, however,
would not have much effect, since nearly all the seeds and seedvessels
that sink in our rivers sink also in our seas, and a much greater
increase of density is required to make any difference.

Yet, although we might term the sinking of a seed or fruit an accidental
attribute of certain plants, just as we might regard the floating of a
log as an accidental attribute of a pine, since in either case the
specific weight might have been acquired without any direct relation to
the density of water, still the sinking of the seed or fruit signifies a
profound distinction not only, as is stated below, in plant
distribution, but, as we shall see later on, in plant-development.
Especially striking, says Prof. Schimper (p. 153), is the dependence
between an over-sea area of distribution and a station at the coast in
the case of species of the same genus of which some belong to the
littoral and some to the inland flora. In the first place, as has been
often remarked in these pages, we have a wide distribution generally
associated with considerable buoyancy of the seeds or fruits. In the
second case the areas are usually very restricted and there is little or
no buoyancy. The better fitted a seed or fruit is for dispersal by
currents the greater, therefore, is the area of the plant. Whether such
an important relationship depends on an accidental attribute of the seed
or fruit is the question that immediately presents itself. But it is
obvious that in raising such a question we touch on a very vital point
in adaptation, since if attributes developed in one connection have a
profound influence in another we may have to rearrange some of our
fundamental notions of the inner workings of Nature.

Let us, therefore, look a little closer into this matter, and turn again
to the Pacific islands. The present state of things may be thus tersely
described. Whilst the shore-plants dispersed by the currents have
remained relatively the same, changes of all kinds, from the production
of a variety and of a species to the development of a genus, have taken
place in the inland floras. Now, let us imagine that all this is altered
and that every seed or fruit is buoyant. There would then be but little
distinction between the strand and inland floras, since they would be in
a constant state of interchange, and most species would be widely
distributed. A relatively monotonous aspect would belong to all insular
floras, and indeed to much of the plant-world, since isolation, one of
the principal conditions for the origin of new species and new genera,
would often not exist.

On the other hand, let us suppose that all seeds and fruits were
non-buoyant. The agency of birds would then be alone available for
stocking new islands with most of their plants. The conditions of
isolation would be intensified. There would be no widely-ranging
strand-flora, since every island and every stretch of continental
sea-board would possess its own littoral plants that could only reflect
the peculiarities of the inland flora. The only determining factor
between coast and inland plants would be the presence or absence of the
capacity or organisation for occupying a station on the sea-shore.

We have now proceeded far enough to disclose the far-reaching influence
on plant-distribution and on plant-development that the relation between
the specific weight of seeds and fruits and the density of sea-water
must possess. Yet it has been shown that when such a relation is viewed
statistically it has an accidental aspect. We will accordingly devote
the next few chapters to the discussion of the buoyancy of seeds and
fruits from the structural standpoint.

_Summary of the Chapter._

(_a_) The great majority of seeds and seedvessels (quite 80 per cent.)
are much heavier than sea-water, but a noticeable proportion are
considerably lighter than fresh water, whilst those with a specific
weight near that of fresh water or of sea-water are very few.

(_b_) The buoyancy of seed and fruit has no direct relation to the
density of sea-water, and even if the ocean was deprived of all its
dissolved salts, the agency of the dispersal of plants by currents would
not be materially affected.

(_c_) Small changes in sea-density, such as the physicist would allow,
would, therefore, have no appreciable influence on the operations of the
currents as plant dispersers; and only great changes in density, such as
are presented by the waters of the Dead Sea, would add materially to the
number of floating seeds and fruits.

(_d_) Although the specific weight of seeds and fruits in its relation
to sea-density may be regarded as an accidental attribute, their
non-buoyancy in the great majority of plants has had a far-reaching
influence not only on plant-distribution, but on plant-development. The
plant-world would be transformed if all seeds and fruits floated in
sea-water.

(_e_) If the floating seed or fruit displays a quality that, so far as
the density of the sea is concerned, has been developed in quite another
connection, we have next to inquire whether the structure of such
buoyant seeds and fruits also affords evidence of non-adaptation.

CHAPTER XI

ADAPTATION AND MEANS OF DISPERSAL

BEFORE entering into a discussion of the causes of the floating powers
of seeds, it is necessary that I should state my general position on the
relation between capacities of dispersal in the organic world and the
question of adaptation. Adaptation runs through all the organic and
inorganic worlds, and we cannot conceive an universe without it. The
naturalist who looks only for the end in the purpose served makes but a
partially legitimate use of the phrase. On the other hand, it has been
improperly appropriated by those who hold to the theory of Natural
Selection, as indicating the result of small fortuitous variations that
have chanced to be of service to the species in the struggle for
existence. There is no question here of any end in view. Nature is
represented as working blindly, and the result of such “fortuitous
variation” is termed an adaptation. We cannot, however, pick and choose
only adaptations that are very evident in their character. We must
include everything in the organic world as an adaptation, whether
apparent or not, that is in direct relation with the organism’s
conditions of existence. It is not conceivable that an organism can be
adapted to conditions outside its environment, and yet many so-called
adaptations are of this character.

Nature—and I here confess my belief in a determining agency working
above and through all living and dead matter, but largely controlled and
checked by the laws of the physical world—Nature, as I apprehend, has
never concerned herself directly with providing means of dispersal
either for plants or animals. With regard to plants, she makes no direct
provision for the distribution of their fruits or seeds. If she had done
so, she would have employed some uniformity in her methods, as in the
instance of the means of reproduction; whereas the modes of dispersal
are almost infinite in their variety. When I say that Nature makes no
direct provision for the dispersal of plants and animals, I mean not in
the sense that a bird is adapted for an aerial life, or an aquatic plant
for a more or less submerged existence. That a bird is often able to
distribute its kind over a great area is the “accident” of its
conditions of existence. In a similar way the wide distribution of the
“ticks” that they carry round the world is due to the parasitical habits
of these insects, habits that have been acquired without any view to
their mode of dispersal by birds.

Similarly it cannot be said of seeds or fruits that are transported by
birds, whether adhering to their plumage by means of hooks or hairs, or
through some viscid excretion, or inclosed in soil adhering to the feet
or legs, or carried in the stomach and intestines, that Nature has made
any special provision for their dispersal. The dispersing agencies take
advantage of certain capacities or characters of a seed or fruit that
have been developed in the plant for quite other reasons and in
conformity with quite other principles. There may be mentioned as
examples the mucosity of seeds, the fleshiness of fruits, the occurrence
of hairs and prickles, &c. Yet as far as their connection with dispersal
is concerned, such capacities and characters are blind results in the
history of the plant’s development, the dispersing agencies making use
of what was not intended for them.

“Adaptation to definite life-purposes,” as Sachs terms it (_Physiology
of Plants_, 1887, p. 122), is seen everywhere; but it is adaptation
restricted to the organism’s conditions of existence. It is not
conceivable, as I have said, that an organism can be adapted to
conditions outside its environment. If there is such a seeming
adaptation, it is but a blind result, the accidental outcome of
collision or contact between two sets of conditions. If we represent a
number of these sets of conditions by several circles gradually
increasing in size until they encroach on each other, we find that the
circles lose their form and acquire a polygonal shape. All characters
seemingly connected with modes of dispersal have only this indirect
relation to such agencies; and their utility in these respects is an
accident in the plant’s life. They have not been acquired in connection
with the dispersing medium, but are the products of the laws of growth
and heredity, guided by a determining agency, and acting within the
organism’s conditions of existence. It is within these narrow limits
that all evident adaptations lie. In matters outside the conditions of
the development of seeds and fruits, the evolutionary or determining
principle “lets them go.” Detached from the plant, they come in contact
with conditions for which they were never created. The predominant power
in Nature, that brings to a successful issue the development of an
organism, has its limitations, and this is one of them, the evolutionary
or determining influence being ever checked and hampered by the laws of
the inorganic world.

I can only refer briefly to some of the reasons that have led me to
apply this view of the duality of forces in Nature to the subject of
plant-dispersal. The principles of evolution and adaptation rule the
world except in matters of dispersal. Take, for instance, the fleshy
fruits which the gardener often makes more attractive to birds than they
are in the wild condition. The result is certainly to increase their
facilities for dispersal by birds; but such a result was as little
intended by man as it was by Nature when species of Cornus, Ficus,
Prunus, Viburnum, and other genera matured their drupes, berries, and
fleshy fruits in the Cretaceous epoch.

Children are now taught in several excellent little books on
“Nature-Study” that fleshy fruits are specially adapted to be eaten by
animals to secure the distribution of the seeds. We read in one book
that plants produce these fruits “on purpose to be eaten,” in another
that they are “intended to be eaten,” and in a third that the
seed-coverings are adaptations, all with the ulterior object of
distribution by frugivorous animals. I must be pardoned if I venture to
express my dissent from these statements, more especially since they are
made by authors from whom it might be thought almost impertinent for me
to differ. Yet authority can be claimed for holding the opposite view.

When the botanist speaks of “useless secretions” in a plant, he is
alluding amongst other things to the sugar and organic acids of fruits.
“How and why all these substances originate is,” as Professor Sachs
observed in the work before quoted, “not known.” It is, however,
suggested by Dr. Kerner, in his _Natural History of Plants_ (Engl. edit.
i, 460-462), that such secretions, though useless to plants, may exist
for the purpose of alluring animals to assist in seed-dispersal. There
are some botanists, it may be remarked, that would reject such a view of
the nature of fruits. Dr. Stapf in his memoir on the flora of Kinabalu
observes in this connection that the fact that a fruit is fleshy and
attractive to birds is “no proof that it is really devoured by them, and
still less that it is dispersed by them.” Neither in fleshy fruits, nor
in minute seeds, nor in seeds capable of being transported by the wind
does he regard the general object of the particular character as
primarily to act as a means of dispersion.

The same plea is made for the mucosity of seeds like those of Capsella
and Plantago (see Note 43), or for the “stickiness” of other seeds and
fruits like those of Pisonia, qualities that favour adherence to passing
objects. This is the reason, we are told, why seeds are “sticky.” Such
secretions I infer are often materials lost to the plant; and being in
that sense excretory we are not called on to supply a use for them. They
can, therefore, not be regarded as having any teleological significance,
since adaptation arises only from the requirements of the plant’s
conditions of existence. If they are serviceable in assisting the
distribution of seeds, such an event can only be described as an
accident in the plant’s life arising from chance contact with another
environment.

The appendages of seeds and fruits, such as hooks and hairs, that render
them liable to adhere to fur or feathers, are also regarded as special
adaptations to this end. Without entering into the physiological
significance of hairs and prickles generally, concerning which, as many
of my readers will know, much might be said not in favour of such a
view, I would refer to cases like that of Cæsalpinia Bonducella, where
the large prickly pods could not possibly be intended to aid the plant’s
dispersal, whilst the leaf-branches are also prickly, and the seeds are
well known to be distributed by the currents. There are other cases like
that of Bidens cernua where the achenes, by reason of their barbed
bristles, and on account of a layer of “buoyant tissue” in the
fruit-coats, are dispersed both by birds and by water. We may fitly ask
to which capacity the theory of adaptation should be applied. Spiny
fruits may be sometimes so large, as in the instance of Trapa natans,
that the question of adaptation to dispersal cannot be raised.

The great variety of the modes of dispersal of seeds is in itself an
indication that the dispersing agencies avail themselves in a hap-hazard
fashion of characters and capacities that have been developed in other
connections. Seeds and fruits, having developed certain characters under
a particular set of life-conditions, on being detached from the parent
plant are brought into contact with conditions quite outside their
original environment. Qualities and capacities are then brought into
play which have no connection with the life-history of the plant. The
care with which the mother plant guards the maturing seeds, and the
protection of the environment, are at a certain period withdrawn, and
the seeds are left to take their chance under strange conditions. It
would be idle to see anything purposeful in the waste that results.
Rather we would see in it the effect of one of the numerous limitations
of the determining or evolutionary power in Nature. Such a power has to
adapt its workings to the laws of the physical world, checked here,
frustrated there, at times, as in this particular case, losing all
control, but in the end prevailing.

My general position may be thus summarised. As concerning the
distribution of fruits and seeds, the dispersing agencies take advantage
of characters and capacities that were never intended for them,
characters and qualities indeed that are often only brought out in
relation to another environment. Thus no question of adaptation as
regards means of dispersal can arise, since such capacities for
dispersal have no connection with the plant’s life-history. That seeds
are dispersed at all is a blind result of the ever-continued struggle
between the opposing forces of evolution and adaptation; that is to say,
between the determining power that lies behind organic life and the
physical conditions to which it has to adapt its ends.

CHAPTER XII

THE CAUSES OF THE BUOYANCY OF SEEDS AND FRUITS OF LITTORAL PLANTS WITH
ESPECIAL REFERENCE TO THOSE OF THE PACIFIC ISLANDS

IN the following pages I have adopted in its main features the
classification of buoyant seeds and fruits employed by Professor
Schimper in his work on the strand-flora of the Indo-Malayan region. The
causes of buoyancy, as he points out, are very various, but they can be
arranged in a few categories; each category, however, usually admitting
great variety within its limits. It is this want of uniformity that
first attracts our attention when we come to study the structure of
seeds and fruits from the standpoint of their buoyancy. Whilst in the
Pacific I went over most of the field traversed by Professor Schimper in
Malaya (the majority of littoral plants of these regions being common to
both), and as a result I have added not a few plants to his original
groups.

It will be seen from the following synopsis that there are three
principal groups. The first group includes those seeds and fruits where
the buoyancy is derived from unfilled space in the seed or fruit cavity.
The second group comprises those seeds or fruits where the floating
power is due to the buoyant kernel or nucleus. The third group includes
those where the buoyancy arises from the existence of air-bearing tissue
in the coverings of the seed or fruit.

The first two groups I will term the mechanical or non-adaptive groups,
not only on account of the structure inducing the buoyancy, but because,
as Professor Schimper remarks, the same structure often occurs with
inland fruits and seeds possessing little or no floating power. In many
of these cases, as he points out, the question of adaptation to
dispersal by ocean currents cannot, therefore, be raised. The third
group may be named the adaptation group, because it is on these examples
of buoyant seeds and fruits that this investigator chiefly based his
contention that in the main the structures concerned with buoyancy
represent adaptations to dispersal by currents effected through the
agency of Natural Selection. It is accordingly to this group that
Professor Schimper especially directed his attention, and the result of
his observations made in the home of the plants and of his
investigations in the laboratory has been the elucidation of many
difficult points in the structure of their fruits and seeds. To the two
“mechanical” groups he did not pay the same attention; and as their
examination came more within the limits of my own capacity as an
inquirer I have worked them out with some detail, the subdivisions of
the first group being my own as well as much of the material.

_Synopsis of the buoyant fruits and seeds of littoral plants of the
tropical Pacific classified according to the cause of buoyancy._ (The
authorities are indicated by the initial letter, S = Schimper, G =
Guppy. Details are given under some of the species in latter part of
volume.)

FIRST GROUP.—The floating power is derived from unoccupied space in the
cavity of the seed or fruit, no part of the seed or fruit as a rule
possessing independent floating power.

SUB-GROUP I., where the seed is concerned.

SECTION I. The seeds have little or no albumen, and neither the tests
nor the seed-contents have any buoyancy. The cotyledons are generally
large, foliaceous, and crumpled or folded, or otherwise arranged, so
that the seed-cavity is incompletely filled.

S. G. Hibiscus tiliaceus.
G. Hibiscus diversifolius.
S. G. Thespesia populnea.
S. Suriana maritima.
G. Kleinhovia hospita, _variable_.
S. G. Colubrina asiatica.
S. Dodonæa viscosa.
G. Argyreia tiliæfolia, _variable_.
G. Ipomœa bona nox, _variable_.
G. Ipomœa glaberrima, Boj.
S. G. Ipomœa grandiflora.
S. G. Ipomœa pes capræ.
G. Ipomœa turpethum, _variable_.
G. Cassytha filiformis.
S. Euphorbia atoto.

_Notes._—The species marked “variable” have seeds that sometimes sink
and sometimes float. With the exception of Kleinhovia they are only at
times littoral in station.

The plants of the British flora are represented by Convolvulus
soldanella and C. sepium, the last being “variable” and not a littoral
species.

SECTION II. All the seeds belong to the Leguminosæ. Neither the tests
nor the seed-contents have any buoyancy, the floating power arising
from a large central cavity produced by the bending outward of the
cotyledons during the final shrinking stage of the maturation of the
seed.

S. Mucuna (generically).
G. Mucuna urens D.C. (Hawaii).
G. Mucuna, species of.
S. G. Vigna lutea.
S. G. Cæsalpinia bonducella.
G. Cæsalpinia bonduc.
G. Entada scandens.

SUB-GROUP II., where the fruit is concerned.

SECTION III. The seed only partially fills the fruit-cavity, and as a
rule is not buoyant. The fruit shell, usually woody, may be also
buoyant.

S. G. Heritiera littoralis.
G. Smythea pacifica.
G. Dalbergia monosperma.
S. G. Derris uliginosa.
S. G. Pongamia glabra.
G. Desmodium umbellatum.
G. Gyrocarpus jacquini.

SECTION IV. The floating power is derived from empty seed-cavities,
where owing to abortion of the ovule or some similar cause the seed
is not developed.

S. G. Morinda citrifolia.
G. Premna tahitensis.

_Note._—Professor Schimper, in the case of Morinda citrifolia, holds the
view that we have here a special adaptation to dispersal by currents.

SECOND GROUP.—Here the floating power is due mainly or entirely to
buoyant kernels. In the case of seeds the tests are non-buoyant; but
with “stones” the floating capacity may be aided by a layer of
air-bearing tissue inside the shell.

SECTION I. Non-Leguminous.

S. G. Ximenia americana (drupe).
S. G. Calophyllum inophyllum (drupe).

_Note._—Professor Schimper would place these two plants in the second
section of the third group on account of the layer of air-bearing tissue
inside the shell of the “stone”; but they are assigned to this section,
since the floating power is mainly due to the buoyant kernel.

Arenaria (Honckeneya) peploides, a British beach plant, belongs here.

SECTION II. Leguminous seeds.

G. Dioclea.
G. Strongylodon lucidum.
S. Canavalia (generic).
G. Canavalia sericea.
S. G. Canavalia obtusifolia.
S. Erythrina (generic).
S. G. Erythrina indica.
P. Erythrina ovalifolia (Penzig).
S. G. Sophora tomentosa.
G. Afzelia bijuga.
G. Lathyrus?

THIRD GROUP.—The floating power is due to the presence of air-bearing
tissue in the seed-tests or fruit-coats.

SECTION I. The buoyant tissue occurs at the outside or forms the
periphery of the seed or fruit. Unless otherwise indicated the fruit
is implied in the list below.

S. G. Carapa moluccensis (seed).
S. G. Carapa obovata (seed).
G. Inocarpus edulis.
G. Serianthes myriadenia.
G. Parinarium laurinum.
S. G. Barringtonia speciosa.
G. Barringtonia racemosa.
S. G. Pemphis acidula (seed).
S. Terminalia (generic).
S. G. Terminalia katappa.
G. Terminalia litorea.
S. Lumnitzera (generic).
S. G. Lumnitzera coccinea.
S. G. Guettarda speciosa.
G. Wedelia strigulosa.
S. G. Scævola Kœnigii.
S. G. Cerbera Odollam.
G. Ochrosia parviflora.
S. G. Cordia subcordata.
S. G. Tournefortia argentea.
S. G. Clerodendron inerme.
G. Vitex trifolia.
G. Vitex trifolia, var. unifoliolata.
G. Tacca pinnatifida (seed).
S. Nipa fruticans.
S. Cocos nucifera.
G. Scirpodendron costatum.

_Additions of shore-plants from Malaya and tropical America mostly given
in Schimper’s work on the Indo-Malayan strand-flora._

S. Cynometra cauliflora.
S. Conocarpus erectus.
S. G. Laguncularia racemosa.
S. Lumnitzera racemosa.
S. Sonneratia (seed).
S. Barringtonia excelsa.
S. Scyphiphora hydrophyllacea.
S. Wollastonia glabrata.
G. Hippomane mancinella.

_Note._—Here belong a species of Vitex, probably V. agnus castus, the
fruits of which occur in the stranded drift of the Sicilian beaches, and
also the British littoral shore-plants, Cakile maritima, Crithmum
maritimum, Matricaria inodora, and Scirpus maritimus.

SECTION II. The buoyant tissue forms a layer inside the hard test of a
seed or inside the shell of the “stone” of a drupaceous fruit, and to
this cause the floating power is mainly or entirely due.

G. Mucuna gigantea (seed).
S. Hernandia peltata.
S. Excæcaria agallocha.
S. Cycas circinalis.
S. Pandanus odoratissimus.
G. Anona paludosa (seed) of tropical America.

_Note._—I have followed Schimper in respect to Pandanus, but it might be
by some placed in the first section of this group.

Here belongs Euphorbia paralias, a British littoral plant, the buoyant
seeds of which occur in the stranded seed-drift of English and
Mediterranean beaches.

In the following general discussion of the groups, reference will be
made only to the plants best illustrating the different varieties of
structure connected with buoyancy; whilst mention of the other plants
will in some cases be found in other parts of this volume, as shown in
the Index; and the matter is discussed at some length in not a few of
the species.

THE FIRST GROUP.

Of the first group, where the floating power is due to the unoccupied
space in the cavity of the seed or fruit, the Convolvulaceæ offer the
most typical examples. Here as a rule the crumpled embryo fills the
seed-cavity more or less incompletely; and it is on the relative size of
the unoccupied space that the sinking or floating of the seed depends.
In those plants where the seed sinks the seed-cavity may be almost
filled, as in Ipomœa tuberculata, or densely packed, as in Ipomœa
pentaphylla, and in species of Cuscuta. When the seed floats, as with
Ipomœa pes capræ, I. glaberrima, &c., the unoccupied space is relatively
large; and when, as with I. bona nox and I. turpethum, the behaviour of
the seeds is irregular, some floating, and others sinking, a
corresponding variation exists in the extent to which the seed-cavity is
filled. This applies also to the irregular behaviour of the seeds of
Ipomœa peltata and of Argyreia tiliæfolia. A singular instance is
afforded by the seeds of Ipomœa insularis, collected by me in Fiji and
Hawaii. Those from Fiji were incompletely filled, and consequently
buoyant. Those from Hawaii were more densely packed and sank.... The
three British species of Convolvulus illustrate the same principle,
namely, C. arvensis, with non-buoyant seeds; C. soldanella, with buoyant
seeds; and C. sepium, with seeds irregular in behaviour.

In the case of plants of the Convolvulaceæ, possessing buoyant seeds,
there is always evidence of marked shrinking of the seed-contents before
the final setting and hardening of the seed-coats. The embryo often
appears shrivelled and dried up, and is almost brittle, so that large
spaces are produced in the seed-cavity. If we partly divide such a seed
and place it in water, the embryo absorbs water rapidly, and within an
hour is soft, healthy-looking, and much swollen, the interspaces being
filled with a jelly-like mucilage. It is therefore evident that absolute
impermeability of the seed-coats is essential for the successful
transport by sea-currents of the floating seed; and we can only suppose
that the shrinking of the seed-contents takes place before the final
setting of the tests. That with the buoyant seeds the coats are quite
waterproof was illustrated in many of my experiments where, after a
period of flotation covering several months, and sometimes a year or
more, the seed-contents were still quite dry and shrunken. The limit of
buoyancy, as I have shown in Chapter IX., depends on an attempt at
germination on the part of the floating seed, which then absorbs water,
softens, swells, and sinks.

It is, therefore, not a matter of surprise that non-buoyant seeds of the
Convolvulaceæ do not gain floating power after prolonged drying of many
months. It is also to be expected that, as we find in Fiji, when a
characteristic shore-species with buoyant seeds like Ipomœa pes capræ
extends far inland, the seeds retain their floating powers.
Seed-buoyancy of this description is, on the face of it, purely
mechanical.

EXPLANATION OF THE DIAGRAMS ILLUSTRATING THE CAUSES OF SEED-BUOYANCY

1. _Entada scandens_ (natural size): (_a_), the shell; (_b_), the
kernel; (_c_), the intercotyledonary cavity. The shell consists of
three coats—an outer and an inner hard chitinous coat, and an
intermediate layer of brown cellular tissue containing little or no
air. The buoyancy is due entirely to the central cavity, neither the
seed-tests nor the seed contents possessing any floating power (see
page 181).

2. _Mucuna urens_, from Hawaii (natural size). The kernel (_b_) sinks,
and the shell has no floating power except where it possesses (under
the raphe) a layer of dark brown, air-bearing, spongy tissue (_a_).
This, however, is not sufficiently developed to endow the seed with
buoyancy, which is due to the intercotyledonary cavity (_c_). (see
page 111).

3., 4. _Mucuna gigantea_, from Fiji (natural size). The kernel (_b_)
sinks, and the seed owes its floating power entirely to the existence
in the shell (_a_) of a layer of brown, spongy, air-bearing tissue
which is mostly developed at the circumference and is almost wanting
at the flat sides of the seed (see page 115).

5., 6. _Dioclea_ (_violacea?_), from Fiji (natural size). Here the
kernel (_b_) is buoyant and endows the seed with floating power.
Though the shell (_a_) possesses a thick layer of reddish-brown
cellular tissue, this tissue contains but little air and aids the
floating power but slightly (see page 113).

7. _Strongylodon lucidum_, from Fiji (natural size). The floating power
is due entirely to the buoyant kernel (_b_). There is a very scanty
amount of loose brown tissue (_a_) under the raphe; but it has no
appreciable effect on the buoyancy (see page 113).

8., 9., 10. _Cæsalpinia bonducella_ and _C. bonduc_, from Fiji (natural
size). Neither the seed-tests (_a_) nor the kernel (_b_) have any
floating power in themselves, the buoyancy being connected with a
large internal cavity (_c_), which normally is intercotyledonary, as
in Fig. 8 (C. bonducella). With both plants, but more especially with
C. bonduc (Figs. 9 and 10), there may be a lateral cavity (_d_), or
the kernel may be loose in the shell (Fig. 10), but this does not
necessarily imply buoyancy (see page 194).

11., 12. _Arenaria peploides_ (enlarged: seeds 4 mm. in size). Here the
curved embryo (_a_) sinks, and the spongy air-bearing albumen (_b_)
gives buoyancy to the seed (see page 116).

13. _Euphorbia paralias_ (enlarged: seeds 3 mm. in size). The kernel
(_b_) sinks, and the seed owes its buoyancy to a layer of air-bearing
tissue (_a_) in the shell (see page 116).

14. _Morinda citrifolia_ (enlarged pyrene 7 mm. long). The floating
power is due to the bladder-like air cavity (_a_). The seed (_b_)
proper is enclosed in the woody tissue behind the bladder (see page
112).

15. _Cucurbita_ (seed enlarged), from the Valparaiso beach-drift (see
page 125). The kernel (_b_) has no buoyancy. The shell (_a_) is formed
of two layers of air-bearing tissue, the outer composed of prismatic
cells and the inner of a spongy vacuola-material.

[_To face page 111._

[Illustration: Diagrams illustrating some of the causes of
seed-buoyancy.]

Another type of the buoyant seeds of the first group is presented by
several species of Leguminosæ, as with Entada scandens, some species of
Mucuna, and Cæsalpinia bonducella. As with the Convolvulaceous seed, the
embryo sinks and the seed-shell has no buoyancy; but here the floating
power is due to the existence of a more or less symmetrical long central
cavity produced by the arching or bending outwards of the large
cotyledons which lie usually in close contact with the seed-shell. This
arching outward of the cotyledons depends on a shrinking process in the
setting or final stage of the maturation of the seed. The stages of the
process may be traced in the immature seeds, which are much larger and
in some cases twice the size of the mature seed. In this immature
condition the seed-coats are soft, and the flabby fleshy and thick
cotyledons fill up the seed-cavity. As the hardening and setting process
continues, the cotyledons diminish in size, become firmer, and gradually
bend outward, leaving a central cavity. This arching outwards is no
doubt in part the result of the contraction of the seed-tests during the
shrinking process. Considerable variation prevails in the results, and
where the cavity is very small the seed sinks. Further details relating
to this subject will be given in my treatment of some of the plants, and
especially under Cæsalpinia. But it may be here remarked with reference
to Hawaiian seeds of Mucuna urens D.C., that although they are strictly
referable to this group, they display beneath the hard test, on the side
beneath the raphe, a scanty layer of dark spongy air-bearing tissue
which is sufficiently buoyant to float up detached portions of the test,
but does not of itself give buoyancy to the seed. The significance of
this structure will be subsequently pointed out. The seed owes its
floating power to the large central cavity, but this layer of spongy
tissue adds to its buoyancy.

The section where the buoyancy of the fruit is connected with unoccupied
space in the fruit-cavity is extremely heterogeneous in its composition.
Every fruit has a method of its own, and the great variety of causes of
buoyancy of a mechanical character is here exemplified. For instance,
with Gyrocarpus jacquini and Cassytha filiformis the cause of buoyancy
is in the main the same as that described in the case of the
Convolvulaceæ. The origin of the floating power of the pods of Derris
uliginosa is two-fold. In the first place the seed or seeds but partly
fill the pod, and in the second place the seed is able to float of
itself by reason of its possessing, as in the seeds of Entada scandens,
a large central cavity produced by the arching out of the cotyledons
during the final stage of maturation. A double cause is also to be
assigned to the buoyancy of the fruits of Heritiera littoralis and of
Smythea pacifica, where, in addition to the unoccupied space produced by
the shrinking of the seed, the fruit-case itself floats, though nothing
but a mechanical explanation is to be given of the floating of empty
ligneous fruits.

One of the most suggestive types of buoyancy belonging to the first
group is presented by those cases, which are, however, not very
frequent, where the floating power is to be attributed to empty
seed-cavities produced by the abortion of the ovule or failure of the
development of the seed. A significant instance of this is afforded by
the fruits of Premna taitensis, a coast plant. The buoyant “stone” of
the drupe, which is often found afloat in the Rewa estuary in Fiji, is
4-locular, each cell containing normally one seed, but as a rule only
one cavity contains a mature seed, the three other cavities becoming
more or less empty through the failure of their seeds. It can be proved
that neither the seeds nor the substance of the “stone” are buoyant, and
that the “stone” owes its capacity of floating for months to the empty
cavities arising from the failure in development of three out of the
four seeds. In Fiji we see the rivers distributing these small fruits,
and we find the “stones” stranded on the beaches and floating in the
currents amongst the islands; and there can be no doubt that this is one
of the effective modes of dispersal of the species; yet, if there was
ever a case of accidental buoyancy concerned with dispersal by currents,
we have it here. Further details are given in Note 32.

It is probably also to the abortion of the ovule, or to the failure of
the seed, that the remarkable air-cavity (see Note 8) to which the
pyrenes of Morinda citrifolia owe their floating power, is to be
attributed. To this structure Professor Schimper (pp. 165, 183, 200)
attaches considerable importance as an example of special adaptation to
dispersal by currents through the influence of Natural Selection. He
suggests, however, that possibly its morphological significance may be
found in its being a peculiarly modified seed-chamber. The case of
Premna taitensis above cited indicates that the latter view is the most
probable. The subject awaits a careful microscopical study of the
seed-development of the genus Morinda since, as elsewhere remarked, the
non-buoyant pyrenes of inland species have not such an air-chamber. An
outline sketch of a pyrene of Morinda citrifolia is given in the
preceding plate. A good figure of it occurs in Schimper’s _Plant
Geography_, p. 28. A very suggestive instance of this nature is
described under Brackenridgea in Note 46 and in Chapter XIII.

THE SECOND GROUP.

Here are included those seeds and stone-fruits that possess buoyant
kernels. Professor Schimper points out that since this is a feature both
with inland as well as coast plants such a character cannot be viewed as
an adaptation to dispersal by currents. The plants concerned belong
mostly to the Leguminosæ, and we find here some of the most widely
spread of strand species, such as Canavalia obtusifolia and Sophora
tomentosa, as well as some of the giant climbers of the coast forests
belonging to the genera Dioclea and Strongylodon. The kernels when
divested of their coverings float buoyantly, but they soon absorb water
and sink usually in a day or two, a circumstance indicating that it is
to the impervious coverings that they indirectly owe their capacity to
keep the seed or fruit afloat. It is noteworthy that seeds of
Strongylodon lucidum from Fiji display beneath the raphe a trace of an
internal layer of loose cellular tissue which, however, has no
appreciable effect on the buoyancy; whilst with seeds of Dioclea
(violacea?) from the same locality there is a thick layer of loose
tissue which aids the floating power of the kernel but is not of itself
sufficiently aeriferous to buoy up the seed.

This leads one to refer to two other plants belonging to this group,
Calophyllum inophyllum (Guttiferæ) and Ximenia americana (Olacineæ),
where, though the floating power is mainly due to the buoyant kernel, it
is also aided by a layer of air-bearing tissue inside the hard shell of
the “stone” of the drupe. Professor Schimper places these fruits in the
third or adaptive group on account of the layer of buoyant tissue, but
it would be more correct to class them according to the predominant
cause of their buoyancy. It can be shown that with a non-buoyant kernel
the “stone” no longer floats. This double cause of the floating power
renders an explanation very difficult, since it would seem indefensible
to give conflicting interpretations of their nature. With Ximenia
americana there is another great difficulty. Its drupes are known to be
dispersed by fruit-pigeons (_Introd. Chall. Bot._ p. 46); and judging
from the rare occurrence of the “stones” in the drift there is good
reason to believe that bird agency in the Western Pacific is predominant
in the dispersal of the plant. It is by such test cases as this that we
must put to the proof the reality or non-reality of the influence of
adaptation on seed-buoyancy.

THE THIRD GROUP.

We have here those plants where the floating-power is entirely or mainly
due to an air-bearing tissue in the seed-tests or fruit-coats. Several
of the fruits are figured in Schimper’s _Indo-malayische Strand-flora_,
and one or two are figured in the English edition of his work on
_Plant-Geography_, p. 29.

In the first section, where the buoyant tissue occurs at the outside or
forms the periphery of the seed or fruit, are included several of the
most familiar of the littoral trees and shrubs of the Pacific islands,
such as Barringtonia speciosa, Cerbera Odollam, Guettarda speciosa,
Pemphis acidula, Scævola Kœnigii, Terminalia katappa, and several others
named in the synopsis. I cannot enter into detail here, but the reader
will find fuller particulars of each plant in most cases in Professor
Schimper’s work, and in some instances in my separate discussion of the
plants concerned. In nearly all cases we are concerned here with the
fruits, and only in a few cases with the seeds, as with Carapa and
Pemphis acidula.

This investigator observes that to this sub-group belong the fruits and
seeds usually described in systematic works as provided with corky or
suberous coverings; but he points out (p. 167) that the resemblance is
nearly always quite superficial, and is limited to colour and
consistence, suberous tissue occurring in only a few cases, as in the
fruit-coats of Clerodendron inerme. The buoyant tissues, he remarks, are
often more or less ligneous, and in those cases where there is no lignin
reaction they resist the action of sulphuric acid much more effectively
than pure cellulose; whilst in their physical characters, as well as in
their behaviour with reagents, they differ just as much from ordinary
cork. Thus, they are but little elastic and often easily crumble away;
whilst in large fruits, like those of Cerbera and Terminalia, they would
soon be stripped off entirely when subjected to the “wear-and-tear” of
transport by currents, if they were not traversed by numbers of stout,
tough fibres which hold the materials together. Where the buoyant
tissues are firmer, as with Clerodendron inerme and Cordia subcordata,
the fibrous framework is scanty or absent, whilst very small seeds or
fruits, like those of Tournefortia argentea and Pemphis acidula, where
the “wear-and-tear” would be comparatively slight, often possess no
protecting fibres in the buoyant tissues.

In one or two fruits, like those of Cerbera Odollam, these tissues
display large intercellular spaces; but in the majority of cases such
spaces are insignificant in size or absent altogether. Speaking
generally, however, there is, as Professor Schimper observes, great
similarity in the structure of the buoyant tissue in the coverings of
these fruits and seeds. The cell-walls are thin or only slightly
thickened, and detached air-bearing portions of the tissue will float
for many weeks. The great floating capacity of these fruits and seeds is
stated by this investigator to be entirely due to the tenacity with
which the air is retained in the covering tissues. It is, however,
noteworthy that in the case of Scævola Kœnigii the fruits are just as
well suited for dispersal by frugivorous birds as by the currents, a
significant circumstance discussed in the next chapter.

The second section contains those plants where the buoyant tissue occurs
inside the hard shell of the fruit or seed, such as is found, for
example, in Anona paludosa, Mucuna gigantea, Hernandia peltata, Cycas
circinalis, &c. Professor Schimper here includes Calophyllum inophyllum
and Ximenia americana; but I have before remarked that the buoyancy of
their fruits is mainly due to their buoyant kernels. This aeriferous
tissue forms a layer between the seed or nucleus and the hard outer
shell. It is described by the above-named authority as soft or friable
and dark brown. The cells contain air and may be closely arranged or
separated by small interspaces, their walls being neither woody nor
suberous.

_The structure of the buoyant seeds and seedvessels of the littoral
plants of the British flora._

The littoral plants with floating seeds or fruits form but a section of
the strand-plants of the British flora, scarcely a third, as is pointed
out in Chapter IV., of the total number. Though small in number they
exhibit great variety in structure; and notwithstanding that as far as
they have been examined they may all be referred to one or other of the
groups and sections of the classification adopted in the synopsis for
the plants of the Pacific islands, nearly every plant presents in the
structure of its seeds or seedvessels a type of buoyant structure
different from the others.

The first group is represented by the seeds of Convolvulus soldanella,
which owe their floating power to the incomplete filling of the
seed-cavity. The second group, where the buoyancy arises from the
buoyancy of the kernel or nucleus, is illustrated by the seeds of
Arenaria (Honckeneya) peploides, but in a fashion quite unique. The test
is thin but impervious, and has no buoyancy; the curved embryo also
sinks; and the floating power arises from the air contained in the loose
spongy albumen, around which the embryo is coiled (see figure). A more
normal component of the second group is represented in some Leguminous
seeds, perhaps of Lathyrus maritimus, that occur regularly amongst the
stranded seed-drift of the north coast of Devon. Here the kernel of the
seed is buoyant. The seeds of Euphorbia paralias are indebted for their
floating capacity to a layer of spongy tissue containing large
air-spaces placed between the kernel and the chitinous outer test,
neither of which possess any floating power (see figure). They thus
belong to the second section of the third group.

The fruits of Cakile maritima, Crithmum maritimum, Matricaria inodora,
and Scirpus maritimus, all belong to the first subdivision of the third
group where the air-bearing tissue exists in the peripheral coverings,
the seed or nucleus in all cases sinking. With Cakile maritima there is
a light spongy outer case of aeriferous tissue, which, however, soon
loses the epidermis, a circumstance that probably explains the limited
period of flotation of about a week. The walls of the mericarp of
Crithmum maritimum are composed of spongy cellular air-bearing tissue
with a persistent epidermis, and the floating powers of the fruits are
consequently great. The achenes of Matricaria inodora have beneath the
epidermis a layer of buoyant tissue, and their structure is similar to
that found with the buoyant achenes of littoral species of Wedelia,
plants of the same order of Compositæ that are found on the Pacific
islands. The cause of the floating power of the fruits of Scirpus
maritimus lies entirely, according to Kolpin Ravn, in the air-bearing
cells of the epidermis. The reader will find the results of my
experiments on the buoyancy of the seeds in Notes 16, 17, and 18.

_Summary of the Chapter._

(1) Following the main lines of Schimper’s classification of those of
the Indo-Malayan region which possesses for the most part the same
species, the buoyant seeds and fruits of the littoral plants of the
Pacific islands are classed in three groups: the _first_ where the
cavity of the seed or fruit is incompletely filled, the floating power
arising from the empty space; the _second_ where the buoyancy is derived
from the buoyant nucleus or kernel; and the _third_ where it arises from
air-bearing tissues in the coats of the seed or fruit.

(2) The first and second groups, in which the question of adaptation to
distribution by currents through the agency of Natural Selection is not
raised, since the same structural characters are found in seeds and
fruits of inland plants not dispersed by the currents, are termed the
mechanical or non-adaptive groups. The third is distinguished as the
adaptive group, because it is here that Schimper finds evidence in
favour of the Selection Theory.

(3) The first group is best represented by the Convolvulaceous and the
Leguminous types. In the former, which is well illustrated by Ipomœa pes
capræ, the seed-cavity is imperfectly filled by the crumpled embryo, the
result of the shrinking process during the final setting of the seed. In
the latter, which is exemplified by Entada scandens and Cæsalpinia
bonducella, the seed displays a large central cavity produced by the
arching outward of the cotyledons during the shrinking process
accompanying the last stage of the maturation of the seed. As an
instance of fruits belonging to the group, those of Heritiera littoralis
may be cited. An uncommon type is presented in the “stones” of the
drupes of Premna taitensis, and in the pyrenes of Morinda citrifolia,
where the buoyancy arises from empty seed-cavities resulting from the
failure of some of the seeds.

(4) The second group with buoyant kernels includes mostly widespread
Leguminous species, such as Canavalia obtusifolia and Sophora tomentosa.

(5) The third or “adaptive” group comprises many of the characteristic
littoral trees and shrubs of the Pacific islands, such as Barringtonia
speciosa, Guettarda speciosa, Terminalia katappa, Tournefortia argentea,
&c., that contain in their fruit-coverings a buoyant cork-like material
often bound together by fibres, but which proves on examination to
resemble cork only in appearance. In another type, illustrated by the
fruits of Cycas circinalis and the seeds of Anona paludosa, the buoyant
tissue forms a layer inside the shell of the seed or “stone.”

(6) Some fruits like those of Ximenia americana and Calophyllum
inophyllum illustrate both the so-called mechanical and adaptive
principles in their structure; whilst with the first-named species they
are as well adapted for dispersal by frugivorous birds and are known to
be a favourite food of fruit-pigeons. The same difficulty arises with
the fruits of some other characteristic littoral plants, as with Scævola
Kœnigii, the drupes of which are equally well fitted for dispersal by
birds and currents.

(7) The same general principles have been at work in determining the
structures concerned with the buoyancy of the fruits and seeds of
British littoral plants. Although the species are few in number they
exhibit in this respect great variety, eight species illustrating six or
seven types of buoyant structure.

CHAPTER XIII

ADAPTATION AND SEED-BUOYANCY

WHEN we speak of a certain structure as an adaptation to dispersal by
currents through the agency of Natural Selection, it is necessary at the
outset to be quite clear as to what is implied. Professor Schimper, who
brought his great and varied knowledge of many other phases of
plant-life to bear on this subject, is careful to clear the ground of
preliminary erroneous conceptions in such a perspicuous and impartial
manner that we cannot do better than follow his guidance. There are, he
observes (p. 178), many mechanisms or contrivances in plants, which,
though they seem to have arisen with a fixed purpose, can in no wise be
regarded as having been developed for that end, since they were produced
in quite a different connection and have merely acquired a new or
supplementary function, of which they are the cause and not the effect.

This is very much the position that I have taken up for the whole
subject of the relation between plants and their dispersing agencies,
and it will be found discussed in Chapter XI. It involves, as I venture
to think, a dominant principle in the organic world, which it is one of
the objects of this work to emphasise, namely, that Nature in dispersing
plants habitually makes use of structures and capacities that were
originally developed in quite another connection. Behind this change of
function, this new purpose, lies the secret of the organic world. There
is for me no more pregnant fact in plant-life than the thistle-seed
blown before the wind, or the seed of our sea-convolvulus floating in
the sea. It proves to my mind that the evolutionary power in nature is
checked and hampered by conditions not of its own creation, and that two
opposing forces are ever at work, the one creating and the other
limiting the creative power, the actual mode of dispersal being but a
blind and accidental result of the struggle.

The question of the operation of Natural Selection is not raised, as
Professor Schimper indicates, until we consider whether the new function
has had any bettering influence on the structure or mechanism with which
it has come to be concerned. If such a modification is thus brought
about it might be legitimately claimed as a result of this agency, and
the term “adaptation” could be used. But if there is no evident change
produced, we should be compelled to assign very subordinate limits to
the capacity of Natural Selection; and in the instance of buoyant fruits
and seeds it would be restricted to determining a plant’s station by the
water-side and in increasing its area. It is only in the first case that
we could speak of them as adaptations in the meaning attached to this
term in the language of the Selection Theory. It would at first sight
seem easy to ascertain whether the characters of fruits and seeds, to
which the buoyancy is due, are adaptations in this sense of the word;
but in reality it is far from being so. We can, however, proceed with
unanimity up to a certain stage in the argument; but there agreement
ends.

It has been before established that in the Pacific islands, and indeed
in the tropics generally, the plants with buoyant seeds or seedvessels
are mainly stationed at the coast. It has also already been shown that
this littoral station is often associated with a special buoyant-tissue
in the coverings of the seed or fruit; and it will now be pointed out
that this tissue is, as a rule, absent or but scantily developed in the
case of inland species of the same genus. Of great importance, remarks
Professor Schimper (p. 179), in relation to the Selection Theory and the
development of adaptations, is the comparison of the fruits and seeds of
strand-plants with those of allied inland species; and he finds here
evidence in support of the Darwinian view. He takes the cases of the
genera Terminalia and Calophyllum, which contain both inland and
littoral species; and he shows that although the same buoyant-tissue
occurs in the fruit-coats of inland species, it is there much
diminished, and in consequence the floating powers are considerably
lessened or lost altogether (see Chapter II.). It is not pretended that
this tissue has had any connection in its origin with dispersal by
currents, but merely that its greater development in the shore species
is an adaptation to this mode of transport.

Further testimony is adduced by this investigator (p. 182) in supporting
his view in the fruits of the genera Barringtonia, Clerodendron, Cordia,
and Guettarda, where the buoyant tissues extensively developed in the
coast species are either non-existent or only represented by a trace in
the inland species of the same genus, a difference in structure
associated with the loss or great diminution of the floating capacity of
the fruits concerned. I have been able to establish other examples in
the cases of the genera Scævola and Tacca, which will be found referred
to in Chapter II.

Professor Schimper (p. 200) points to the circumstance that the
“adaptations” in these fruits all belong to the diagnostic marks of the
genera and the species, and contends that these plants abundantly prove
the erroneous nature of the contention that Natural Selection could have
played no part in the elimination of the strand-flora. My own contention
is that Natural Selection has played such a part, but that in doing so
it has merely availed itself of characters previously existing, without
originating, modifying, or improving them in any way. The foregoing
evidence might with equal fitness be employed to show, as pointed out in
Chapter II., that in the course of ages there has been a great sorting
process by which, excluding the mangroves, plants of the xerophilous
habit possessing buoyant seeds and fruits have been sorted out and
placed at the coast. Direct evidence does not lead us farther than to
the establishment of a littoral station for plants thus endowed. The
problem whether the characters of their fruits and seeds that are
concerned with buoyancy may be regarded as adaptive in the Darwinian
sense lies beyond the reach of direct testimony. We can, however,
approach it from the outside by several directions, and from some of
these we will now proceed to deal with it.

There is first the singular circumstance that in Fiji, when the littoral
plants with buoyant seeds or fruits leave the beach and extend far
inland, they, as a rule, retain their floating powers and, of course,
their buoyant structures. I found this to be true of Cassytha
filiformis, Cerbera Odollam, Ipomœa pes capræ, Morinda citrifolia,
Scævola Kœnigii, and one or two other plants mentioned in Note 44, where
this subject is discussed. My experiments on these plants indicated that
their fruits or seeds floated equally long, whether obtained from coast
or from inland plants. This, at first sight, appears to present a
serious objection to the adaptation theory; but it was not so regarded
by Professor Schimper, who in a letter to me, dated March 8th, 1900,
observed that he did not see “why littoral plants growing inland should
lose their adaptations to littoral life, especially if those adaptations
are not conflicting with the conditions of life beyond the littoral
zone, and if the competition does not require special adaptations.”

My view, however, is that any process of adaptation is unnecessary. All
these plants, it is contended, were originally inland plants that
acquired the buoyant qualities of their seeds and fruits in the inland
stations, and ultimately found a station at the coast through the
sorting process above referred to. In the case of plants like Ipomœa pes
capræ and Cassytha filiformis this would be conceded, since they belong
to the acknowledged non-adaptive groups discussed in the preceding
chapter. It is only to some of these plants, such as Scævola Kœnigii and
Cerbera Odollam, that the adaptation view of Professor Schimper is
applied; and the question arises whether we are justified in making such
a distinction, or, in other words, whether it is antecedently probable
that two independent principles have been at work in determining the
fitness of seeds and fruits for dispersal by the currents.

The plants for which the influence of adaptation through Natural
Selection is claimed belong, as stated in Chapter XII., almost entirely
to the third group. It is admitted that with the other two groups the
utmost that any sorting or selecting process would effect would be to
determine a station at the coast and to extend the area of distribution.
The numerical aspect of the question therefore acquires some importance;
and the reader’s attention is accordingly directed to the results
tabulated in Note 45, where it is shown (assuming for the time that
there is no difference of opinion about the adaptive significance of the
seeds and fruits concerned) that the plants of the third or adaptive
group make up only about half the total. It would therefore appear that
if the agencies of Natural Selection have been at work here either in
bettering or in developing buoyant structures, half of the shore-plants
with buoyant seeds or fruits have not come within their influence.

But the subject takes another aspect when we reflect that in some
buoyant fruits, as with Ximenia americana and Calophyllum inophyllum,
the two principles would seem to have been at work. Whilst from this
standpoint Natural Selection is regarded as having either developed or
increased in amount the layer of buoyant tissue in the fruit-coats, the
buoyant kernels are not viewed as adaptive in their origin. In the case
of Ximenia americana the dispersing agency of frugivorous birds adds
another factor, since, as before stated, its drupes are known to be
dispersed by fruit-pigeons. In the cases of Scævola Kœnigii and of Vitex
trifolia, two plants belonging to the adaptive group, Professor Schimper
(pp. 156, 188) admits also the dispersing agency of frugivorous birds,
and he claims it for Morinda citrifolia, in the pyrenes of which he also
detects a special adaptation to dispersal by currents. It may be added
that, as he also points out, fruits of the non-adaptive group of
littoral plants, such as Premna integrifolia (P. taitensis) and Cassytha
filiformis, would sometimes also attract birds. In fact, those of the
last-named have been found in the crops of pigeons (_Introd. Chall.
Bot._, p. 46).

Looking at all these littoral plants with fruits that are equally fitted
for dispersal by birds and by currents, we may now ask, Where does the
general principle of adaptation to dispersal lie? Whatever view we
adopt, we must apply the same view to all, whether it be a question of
dispersal by birds or by currents. We cannot choose between two sets of
principles determining the buoyancy of seeds and fruits any more than we
can regard a fleshy drupe and a buoyant seed as illustrating different
principles regulating the dispersal of plants. Nature works with
uniformity in these matters, and if the Natural Selection theory is held
to explain one case it ought to account for all. Yet nobody would go so
far as this; and this view of dispersal is on many grounds antecedently
improbable. These difficulties disappear if we assume that in all cases
the dispersing agencies have without modification made use of characters
and capacities that were developed, as we now see them, in quite other
connections and under quite other conditions.

It will now be necessary to look a little closer into the subject of the
buoyant tissue, to the existence of which in their coats about half of
the littoral plants concerned owe the floating power of their fruits or
seeds. In the first place, it is to be remarked that in the case of some
of the seeds of the plants of the non-adaptive groups it is also
represented to a small degree in the seed-coats, although, as with
Strongylodon lucidum and Mucuna urens, it is not present in sufficient
amount to float the seed. In the next place, it should be noted that
with some genera possessing, like Terminalia, both inland and coast
species it is to be found alike in the fruit-coverings of inland and of
littoral plants, though in a less degree in the case of the fruits of
inland species, the floating power of which is proportionately
diminished. There are, however, a few cases where this buoyant tissue is
developed in inland species which belong to genera or subgenera that
have no littoral species. This is what we would expect, if Natural
Selection has merely concerned itself with placing plants of xerophilous
habit possessing buoyant seeds or fruits at the coast. Under such
conditions we would now and then expect to find an inland plant
possessing buoyant fruits or seeds of this description that has never
been able to establish itself at the coast.

A good instance is afforded by Pritchardia Gaudichaudii, a fan palm
peculiar to Hawaii, the drupes of which float for several weeks and have
a covering of spongy buoyant tissue (see Chapter XXV.). The seeds of
Hibiscus Abelmoschus, a species distinguished subgenerically from the
littoral Hibiscus tiliaceus, offer another example. They float for
months, and owe their buoyancy to a layer of air-bearing tissue between
the kernel and the test, in this respect differing from the seeds of the
littoral species, where the floating power is due to unoccupied space in
the seed-cavity. The buoyancy of the seeds of Hibiscus Abelmoschus thus
offers another example of ineffectual floating power, since it is not a
littoral plant, is often cultivated, and has accompanied aboriginal man
over much of the tropical zone.

A singular instance of the dispersal by currents of an inland plant that
occurs both wild and cultivated in tropical America, the West Indies,
and on the West Coast of Africa, is afforded by Spondias lutea, Linn.,
which is referred to at the end of Chapter XXXII. Its “stones,” which
are provided with a cork-like covering much as we find with those of
Cordia subcordata and Guettarda speciosa, possess great buoyancy, and
are found in the river and beach drift of those regions with the seeds
in a sound condition.

A very remarkable case of ineffectual buoyancy is presented by the
seedvessels of Brackenridgea, which have been found floating in the
drift off the coast of New Guinea. They owe their floating power to
closed cavities which would seem to arise from the failure of one of the
seeds or from the abortion of an ovule. But, according to Beccari, their
fleshy coverings would aid their dispersal by frugivorous birds; and
since the species are all much localised and are rarely littoral in
their habit, it is very probable that birds have mainly effected the
dispersal of the genus (see Note 46). It has, however, been shown in the
previous chapter that Premna taitensis and Morinda citrifolia owe their
dispersal by currents to similar cavities in the seeds or “stones.”

Amongst the inland plants possessing seeds or fruits that are dispersed
by the currents without aiding the distribution of the species may be
recognised types of both the adaptive and non-adaptive groups. A
singular instance is afforded by the large seeds almost an inch long of
a huge pumpkin (Cucurbita) which, in sound condition, form one of the
commonest constituents of the beach drift on the coast of Chile from
Valparaiso northward to Iquique. The fruit is commonly eaten by the
lower classes. The seeds, which are very buoyant, contain a kernel that
does not float, the buoyancy being due to the water-tight coats which,
as shown in the plate in Chapter XII., possess well developed
air-bearing tissues. It may here be observed that Martins refers to the
germination of seeds of Cucurbita pepo after 45 and 93 days’ flotation
in sea-water.

One sometimes finds buoyant tissue developed in the seeds of
bottle-gourds, where it can serve no useful purpose of dispersal. Thus
small bottle-gourds, seemingly of the genus Cucurbita rather than of
Lagenaria, are to be commonly found afloat in the Guayaquil River and
stranded on the Ecuador beaches. They will float for many months, and
contain the seeds dried up into a small loose compacted mass in their
interior. These seeds, which contain a layer of spongy air-bearing
tissue in their coverings, will in several cases float for months. Some
that I had been keeping two months afloat in sea-water germinated
freely. It is shown in Note 47 that bottle-gourds containing sound seeds
are dispersed far and wide by the currents. In some species the seeds
are buoyant, and in others they sink in sea-water; but the gourds
themselves will float for probably a year or more, and the floating
capacity of the seeds when it exists is too insignificant to affect the
fruit’s buoyancy.

Other instances of the useless buoyancy of fruits of inland plants are
afforded by different species of Citrus. In the floating drift of the
Fijian rivers the fruits of the wild and indigenous Shaddock (C.
decumana) and of an inedible Orange, also wild and indigenous (C.
vulgaris?), are at certain times to be found, the latter often in
numbers. The first-named floats four to five weeks in sea-water, and the
last-named nearly two months, and both are to be observed floating out
at sea between the islands. The fruits of the Tahitian Orange, a variety
of C. aurantium, floated in sea-water between three and four weeks. The
seeds of these and other species of Citrus sank in from a few hours to a
day or two. The buoyancy of the fruit depends on the rind—the thicker
the rind the greater the floating power. This was not only shown in the
length of the period of flotation, but also in the buoyant behaviour of
the fruit. With the Tahitian Orange, where the rind is relatively thin,
the fruits floated heavily in sea-water and only protruded slightly
above the surface. With the Shaddock and with the other indigenous
species of Citrus, the fruits floated lightly and protruded half-way out
of the water.

There is nothing trivial in these examples of buoyant fruits. That they
have at times aided in the dispersal of the genus, with man’s assistance
in planting the seeds of the stranded fruits, I cannot doubt; but
unaided by man such buoyant capacities would be useless for purposes of
effective dispersal by currents. Between the two genera Terminalia and
Citrus there is this great distinction, that the former is more or less
halophilous, some of its species being at home on the sea-beaches,
whilst the latter, as Schimper would term it, is salt-shy, and includes
no halophytes or plants of the sea-shore amongst its species. The only
effect of buoyancy of the fruits on the distribution of the species of
Citrus would be to place them by the side of the river and the pond.
This has evidently been its result in the case of the Shaddock in Fiji,
where, as Seemann remarks, it often thickly lines the banks of the
rivers.

As also indicating that the buoyancy of the seed or fruit would never,
apart from the halophilous habit, endow an inland plant with a littoral
station, the examples of the Oak (Quercus robur) and of the Hazel
(Corylus avellana) may be taken. As shown in Note 48, these fruits
acquire floating power by drying, on account of the space formed by the
shrinking of the kernel. They occur commonly in beach drift, but rarely
in a sound condition; yet experiment has proved that they will sometimes
germinate after prolonged sea-water flotation. The fruits of other
species of Quercus are also transported in tropical regions by the
currents, but never, as far as I could learn, effectively. The Amentaceæ
as an order are “salt-shy,” and with only a few exceptions shun the
sea-beach.

In the great sorting-process, by which xerophytic plants with buoyant
seeds or fruits have been placed at the coast, and hygrophytic plants
with similar fruits or seeds have been stationed at the riverside or by
ponds and lakes, one might expect to find that other influences may have
at times been in conflict with the selecting operation here indicated.
To this cause may probably be attributed the cases of “useless buoyancy”
above referred to. Here we find in some inland plants fruits and seeds
with buoyant tissues in their coverings that in the case of littoral
plants would have been regarded as the result of adaptation to dispersal
by currents. Such cases go to emphasize the conclusion already indicated
that these tissues could not have been developed through the agency of
Natural Selection. But the great objection against the application of
the Darwinian view to the general subject of the buoyancy of the seeds
and fruits of littoral plants lies in the circumstance that quite half
of the plants concerned are admitted to be outside the scope of the
theory, and that for these another explanation has to be found. I think
we may fairly claim that in a matter which finally resolves itself into
a question of buoyancy one explanation should cover all. We have thus to
decide whether to regard as adaptations to dispersal by currents the
structures of the buoyant seeds and fruits of littoral plants; or
whether to hold the view that as far as dispersal by currents is
concerned such structures are purely accidental, and that Nature has
never directly concerned herself in the matter at all. The first
explanation lies under the disadvantage above alluded to, and it remains
to be learned whether the second view could be made to cover all cases
of dispersal by currents. Further investigation on many points is yet
required; but, apart from the evidence against Natural Selection as the
principal agency that has been produced in this chapter, a powerful
argument in favour of the view that the buoyancy of seeds and fruits is
not concerned with adaptation is, that as a rule the floating capacity
of the seed or fruit has no direct relation with the density of
sea-water. Generally speaking, as shown in Chapter X., these seeds and
fruits are much more buoyant than they need to be, that is to say, if
they owe their floating power to adaptation to dispersal by currents.
This is quite in accordance with the argument developed in Chapter XI.
with regard to the general question of plant-distribution, that
dispersing agencies make use of characters and capacities of seeds and
fruits that were never intended for them.

_Summary of the Chapter._

(1) There are many mechanisms or contrivances in plants that now serve a
purpose for which they were not originally developed.

(2) Of this nature, it is contended, is the relation between fruits and
seeds and the agencies of dispersal.

(3) If, however, the structure or mechanism is made more effective by
the new function, such a modification may be regarded as an “adaptation”
in the language of the theory of Natural Selection.

(4) It is held by Professor Schimper that the structures connected with
the buoyancy of the fruits or seeds of several tropical littoral plants
are, in the above sense, adaptations; and he points to several genera
where the buoyant tissues in the coverings of the fruits or seeds of the
coast species are scantily represented or absent in the inland species
of the same genus, a difference corresponding with the loss or
diminution of the floating powers.

(5) This contrast in structure and in floating capacity between the
fruits or seeds of inland and coast species of the same genus is beyond
dispute, and the author adduces fresh data in support of it.

(6) But he contends that it is not proved that the relatively great
development of buoyant tissues in the case of littoral plants is the
effect of adaptation; and that if the selecting process had been
confined to sorting out the xerophilous plants with buoyant seeds or
fruits and to placing them at the coast, the same contrast would have
been produced.

(7) In support of this contention he points out that when such littoral
plants extend inland the floating capacity and the buoyant tissues are
as a rule retained; and that in those exceptional cases where inland
plants possess buoyant fruits or seeds these tissues are sometimes well
developed under conditions in which they could never aid the plant’s
dispersal.

(8) But the most serious objection against the adaptation view is that
admittedly only about half of the shore-plants with buoyant fruits or
seeds come within its scope. Therefore a second explanation has to be
framed for the other plants concerned.

(9) As showing the difficulties raised by regarding some of the
structures connected with buoyancy as “adaptive” and others as
“accidental,” it is pointed out that some fruits possess the two kinds
of structure. It is also shown that in several cases fruits endowed with
buoyant tissues are just as well adapted for dispersal by frugivorous
birds; and the instance of Ximenia americana is cited where a drupaceous
fruit, known to be dispersed by fruit-pigeons, possesses also in its
“stone” both the “adaptive” and “non-adaptive” types of “buoyant
structures.”

(10) It is urged that whatever is the relation between the buoyancy of
the seeds and fruits of shore-plants and dispersal by currents, there
has been a uniform principle affecting all.

(11) The weight of evidence is regarded as adverse to the Natural
Selection theory, an inference which is consistent with the conclusion
arrived at in Chapter X. that there is no direct relation between the
density of sea-water and the buoyancy of seeds and fruits, the floating
capacities being as a rule far greater than the adaptation view would
explain. Nature, it is held, has never made any provision for dispersal
by currents, the buoyancy of seeds and fruits being, as concerns the
currents, a purely accidental quality.

CHAPTER XIV

THE RELATION BETWEEN LITTORAL AND INLAND PLANTS

IN discussing the relation between the littoral and inland floras in the
Pacific it will be at first necessary to pick up some of the threads of
the various lines of investigation dealt with in the previous portion of
this work. Apart from considerations connected with the genetic history
of the plants concerned, when we come to inquire into the sources of any
individual strand-flora, whether in the temperate or in the tropical
regions, we arrive at the rough and ready inference that it is composed
of “what the sea sends and the land lends.” But it has been already
shown that the relative proportion of the current-borne and in
consequence widely dispersed plants in a strand-flora varies greatly in
different regions. Thus in the Pacific islands, as typified by those of
Fiji, about 90 per cent. have buoyant seeds or seedvessels originally
brought from distant localities; and in the tropics, as a rule, the
average would probably be never under 75 per cent. On the other hand, in
a temperate region the plants derived from inland would be most
predominant, making up probably some three-fourths of the whole, whilst
the proportion of current-dispersed plants hailing from distant places
would be relatively few.

It is on this account that there is such uniformity in the general
composition of the strand-flora over a large part of the tropics, since
current-dispersed plants are widely spread. But in the temperate regions
we find a great contrast in this respect. There are, it is true, a few
current-borne plants that one meets everywhere. For instance,
Convolvulus soldanella is to be gathered on English beaches and on those
of New Zealand and of the coast of Chile. But these littoral plants with
buoyant fruits hardly give a feature to the strand-flora. A multitude of
intruders, either characteristic of the inland flora of the region or
confined only to the seaboard of that part of the world, also make their
home on the beach and frequently endow a beach-flora with its leading
features. The possible associations of plants on a beach in a temperate
region are thus very great; and I have already discussed this in part in
Chapter IV. as concerning the British shore-flora. One has only to look
at a work like that of Dr. Willkomm on the vegetation of the strand and
steppe-regions of the Iberian peninsula to realise how the few littoral
plants familiar to the English eye cut but a sorry figure amongst the
numbers of strange intruders from the arid regions inland. So again, as
I found on the Chilian beaches, Convolvulus soldanella finds odd
associates amongst the species of Nolana and Franseria that are peculiar
to the coasts of that part of the globe (see Chapter XXXII.); and
different grotesque American forms of the Cactaceæ with a
Mesembryanthemum and a host of strange-looking plants descend from the
arid slopes of the hills behind to keep company with the far-travelled
English beach-plant (see Note 49). Or again, a glance at the pages of
Professor Schimper’s great work on _Plant-Geography_ will bring the same
fact home in a still more varied fashion.

Yet on tropical coasts the intruding inland element is also
distinguishable, though it may influence only to a small degree the
general character of the strand-flora. Dividing it, as we have described
in Chapter V., into the plants of the sandy beach and of the
mangrove-swamp, we find in the mangroves the most stable element and in
the beach-plants those most liable to change. Professor Schimper
observes that whilst the physiognomy of the beach-flora varies to some
extent with the alterations in the inland flora, the mangrove-formation
makes but a slow response to such changes. As he points out in his work
on the Indo-Malayan Strand-Flora (p. 199), seeds and seedvessels are
being continually brought down to the sea-coast through the agencies of
rivers, winds, and birds; and in this manner, in the course of ages, the
beach-flora is recruited from the inland plants. But for the mangroves
such additions to their numbers are rarely possible. Whilst the same
genera are often shared by both the beach and inland floras, we have in
the mangrove-formation families, sub-families, and genera almost
peculiar to itself, and including plants, like those of the Rhizophoreæ,
that in their characters betray but little kinship with others and give
but little indication of their descent. The mangroves have remained
through the ages as something apart from other coast-plants, isolated
both in their history and in their characters, and especially
distinguished by their “adaptations” to their surroundings.

Such is the line of argument followed by this eminent German botanist in
his account of the development of a tropical strand-flora. In various
parts of this work I have ventured to suggest that the mangroves may be
the remnant of an ancient flora widely distributed over the lower levels
and coastal regions of the globe in an age when vivipary (meaning,
thereby, germination on the plant) was the rule rather than the
exception. At such a period, as I imagine, the climatic conditions of
the earth were much more uniform than they are at present, at least in
the lower levels; and a warm atmosphere, charged with aqueous vapour and
heavy with mist and cloud, enveloped a large portion of the globe. The
mangroves, it may be remarked, are by no means universally distributed
on tropical coasts in our own time. (Professor Schimper describes their
distribution in his _Indo-malayische Strand-Flora_, pp. 85, 86, and in
the English edition of his _Plant-Geography_, p. 409.) They are not
found on rainless coasts even when under the Line, except where there
happen to be large estuaries; but where a rank and luxuriant inland
flora betokens a high degree of humidity, there they thrive. This is
well illustrated on the rainless shores of tropical Peru, a locality
described in Chapter XXXII. of this work.

Yet if, as it is here contended, the mangroves form a remnant of a once
widely spread viviparous flora, it might be expected that the
beach-plants of that age would have been also viviparous, and that with
their present descendants, as well as with some of the inland plants
allied to them, we ought to find in the anomalous structure of the seed
some indication of the lost viviparous habit. This appears to be the
case, as described in Note 50, with the Barringtoniæ, a tribe that has
supplied some of the most characteristic beach-trees, and also with some
genera of the Guttiferæ. Perhaps, indeed, when the seeds of several
other littoral beach-trees come to be examined, for instance, Guettarda,
analogous structures may be found.

Although the beach-flora of the tropics is less stable in its
composition than the mangrove-formation, it is not to be assumed that in
the Pacific region or in the tropics generally it is at all modern in
its character. Though in the main, no doubt, more recent than the
mangroves, since it is likely that in early geological periods the swamp
rather than the sandy beach formed the predominant feature of the
sea-border throughout the tropics, yet it bears in several respects the
impress of a high antiquity. There are few beach plants in the tropical
Pacific that are not found over the tropics of a large portion of the
globe, a circumstance that would in itself warrant our assigning a great
age to the beach-flora; and it is highly probable that some at least of
the beach plants of the Pacific that occur on the east and west coasts
of tropical America are, for reasons given in Chapter XXXII., older than
the barrier now interposed by Central America between the Atlantic and
Pacific oceans. There are, it is true, a few species, like Acacia
laurifolia and Drymispermum Burnettianum, which, on account of their
restriction to the beaches of the Western Pacific and their lack of
capacity for dispersal by currents, may be regarded as local
productions; but for the great majority, ranging as they do over much of
the tropics, it is not possible to determine when and where they assumed
their littoral habits. That except in a few instances their home in some
bygone age lay outside the Pacific can scarcely be doubted.

It is therefore to be expected that in a discussion of the relation
between the strand and inland floras in the Pacific islands the problem
will be mainly concerned with the possible derivation of inland from
littoral plants. In such a discussion the relation between the beach and
inland species of the same genus becomes a subject of great interest. It
is a subject that had a peculiar fascination for Professor Schimper, who
refers to it more than once in his pages; and though never able to take
it up, he viewed it as a very promising field of inquiry. The question
has been frequently alluded to in this work; and it is especially dealt
with in one connection in Chapter II. It is there shown that whilst, as
a general rule, the seeds or seedvessels of the coast species possess
great floating power, those of the inland species of the same genus have
little or none, and that both may have independent modes of dispersal,
the first by currents, and the last through frugivorous birds.

A close connection between the beach and inland floras is apparently
displayed in the circumstance that quite a third of the genera of the
Pacific insular floras containing littoral species (some 70 in all,
excluding the mangroves) possess in this region also inland species. But
the further examination of this interesting group of genera, which are
enumerated in the list below, goes to show that the connection between
the inland and coast species of a genus is by no means always so close,
or of such a character, as one might have expected. It will not be
possible, however, to do much more than indicate in this chapter the
results of this inquiry; but the details will usually be found either in
the separate discussion of the genus or in other parts of this work. For
convenience of treatment these genera may be grouped in the following
sections.

_Grouping of the Plant-Genera of the Islands of the Tropical Pacific
that possess both Littoral and Inland Species._

Section I. Where the littoral and inland species are most probably of
independent origin, both possessing their own means of dispersal;
Calophyllum, Hibiscus, Colubrina, Morinda, Scævola, Cordia, Ipomœa,
Vitex, Tacca, Casuarina.

Section II. Where the littoral species have probably given rise to
inland species, and both still exist in the group of islands: Vigna,
Premna.

Section III. Where inland species have been probably developed from
littoral species no longer existing in the group: Canavalia, Erythrina,
Sophora, Ochrosia.

Section IV. Where the littoral and inland species are evidently of
independent origin, and there is no means of accounting for the
existence of the inland species by agencies of dispersal at present in
operation: Barringtonia, Pandanus.

Section V. Where in the same genus some inland species are derivatives
of the coast species and others are of independent origin: Guettarda.

Section VI. Where the coast species, having little or no capacity for
dispersal by currents, are regarded as derived from the inland species
in one group of islands and as afterwards distributed to those in the
vicinity: Eugenia, Drymispermum, Acacia.

SECTION I

This group, which includes those genera where the coast and inland
species are regarded as of independent origin, both possessing their own
means of dispersal, contains about half of the total number of genera
here concerned. We will first deal with the genera Calophyllum, Morinda,
and Scævola, where the littoral species have buoyant fruits or seeds
that are dispersed by currents, whilst the inland species have more or
less non-buoyant fleshy fruits that could only be dispersed by
frugivorous birds. Here the inland and coast species could have arrived
independently at the island, and we are not called upon either on this
ground or by reason of affinity of characters to connect the one with
the other.

The genus Scævola is very typical of its kind and has been already in
part discussed in Chapter II. The wide-ranging shore-species, S.
Kœnigii, that is distributed over the Pacific may sometimes, as in
Hawaii, be accompanied by numerous inland species, all endemic, seven of
them being enumerated by Hillebrand; or, as in Fiji and Tonga, there may
be associated with it a solitary inland species, S. floribunda (see Note
51); or, as in Tahiti, it may exist by itself. On the other hand, as in
the Kermadec Islands, a single inland peculiar species may alone
represent the genus. The inland species have fleshy drupes which, as far
as examined, have no floating power and possess no buoyant tissues in
their coverings; and their independent dispersal by birds cannot be
doubted. The endemic character of most of the inland species of the
Pacific islands is most probably due to the suspension of the
transporting agency of frugivorous birds, just as the wide range of the
solitary littoral species may be attributed to the uninterrupted agency
of the currents. There is nothing in the description of the endemic
species given in Hillebrand’s _Hawaiian Flora_ to indicate any especial
genetic connection between the inland species and the beach plant, S.
Kœnigii; and the occurrence of a solitary inland peculiar species in the
Kermadec Islands clearly proves an origin independent of any littoral
plant.

Morinda is another critical genus in this discussion. Besides the
widespread littoral species (M. citrifolia) that is distributed by the
currents and is also dispersed by man, there are in the Pacific islands
a number of inland species, mostly climbers and denizens of the forests.
In the _Index Kewensis_ six are accredited to Fiji and five to New
Caledonia. Hillebrand gives a peculiar Hawaiian species, and there is a
widespread species (M. Forsteri) that ranges over the South Pacific from
New Caledonia to the Marquesas and the Paumotu Islands. Since, as
indicated in Chapter II. and in Note 8, the pyrenes of the fruits of the
inland species are not dispersed by the currents and could readily be
transported by frugivorous birds, we are not called upon to connect them
in their origin with M. citrifolia, the wide-ranging species of tropical
beaches.

The fact of the dispersal of certain inland species of the genus over
large areas of the tropics, such as in the case of Morinda umbellata
through tropical Asia and Malaya, and M. Forsteri in the Pacific, is
indeed sufficient proof that these inland plants are independent of any
littoral species in the Pacific and possess their own means of
distribution. Though the genus, comprising at least forty species, is
mainly confined to the Old World, there are a few species in America;
but M. citrifolia, the familiar beach species of the Old World and the
Pacific, is not indigenous there, and, as far as I can gather, all the
American species belong inland. Facts of distribution of this nature
negative the possibility that the Pacific islands have received their
inland species of Morinda through the intervention of the far-ranging
littoral plant.

As respecting Calophyllum, which is represented all over the tropical
South Pacific by the wide-ranging C. inophyllum and by a tree of the
inland forests found also in Malaya and in Ceylon (C. spectabile), there
are, apart from questions of affinity, grave objections against the
derivation of the same inland species from the coast species all over
this area. The fruits of the two inland species of Fiji, C. spectabile
and C. burmanni, have sappy outer coverings and are quite suited for
dispersal by fruit-pigeons. As observed in Chapter II. and Note 9, they
have limited floating capacities and their dispersal by birds is
necessary to explain their distribution. Since the timber is greatly
valued by the Polynesians, it is not unlikely, however, that those
islanders have assisted in the distribution of the inland species. It is
not possible to do more than touch on this subject here; but it may be
inferred that the history of Calophyllum in the Pacific has not been one
that would warrant our regarding the inland trees as derivatives of a
coast species.

There are other genera of this section where, for reasons of a different
character, there is no cause for assuming that the inland species are
derived from the coast species, or _vice versâ_. Thus, in Fiji,
Casuarina equisetifolia, a widely distributed species of the Old World,
occurs at the coast and in the scantily wooded plains behind; while C.
nodiflora, a New Caledonian species, finds its home in the lower
forests. There are many endemic species in Australia and New Caledonia;
and we are not called on to connect together these two species in Fiji.
In the same way we are not under any obligation in the case of the
numerous inland species of Ipomœa of the Pacific islands to connect them
with the coast species. They are all widely ranging species, and their
seeds have been carried to the islands, each in its own fashion. So
again with the inland species of Hibiscus found in the Polynesian
islands and often cultivated, we cannot either from the point of view of
dispersal or of affinity connect them with the far-ranging littoral
species, H. tiliaceus, which belongs to a section of the genus distinct
from those sections to which the inland species belong.

In a similar way there is no ground for supposing that Cordia aspera, an
inland species confined to Fiji, Tonga, and Samoa, is derived from C.
subcordata, the widely distributed littoral species of the Pacific and
of the Old World, since they belong to different sections of the genus.
But, apart from any question of affinity, the drupes of inland species
of Cordia are known to be well suited for dispersal by frugivorous
birds, though, unlike the littoral species above named, not adapted for
transportal by the currents. The genus Vitex, which is represented by a
wide-ranging littoral species in the Pacific (V. trifolia), appears to
be associated with inland species only in Fiji, where one or two,
seemingly endemic, occur. But there is nothing in Dr. Seemann’s
description of V. vitiensis, one of these species, that at all suggests
its derivation from the strand species, a very variable plant that often
extends far inland into the plains, adopting a different habit of growth
in those localities. It is known that Vitex fruits can be dispersed both
by birds and by currents. This genus is more fully discussed in a later
chapter.

Of the genus Colubrina there seem to be only two Pacific species
known—one the widely distributed shore-plant, C. asiatica, a straggling
shrub with alternate leaves found in all the Pacific groups and on the
beaches of much of the tropics of the Old World; the other a tree, C.
oppositifolia, with opposite leaves, that is peculiar to the Hawaiian
islands, where it frequents the open-wooded and scrubby inland
districts. The seeds of the shore-plant float unharmed for many months,
whilst the fruits of the inland plant, which differ in some important
respects (see Note 52), would float only for a week or two. The strand
species is also quite at home inland in many parts of the world; and
there is nothing from the standpoint of affinity to indicate that in
Hawaii it has given birth to an inland species so divergent in habit and
in character. There is of course the difficulty of explaining how a
plant like C. oppositifolia, with such a dry, unattractive fruit, could
be indebted to birds for its original introduction into the group; but
the same difficulty arises with a host of Hawaiian plants. It is,
however, evident from its distribution over the islands of this
archipelago that it possesses or has possessed some means of
inter-island dispersal, and since it is not of much service to the
aborigines we must look therefore to the bird.

In the instance of the genus Tacca there is in Fiji an inland species,
T. maculata, associated with a wide-ranging beach species, T.
pinnatifida, which also grows inland. The first-named is recorded from
the north coast of Australia and from Samoa, and though, unlike the
beach plant, its seeds are unfitted for dispersal by currents (see
Chapter II.), they might be distributed by birds. Dr. Reinecke describes
another inland species from Samoa, T. samoensis. The beach plant, T.
pinnatifida, grows so typically (sometimes side by side with T.
maculata) in the inland plains of Fiji that one would not be justified,
apart from questions of affinity, in regarding it as the parent form of
inland species in the Pacific islands.

For food and other purposes Tacca pinnatifida is or was much valued by
the Pacific islanders, and it grows so abundantly that cultivation is
rarely practised. That the Polynesians have aided the currents in the
distribution of the plant there can be no doubt, and this is
particularly indicated by its occurrence in Hawaii. The genus contains
ten or a dozen species, of which at least three are peculiar to America;
but T. pinnatifida, the characteristic shore-plant of the Old World, and
according to Schimper the only one that can be so designated, is not
found in America, where, as far as I can gather, there is no
widely-spread beach species dispersed by the currents from which the
peculiar species could have been derived. In the case of the Pacific
species, however, it should be noted that I am not endeavouring to prove
the improbability of the inland species having been derived from the
coast species in other regions, as in Australia, but that my point is to
show there is no reason to suppose that this has taken place in the
Pacific. There is no difficulty in attributing the dispersal of inland
species to birds; and we are therefore not called on to connect them
with the beach plants.

SECTION II

This division includes those genera where the littoral species has
apparently given rise to one or more inland species and both still exist
in the same group of islands. Two genera alone, Vigna and Premna, come
into this category. The first-named seems to present a good case for the
derivation of an inland from a coast species in Hawaii. Besides Vigna
lutea, the beach species, which is found not only all over the Pacific
islands but on the tropical beaches of the Old World, there are in
Hawaii two endemic species (V. sandwicensis and V. oahuensis) that occur
in the mountains, usually at elevations of from 1,500 to 5,000 feet; but
I do not find any more inland species recorded from the other Polynesian
archipelagoes. It may at first be noted that Vigna lutea, which in some
parts of the world strays inland, displays considerable variety in its
littoral station in the Pacific. Thus, in Hawaii, I found it sometimes
on the sandy beach, sometimes on a rocky shore, and sometimes on the
edge of old lava-cliffs overlooking the sea. In Fiji, though usually a
trailer on the beach, it may become a climber hanging from the trees
bordering the creeks in the mangrove-swamps. Though Hillebrand makes no
mention of forms intermediate between coast and inland species in
Hawaii, I found in one locality at the coast some specimens of Vigna
lutea displaying the twisted pods and two callosities on the standard
that are characteristic of V. sandwicensis, one of the inland species.
The seeds of Vigna lutea float in sea-water unharmed for months, and
they are to be found in the stranded drift of the Hawaiian and Fijian
beaches, and floating in the drift of the Fijian rivers. I was unable to
obtain the mature seeds of the inland species, and it has therefore yet
to be determined whether they follow the rule in the loss of buoyancy.
It may be added that a plant of Vigna lutea raised in Hawaii from seed
displayed some small tubers of the size of a pea on its roots.

The case for Premna is stated in Note 32. In this genus, as with Vigna,
the final test of experiment is needed; but the data at my disposal
point to the probability that an inland species has here been derived
from a littoral plant.

The summary of this chapter is given at the end of Chapter XVI.

CHAPTER XV

THE RELATION BETWEEN LITTORAL AND INLAND PLANTS
(_continued_)

Inland species of a genus developed from littoral species originally
brought by the currents but no longer existing in the
group.—Illustrated by the Leguminous genera, Erythrina, Canavalia,
Mezoneuron, and Sophora, and by the Apocynaceous genus, Ochrosia.—The
Hawaiian difficulty.

SECTION III

HERE we have three genera of the Leguminosæ, namely, Erythrina,
Canavalia, and Sophora, and one Apocynaceous genus, Ochrosia, in which
it is considered that inland species have been probably developed from
littoral species no longer found in the group. In this case the shore
species, possessing buoyant seeds or fruits that are known to be
dispersed by the currents, is absent from the particular group in which
the inland species occurs; and since the last-named displays no capacity
for distribution by currents, or seemingly by birds, we are driven to
infer that it was originally derived from a coast species, brought by
the currents, that has since disappeared.

Hawaii is the only region concerned here; and these four genera may be
said to well illustrate the particular “Hawaiian difficulty.” If this
explanation of the origin of the inland species is legitimate, then it
offers us a mode of explaining still more perplexing cases in the
Hawaiian flora, such as those relating to the endemic species of
Mezoneuron (Leguminosæ) and to Hillebrand’s Vallesia (Apocynaceæ), where
there is apparently no littoral species known from any region.

Dealing with the three Leguminous genera, it is at first to be remarked
that the great floating powers of the seeds of the littoral species are
in all three cases to be attributed to the buoyant kernel; whilst on
account of the non-buoyancy of the kernel the seeds of all the inland
species possess no floating power. Some very interesting points are
raised in each of the three genera, and I will first deal with the genus
Erythrina.

ERYTHRINA.

If we look over the Pacific islands in search of a critical locality for
the investigation of the genetic relation between the littoral and coast
species of Erythrina, we discover it, as far as I can gather, only in
one group. In Fiji, Tonga, and Samoa we find only the littoral species;
in Hawaii there is only an inland species; whilst in Tahiti occur both
the littoral and the inland species—E. indica, the wide-ranging
shore-tree of the South Pacific, and E. monosperma, the inland tree of
Hawaii—the last found nowhere else in Polynesia, and confined to the
Pacific. In Tahiti there are no other species, and it is between these
two species that the connection, if it exists, is to be sought. (Further
details relating to the genus are given in Note 53. In this place only
the facts bearing on the argument will be discussed.)

The buoyant seeds of Erythrina indica are well known to be dispersed by
the currents; whilst those of E. monosperma, as obtained from Hawaii,
have no floating power and sink at once, or in a day or so, even after
drying for two years. In Tahiti the first-named species is a
characteristic plant of the beach, whilst the last grows there in the
valleys and on the mountains up to elevations of 700 to 800 metres. We
have now to inquire whether there is any decided affinity between the
two species, and whether the divergent characters of the inland species
can be connected with its station. With regard to the first query we may
quote in reply the observation of Drake del Castillo, that as concerning
the foliage and the inflorescence E. monosperma is very nearly related
to E. indica, differing only from it in the more hairy calyx, in the
more permanently tomentose and much shorter pod, and in the paucity of
seeds (one or two in number).

We will now see whether it is possible to connect these differences in
character with differences of station. Neither Nadeaud nor Drake del
Castillo give precise descriptions of the station of Erythrina
monosperma in Tahiti; but Nadeaud and Lepine remark that it grows on
precipices as well as in the valleys on the north or dry side of the
island; and we may infer that it affects exposed dry rocky stations. In
Hawaii, according to Hillebrand, it is found on the dry rocky hills and
plains of all the islands up to 1,000 feet. I was particularly
interested in this tree whilst in the group, and found it in the large
islands of Maui and Hawaii thriving in rocky arid districts of little
rainfall, accompanied by Cactus opuntia, Ricinus communis, and
Cæsalpinia bonducella. It is often to be observed on scantily vegetated
lava-flows, a solitary tree growing here and there out of a crack in the
old lava, or it may dot the rocky slopes of some barren declivity. I
found it in the dry gulches behind Lahaina at elevations of 800 to 1,200
feet above the sea, growing amongst huge blocks of stone in clumps of
ten or twelve trees. When one contrasts the inland station of E.
monosperma with that of E. indica on the beach where the atmosphere is
more humid and the conditions more suited for plant-growth, it appears
probable that the differences between these two species may be largely
connected with station, especially as regards hairiness and the
diminished size of the pods.

Assuming, therefore, that Erythrina monosperma is but the inland form of
E. indica and that the differences between the two species are mainly an
affair of station, we have next to account for the occurrence of the
inland species in Hawaii without the littoral species. The agency of
currents in explanation of the existence of E. monosperma in Hawaii is
at once excluded, since the pods dehisce on the tree, and the seeds, as
already remarked, have no floating power. Nor does it seem likely that
beans half an inch (13 mm.) long could be transported unharmed in a
bird’s stomach over the two thousand miles of sea that intervene between
Tahiti and Hawaii. Yet one cannot doubt that the pyrenes and “stones” of
genera like Coprosma, Nertera, Cyathodes, and Osteomeles have been
carried by frugivorous birds to Hawaii. But a bean is somewhat different
from the crustaceous pyrene of Coprosma or the hard “stone” of
Cyathodes; and although, as indicated by the occurrence of an endemic
species of Erythrina in Fernando Noronha, birds may carry large beans
unharmed over a couple of hundred miles of sea, one hesitates to
conclude that they could effect this when the tract of ocean to be
traversed is ten times as great. There are again reasons for believing
that the seeds of Erythrina monosperma are particularly ill-suited for
dispersal by birds, since, notwithstanding their hardness, they soon
absorb water through the micropylar opening; and they germinated so
readily in my experiments that the digestive juices in a bird’s stomach
would probably soon find access and destroy the kernel. It is, however,
known from the observations of the Messrs. Layard in New Caledonia that
a small crow and different species of parrots feed on the seeds of
Erythrina, and they may aid in the local dispersal (_Ibis_, vol. 6,
1882).

To admit man’s agency in carrying to Hawaii the seeds of a tree which is
only useful in supplying him with light wood for his outriggers and his
fishing-net floats would compel us to place in the same category a great
number of plants in some way useful to him which are recognised as
indigenous. The Polynesian ransacks the vegetable world for his wants,
and carries with him in his migrations only his food-plants and the
seeds of his sacred trees.

There remains then the possibility that the parent species, Erythrina
indica, was once in Hawaii but has since disappeared. In order to
establish this, it will be requisite to show not only that the
extinction of a shore-plant is probable, but also to explain why the new
species has selected such arid inland localities for its stations, to
account for the loss of buoyancy of the seeds, and, if possible, to give
an instance of the production of a new species of Erythrina in a small
isolated oceanic island.

A study of the special circumstances of Hawaii leads one to conclude
that a shore-tree may become extinct in one of two ways. It may be
exterminated by insect pests, or it may be forced inland through
unsuitable coast-conditions and there be lost in the resulting new
species. One characteristic shore-tree, Cordia subcordata, has indeed
been almost exterminated by insects, and even Erythrina monosperma is
now from the same cause on its road to extinction (see Note 53); but
there is no indication of their leaving modified descendants behind that
are pest-proof. The most probable view then is that the littoral tree,
having been driven inland through the unsuitability of the
coast-conditions, such as lack of beaches or want of moisture, has there
become modified. This is what has really happened, as I have shown, with
Cæsalpinia bonducella in Hawaii. As indicated in Chapter XVII., this
characteristic beach-plant has here been driven off the beach. There
would thus be no difficulty in assigning a reason why a littoral tree
like Erythrina indica should select arid localities when it extends
inland, since, as is pointed out in Chapter IV. and in other parts of
this work, the plants of the beach and of the arid inland district
possess the same xerophilous habit.

With regard to the loss of buoyancy of the seeds in the case of
Erythrina monosperma, it may be remarked that this is precisely what has
happened with the seeds of Cæsalpinia bonducella, its usual associate on
the old lava-wastes in Hawaii, and with an inland species of Cæsalpinia
in Fiji. It is argued that the same thing has occurred with the inland
Hawaiian species of Canavalia and Sophora, as shown in later pages of
this chapter. It has certainly happened with the inland form of Afzelia
bijuga in Fiji, a tree dealt with in Chapter XVII. These are all
Leguminous genera; and in all of them, with the exception of Cæsalpinia,
where the floating power arises from a central cavity in the seed, the
seeds of the littoral species possess, like Erythrina indica, buoyant
kernels. Whilst most littoral plants with buoyant seeds or fruits retain
the floating capacity of the seed or fruit when they extend inland, the
Leguminosæ often offer exceptions to the rule.

That inland endemic species of Erythrina can be developed in isolated
islands is illustrated by the existence in Fernando Noronha, some two
hundred miles from the coast of Brazil, of a peculiar species, E.
aurantiaca, described by Mr. Ridley. Here also is found an inland
species of Guettarda peculiar to the locality; but in neither genus does
the littoral species occur.

Many difficulties will yet have to be explained before it can be finally
established that Erythrina monosperma has been derived from E. indica or
some similar shore species that was originally dispersed by the
currents; but we are almost driven towards such a view, since it is hard
to believe that the beans were carried to Hawaii by birds over some two
thousand miles of sea. Observers in other regions where littoral and
inland species of the genus occur may perhaps devote their attention to
the relation between the two; and if they are able to supplement
observation and experiment by a microscopical investigation, some
interesting results would be obtained. For instance, I would suggest
that in Queensland a thorough examination of the littoral E. indica and
the inland E. vespertilio might be undertaken; or perhaps there may be
some other littoral form.

With the two other Leguminous genera, Canavalia and Sophora, to be
immediately discussed, we have for the most part the same questions
raised. Both possess wide-ranging current-dispersed littoral species in
other parts of the Pacific, but only endemic inland species with
non-buoyant seeds in Hawaii. The pivot of the discussion will be here
also the impracticability of these inland species ever having reached
the Hawaiian Islands through the agency of the currents, and the great
difficulty in believing that their beans were carried unharmed by birds
over half the breadth of the Pacific Ocean. If we reject alike the
current, the bird, and the parentage of a lost littoral species, we must
fall back on the continental hypothesis, against which in the case of
Hawaii the evidence is overwhelming.

CANAVALIA.

This genus is represented in the tropical islands of the South Pacific
from Fiji to Tahiti by three littoral species, none of which have been
found in Hawaii, where only an endemic inland species exists. Reference
will alone be made here to such facts as bear on the probable history of
the mysterious Hawaiian species, additional particulars being given in
Note 54. The littoral species, Canavalia obtusifolia (D.C.), C. sericea
(Gray), and C. ensiformis (D.C.), have buoyant seeds and are dispersed
by the currents; whilst the inland Hawaiian species, C. galeata (Gaud.),
a forest climber peculiar to that group, has non-buoyant seeds. We thus
have repeated the problem of Erythrina monosperma. The absence of the
littoral species from Hawaii can scarcely be attributed to the failure
of the currents, since Ipomœa pes capræ, which accompanies C.
obtusifolia as a beach-creeper all round the tropical globe, is present
on the Hawaiian beaches. Nor can it arise from lack of floating-power on
the part of the seeds, since experiment indicates that the seeds of C.
obtusifolia will float for months unharmed in sea-water. Nor can it be
ascribed to climatic conditions, since this tropical shore species
extends into cooler latitudes than those of the Hawaiian Islands, being
found in the Kermadec Group and in the Bermudas, which are subtropical
both in position and as regards much of their vegetation. The reason
perhaps we may never learn from the plants themselves, though it may be
possible to obtain some light on the problem from outside sources.

Canavalia galeata differs much in its habits, as well as in some of its
characters, from the existing littoral species of regions outside the
Hawaiian Group. It is a stout climber ascending the forest trees to a
considerable height, though, as is indicated in Note 54, the shore
species sometimes display a tendency in the same direction. It is
described by Hillebrand as occurring “on all islands, in forests up to
2,000 feet.” Like those of the inland species of Erythrina (E.
monosperma), its seeds sink in sea-water even after being kept for four
years, nor could the pods be utilised for dispersal by the currents,
since they float, when unopened, only for four or five days. Here also,
as with Erythrina, the seeds of the inland species no longer possess the
buoyant kernels to which the floating capacity of the seeds of the coast
species is due. Though we have to exclude the currents, we can scarcely
in its case appeal to bird-agency when we wish to account for the
transportal of the original seeds to Hawaii, as that would imply that
birds can carry beans nearly an inch, or 2 to 2.5 centimetres, in length
unharmed in their stomachs over a tract of ocean some 1,500 or 2,000
miles across. We should have to learn much that is unexpected of the
modes of dispersal of the Leguminosæ before we could accept such an
hypothesis.

Canavalia galeata indeed presents to the student of dispersal one of the
enigmas of the Hawaiian flora; and it should be noted that the mystery
of its distribution is concerned not only with the means of transportal
of the seeds of the original species to the group, but also with its
present dispersal among the islands. It is, however, suggestive that Dr.
Hillebrand mentions two varieties, one of them found on Kauai, with
somewhat smaller seeds; so that some inter-island differentiation is
evidently in progress. No attempt is made here to connect this inland
species directly with the absent beach-plants. That is a matter for the
systematist; but we are not tied down to existing shore-plants in
finding an ancestor, since the common parent of the littoral and inland
species may have been a shore-plant dispersed by the currents.

MEZONEURON.

Another closely parallel instance, offering, from the standpoint of
dispersal, the same difficulties presented by Canavalia galeata, is to
be found in Mezoneuron kauaiensis (Hillebr.), a tall inland shrub also
peculiar to the group and belonging alike to the Leguminosæ. The
difficulties are so nearly identical that the same explanation will have
to cover both; but it is significant that with Mezoneuron there is no
littoral species to which we can appeal to extricate us from the
difficulty. Yet the genus is related to Cæsalpinia, and the species was
first described by Mann as C. kauaiensis, so that it may have once
possessed a littoral species that has ceased to exist as such. When we
come to discuss Cæsalpinia and Afzelia (Chapter XVII.) we shall obtain
from those genera many suggestions as to the probable past of Canavalia
galeata and Mezoneuron kauaiensis, two of the greatest riddles presented
by the Leguminosæ of Hawaii.

The flat seeds of this species of Mezoneuron measure about an inch (2·5
cm.), and seem most unsuitable for dispersal by birds over a wide extent
of ocean. Nor can we appeal to the currents, since my experiments in
Hawaii show that the seeds have no buoyancy and that the pods only float
for a week in sea-water. Dr. Hillebrand records this shrub from Kauai,
Oahu, and Maui; I found it also on the lower slopes of Hualalai in
Hawaii and therefore the same question of inter-island dispersal here
presents itself that was connected with Canavalia galeata, since we have
also to explain the transport of the seeds between islands 70 to 150
miles apart. The critical point in the history of these two enigmatic
inland plants of the Hawaiian Islands was doubtless concerned with the
loss of buoyancy of the seeds of the original littoral plant. It will
subsequently be shown that this is what is now in actual operation with
Cæsalpinia and Afzelia in different parts of the Pacific.

SOPHORA.

In this genus, as in Erythrina and Canavalia, we have a littoral
species, Sophora tomentosa, that ranges over the tropical beaches of the
globe, including most of the islands of the Pacific, but does not occur
in Hawaii, where the genus is represented by an endemic inland species,
S. chrysophylla. Here also we find the shore-species with seeds capable
of floating for months on account of their buoyant kernels, and the
inland species with seeds that sink even after years of drying (see Note
56). Unless other inland species of Sophora have recently been described
from the tropical Pacific, the Hawaiian species is the only one of its
kind known from this region.

But the problem wears a different aspect in the case of this genus,
since the endemic inland species of Hawaii is a tree of the mountains
where a temperate climate prevails, whilst Sophora tomentosa is a shrub
of the tropical beach that only at times extends into subtropical
latitudes. The Mamani tree, as the Hawaiians name S. chrysophylla,
extends up to 9,000 or 10,000 feet above the sea, forming, with Myoporum
sandwicense and one or two other trees and shrubs, the highest belt of
the forest in the larger islands. It is in the open woodland between
6,000 and 7,000 feet that it is most at home, and here it attains a
height of 20 to 30 feet. It descends in places to as low as 2,000 feet
above sea-level; but here is living under uncongenial conditions, and,
like Myoporum sandwicense, becomes dwarfed and shrubby. The climatic
conditions under which S. chrysophylla thrives in the Hawaiian mountains
are therefore those of the temperate zone. From the data given in
Chapter XIX., the mean annual temperature at an elevation of 6,000 to
7,000 feet would probably be about 55°, the average temperature of New
Zealand.

We must therefore look to the temperate and not to the tropical zone for
the home of the parent species of Sophora chrysophylla; and if it was
originally derived from a shore-plant dispersed by the currents, the
widespread S. tomentosa could scarcely have been the species concerned.
But this strand-plant is disqualified for another potent reason, since
it belongs to a different section of the genus. Whilst S. tomentosa
belongs to the section possessing smooth pods, S. chrysophylla is
referred to the section Edwardsia having four-winged pods, which
comprises about ten species found in Chile and Peru, Hawaii, New
Zealand, Further India, and the Isle of Bourbon. What strange principle
in distribution, we may fitly ask, has linked together in this odd
fashion the continents of the Old and New World and the islands of the
Indian and Pacific oceans?

Yet, discredited as Sophora tomentosa is as a possible parent of the
Hawaiian mountain species, it may yet afford us a clue. It is
significant that the distribution of this wide-ranging beach-shrub in
the tropics of the southern hemisphere is almost coterminous with that
of Sophora tetraptera, a species widely spread in the south temperate
zone from Chile to New Zealand and extending towards the tropics as far
as Juan Fernandez in lat. 33° S. and to Easter Island in lat. 27° S.
Though not strictly a beach-plant, S. tetraptera is a plant of the
sea-border; and it is remarkable, but not surprising, how in New
Zealand, one of its principal homes, its behaviour in respect of its
vertical distribution presents a great contrast to that of S.
chrysophylla in the tropical latitudes of Hawaii. We have seen that, in
Hawaii, S. chrysophylla, which thrives as a tree 20 to 30 feet high in
the mountains, becomes shrubby when it descends to the lower levels. In
New Zealand, S. tetraptera is, as we learn from Kirk, a prostrate shrub
in the mountains, whilst in the lower elevations towards the sea it
becomes a tree 30 and even 50 feet in height. It can scarcely be doubted
that, if we exchanged the habitats of these Hawaiian and New Zealand
species, each would to a great extent take up the other’s station and
the other’s habit.

The whole problem of the dispersal of Sophora was brought immediately to
my notice at Corral, in latitude 40° S. on the coast of Chile. Here a
small tree of the section Edwardsia was growing in fruit on the lower
slopes of the hills, becoming bushy when descending to the beach.
Specimens of its four-winged pods have been identified at the Kew Museum
as those of Sophora tetraptera; and, as far as the pod is concerned, I
cannot distinguish between my specimens of the Hawaiian S. chrysophylla
and the Chilian species. Subsequently I found the buoyant seeds of the
same plant amongst the stranded beach-drift at Bahia San Vincente,
nearly 200 miles further north. This led to my experimenting on the
capacity of the plant for dispersal by the currents, and as a result it
was ascertained (see Note 56) that whilst, as in the case of S.
chrysophylla, the pods floated only one or two weeks, the seeds on
account of their buoyant kernels floated for several months in
sea-water, retaining their power of germination. The Chilian plant thus
differs significantly in its capacity for dispersal by currents from the
Hawaiian species, the seeds of which sink in sea-water even after years
of drying.

The Mamani tree in Hawaii had always been an object of great interest to
me. I was attracted by the mystery surrounding its origin and had long
suspected that the clue was to be found in the non-buoyancy of its seeds
and in the absence of a littoral species of the genus. When in Fiji it
was to the littoral Sophora tomentosa that I looked in vain for a
solution of the riddle, and seven years afterwards on the coast of Chile
a solution of this enigma of the Hawaiian mountains presented itself in
the form of an argument somewhat in this shape.

On account of the elevated station of the Mamani tree (S. chrysophylla)
in Hawaii it is to be inferred that the original species was a plant of
the temperate regions or of the uplands of some tropical mountains. If
it has had its origin in some shore-plant dispersed by the currents,
that species can only now be found on the coasts of extra-tropical
regions. Such a maritime plant had buoyant seeds; and plants of this
type are presented by Sophora tetraptera and its allied species that are
at home in the cool latitudes of the southern hemisphere, as in Chile
and New Zealand. No difficulty, as I argued, could be connected with the
loss of buoyancy of the seeds of the Hawaiian mountain species, since it
follows the general principle (laid down in Chapter II.) that in the
same genus coast species have buoyant seeds or fruits, and inland
species those that sink; and in support of this view it was recalled
that this is what happens to the seeds of Cæsalpinia bonducella and
Afzelia bijuga when the plants extend inland in the Pacific islands. It
was held, in short, that the original form of Sophora chrysophylla in
Hawaii was a coast plant with buoyant seeds, and therefore indebted for
its presence to the currents. Hailing from an extra-tropical region, it
abandoned the beach and found suitable conditions of existence in the
mountains, where it underwent specific differentiation. Such was the
explanation that presented itself to me on a Chilian beach.

The first objection that offers itself against this view is that Sophora
chrysophylla is one of several species characterising the antarctic
element of the mountain flora of Hawaii, and that many of these plants,
such as those of the genera Astelia, Coprosma, Gunnera, Myoporum, &c.,
could only have reached these islands through the agency of frugivorous
birds (see Chapter XXIII.). There is, therefore, something to be said
for this mode of dispersal; but though one can understand how hard seeds
and the “stones” and crustaceous pyrenes of fleshy fruits might be
transported unharmed in a bird’s stomach half-way across the Pacific
Ocean to the distant group of Hawaii, it is difficult to understand how
Leguminous seeds, except in such cases as Tephrosia piscatoria, could be
ejected unharmed by a bird after an ocean passage of some 1,500 or 2,000
miles. Yet evidence pointing to such a possibility is not lacking. It
was pointed out by W. O. Focke (_Nat. schaft. Ver. zur Bremen_,
Abhandl., Band 5, 1876) that for many Leguminosæ we are driven to the
agency of birds in order to explain their dispersal. In this connection
he mentions the case of a pigeon killed by some beast of prey that he
found in his garden in the early winter. In the following spring he
noticed numerous seedlings of Vicia faba sprouting up from amongst the
feathers that alone remained of the bird. In this observation he
detected the normal method of the dispersal of the Leguminosæ by birds,
the seeds not being ejected by the bird but being set free by its death.
It is well known that Darwin had this idea in his mind when he conducted
his experiments on the dispersal of seeds; and reference may here be
made to one that is recorded in _More Letters of Charles Darwin_ (i.,
436). Out of a number of seeds left in the stomach of an eagle for
eighteen hours, the majority were killed; but amongst the few that
germinated afterwards was a seed of clover (Trifolium). If such a bird
had carried a Sophora seed to Hawaii, this would have involved a
continuous flight of, on the average, 100 miles per hour for a period of
fifteen to twenty hours. This would just come within the limitations
laid down by Gätke as regards length and velocity of flight—a subject
discussed in Chapter XXXIII.

We will now turn to the Sophora seeds themselves for evidence of their
capacity of surviving the perils of such a journey. The seeds of Sophora
chrysophylla, which are about a quarter of an inch (6 to 7 mm.) in
length, possess unusually hard coverings for the order, and in that
respect appear fitted for dispersal by animals. Indeed, in the large
island of Hawaii wild pigs and sheep feed on the pods, and no doubt aid
in the distribution of the plant over the island through the germination
of ejected undigested seeds. But since the species is found on most of
the larger islands, it is apparent that to birds we must look for the
explanation of its inter-island dispersal. Mr. Wilson, in his _Aves
Hawaienses_, remarks that one of the Hawaiian finches (Loxioides) feeds
on the seeds of this tree, which probably, he adds, also serve as the
food of Chloridops kona, another big finch; and it is to be inferred
from the observations of Mr. Perkins, quoted by Mr. Evans in his book on
Birds, that the Drepanididæ, a family peculiar to Hawaii, are in the
habit of splitting the pods of trees like Acacia koa and Sophora
chrysophylla to obtain the seeds. It would, however, seem that the
agency of birds confined to these islands does not carry us very far
when we wish to explain the original transport of the seeds over a
breadth of ocean of some 1,500 miles and more. Yet we know that this
must have happened with some of the Hawaiian plants, such as Osteomeles
anthyllidifolia and Nertera depressa, that are not confined to these
islands and possess fruits that would attract frugivorous birds. But
whether it has occurred with the dry beans of the Hawaiian species of
Sophora is another matter.

On the whole I am inclined to the view, bearing in mind the general
indications of the Leguminosæ in the Pacific, that S. chrysophylla
originally reached Hawaii as a littoral plant through the agency of the
currents. Many points still need investigation; but it may be pointed
out that South America probably received Sophora tetraptera from New
Zealand by the West Wind Drift Current.

OCHROSIA (Apocyneæ).

This genus seems to offer the strongest testimony in support of the
derivation of an inland species from a strand-plant. The drupes are so
large, the minimum size of the “stone” being 1-1/2 or 2 inches (37 to 50
mm.), and so dry and unattractive for birds, that any other agency but
that of the currents appears to be out of the question. Indeed their dry
appearance would suggest to my readers that only birds of the habits of
the ostrich would venture on such a diet. It is, however, worth noting
that whilst in the Keeling Islands I learned that a cassowary that had
been kept on the atoll was a very efficient distributor of the seeds of
Ochrosia parviflora, scattering the undigested stones everywhere, and
causing the young trees to become so numerous that they had to be
destroyed. A similar habit of the cassowary in the Aru Islands is
recorded by Beccari, where the dry fruits of a palm, 2-1/2 inches
across, are swallowed by these birds and the seeds dispersed.
Cassowaries are active agents in dissemination, for they swallow every
kind of pulpy fruit, and convey them long distances undigested; they are
also excellent swimmers and traverse considerable expanses of water
(Beccari, quoted in _Chall. Bot._, iv., 297, 313).

Modern ornithologists would probably not object to our appealing to the
former volant habits of the cassowary and its allies even across a wide
tract of sea; but, excepting in New Zealand and its vicinity, such birds
are not at our disposal in the island groups of the open Pacific. There
is a possibility that the extinct Columbæ and other exterminated birds
of the Mascarene Islands might account for some anomalies in their
floras; and in Chapter XVI. reference is made to the fact that these
islands possess more endemic species of Pandanus than any other oceanic
groups, a genus possessing drupes that in the case of inland species
seem unfit for any mode of dispersal with which we are familiar. In the
islands of the tropical Pacific, however, it is not possible to find
such a way out of the difficulty, since, as shown in Chapter XXXIII.,
the birds are lacking.

The genus, according to the _Index Kewensis_, includes about ten species
distributed over the islands of the Indian Ocean, and found also in
Malaya, Australia, and throughout the Pacific. It is essentially an
insular genus, and two at least of the species are wide-ranging littoral
trees, one, Ochrosia borbonica, mainly distributed over the islands of
the Indian Ocean and of Malaya, and the other, O. parviflora, chiefly of
the islands of the Pacific. It will be out of place to deal here in any
detail with this interesting genus, and my remarks will be confined to
such matters as concern the origin of the inland species of the Hawaiian
Islands, species that are peculiar to that group. Some confusion has
prevailed amongst different authors in the determination of the limits
of the various species, and to avoid this I have mainly followed
Schumann in his monograph on the order (Engler’s _Naturl. Pflanz. Fam._,
Theil 4, Abth. 2, 1895), as indicated in Note 57.

Besides the littoral species Ochrosia parviflora, Hensl., that ranges
over most of the archipelagoes of the Pacific from the Solomon Islands
to Tahiti, but is not found in Hawaii, we have in the Pacific, O.
elliptica, Lab., of New Caledonia and Fiji; another species of New
Guinea and the Ladrones; and one or two inland species of Hawaii.
Ochrosia parviflora was familiar to me on Keeling Atoll, in the coral
islets of the Solomon Group, and on the islets and coasts of certain
parts of Fiji. Its fruits, which are dispersed by the currents, were
found amongst the stranded drift of the Keeling and Fijian beaches.
Although usually a coast-tree in Fiji, it came under my notice in one
locality growing inland; and it is a very suggestive circumstance in
connection with the inland species of Hawaii, that in Tahiti this tree
is only described by the French botanists as growing in the mountains at
elevations of 700 to 800 metres above the sea, it having for some reason
abandoned the beach. The process which we thus see in operation in
Tahiti is completed in Hawaii, and we there find a peculiar inland
species far away in the interior of the islands which is placed by
Schumann in the same section of the genus with the littoral O.
parviflora, that is not, however, found in the group. It may be remarked
that Gray describes only one species from Hawaii, O. sandwicensis, but
Schumann makes two species of it—one, O. compta, Sch., peculiar to the
group and referred to the same section as O. parviflora; the other, the
original species of Gray, which he considers as probably a variety of O.
borbonica. These determinations of the German botanist, who had no
theory to serve, are especially interesting. It is with the littoral
trees now missing from the Hawaiian beaches that he compares the inland
species of the group, trees now chiefly characteristic the one of the
Indian Ocean and the other of the South Pacific; and we can scarcely
doubt that originally one littoral tree ranged over both oceans.

Hillebrand describes Ochrosia sandwicensis of Gray as a shrub or small
tree, 6 to 12 feet in height, growing in the open woods of the lower and
middle regions on all the islands. Its dry ellipsoid fruit is two inches
(5 cm.) long, and possesses a thin suberose covering on one side and a
very thick woody endocarp, one-quarter to one-third of an inch (6 to 8
mm.) in depth. The other species which he characterises as a variety is
not so generally distributed in the group. We have to explain not only
how the original species reached the group, but also how they have been
distributed over the islands. The currents could scarcely have
transported the fruits as we now see them. Those of O. sandwicensis have
only a trace of a buoyant covering, and, judging from some fruits that I
examined, they could possess little or no floating power. Even the most
enthusiastic advocate of dispersal by birds must pause here; and there
remains the view, supported by evidence of a striking character, that
the inland Hawaiian species are derived from littoral species that,
having been originally brought by the currents, like O. parviflora in
Fiji, abandoned the beach and took to the mountains, where they have
become differentiated.

It is probable that the lesson of Ochrosia in Hawaii can be applied to
one or two of the other Hawaiian “difficulties,” and that plants that
now set at defiance all the attempts of the student of dispersal to
explain their occurrence in this group may have commenced their
existence in these islands as littoral species brought originally by the
currents and afterwards driven off the beach. One of the greatest
enigmas of the Hawaiian flora is connected with another small
Apocynaceous tree peculiar to the group and described by Hillebrand as
Vallesia macrocarpa and by other Hawaiian botanists as a species of
Ochrosia. Schumann, however, places it in a new genus, Pteralyxia, near
to Alyxia, a genus already in the islands. However this may be, its dry
drupaceous fruits two inches (5 cm.) in length, and its pyrenes almost
as long, could never have been transported as such by the birds of our
own time; and if they could have been carried in the stomach of a bird
given to the dietetic humours of the cassowary, such birds in their
trans-oceanic passages would have left some trace behind in the groups
of the mid-Pacific. In our perplexity we read again the lesson of
Ochrosia.

_Summary of Chapter_ (see end of Chapter XVI.).

CHAPTER XVI

THE RELATION BETWEEN LITTORAL AND INLAND PLANTS
(_continued_)

SECTION IV

HERE we deal with two genera, Pandanus and Barringtonia, where inland
endemic species occur in the same group with the wide-ranging coast
species, but possess fruits concerning which it is either difficult or
almost impossible to suggest a mode of dispersal by existing agencies.
This section is especially concerned with Fiji, and represents the
peculiar “Fijian difficulty” that is illustrated by other genera as—for
instance, the Coniferous genus Dammara—which are not in any sense
littoral. Further investigation is, however, requisite in the case of
Barringtonia, and to a less degree with Pandanus; and I can only here
point to the general indications of the data at my disposal. We have in
these genera to assume either that the inland species are derived from
the coast species, or that the seeds were brought by one of the extinct
birds of the Western Pacific, by a megapode or by one of the Columbæ, or
by some Struthious bird like the moa or the cassowary, or, if these two
assumptions fail, that there has been a continental connection through
the islands to the westward with the mainland beyond.

PANDANUS.

I take this genus first because the recent monograph on the Pandanaceæ
by Dr. Warburg (Engler’s _Das Pflanzenreich_, 1900) enables me to tread
on relatively safe ground in making my deductions. The three genera of
the order, Freycinetia, Pandanus, and Sararanga, each tell their own
story; and in each and all of them I have taken an especial interest
from the standpoint of their dispersal. Freycinetia is fully discussed
in Chapter XXV., and presents no difficulties respecting its dispersal.
In the discovery of Sararanga the author has had a share. It was first
established by Mr. Hemsley from specimens sent by me to Kew in 1885; and
it has received from the botanist the name given to it by the natives of
the islands of Bougainville Straits in the Solomon Group, where I first
collected it. It contains only one species and was also discovered by
Dr. Beccari, the celebrated Italian botanist, in Jobie Island, New
Guinea. From the other two genera of the order, Pandanus and
Freycinetia, it stands quite apart; and it apparently presents us with a
relic of some ancient flora on the western borders of the Pacific. Its
fleshy drupes (one-half to three-quarters of an inch in size) inclosing
several small osseous pyrenes seem suited for dispersal by birds; and it
is not at first sight easy to understand why its distribution should be
so limited, unless this is connected with its diœcious habit (see
Guppy’s _Solomon Islands_, p. 302; _Journ. Linn. Soc. Bot._ vol. xxx.;
and Warburg’s monograph).

It is, however, with the genus Pandanus that we are here especially
concerned. If the advocate of the previous continental connections of
Fiji and the groups around were to look for evidence in support of his
views, he apparently could not do better than take this genus. Whilst P.
odoratissimus, the littoral species of tropical Asia and Malaya, is
found on the coasts of almost all the Pacific islands from Fiji to
Tahiti and northward to Hawaii, it is only in the archipelagoes of the
Western Pacific, namely, in Fiji and Samoa, that inland endemic species
have been found. (Such species occur also in the more western islands
not dealt with here—New Caledonia, Solomon Islands, &c.) Not even in
Hawaii, with all its botanical evidence of antiquity, has an inland
endemic species been found, although the coast species extends miles
inland, and for nearly 2,000 feet up the mountain slopes. When, however,
we turn to Fiji and Samoa, we find in each group two endemic inland
species. To endeavour to connect the inland species of Fiji and Samoa
with the widespread littoral Pandanus odoratissimus, that owes its
dispersal largely to the currents, is out of the question, at least for
the student of plant-dispersal, since they belong to different sections
of the genus, and in their characters are often far removed (see Note
58).

As regards the agency of birds, it is of course possible that
fruit-pigeons that can disperse the “stones” of Canarium and Elæocarpus
could transport the smaller drupes of Pandanus to oceanic islands like
the Fijis, Samoa, and the Mascarene Islands; and in Note 58 reference is
made to the size of the drupes of the endemic species of Pandanus in
those groups. But my difficulty is that I have not come upon any record
of birds eating these fruits; and I should imagine that amongst living
birds only those like the cassowary and its kin would prefer such a kind
of diet; whilst the only pigeon that could have ever attempted it must
have been able to swallow pebbles like the dodo. It is remarkable that
the Mascarene Islands, the home of the extinct Columbæ, possess more
endemic species of Pandanus than any other groups.

Dr. Warburg points out that, with the exception of some three or four
species dispersed by the currents (P. dubius, P. leram, P. polycephalus,
P. odoratissimus), almost all the species (156 in number) are very
restricted in their areas. When we look at his table of the distribution
of the genus we notice that, excepting the islands of the Hawaiian and
Tahitian regions, nearly all the elevated or mountainous islands of the
tropical and subtropical latitudes of the Indian and Pacific oceans have
their peculiar species, whether in the case of Mauritius, Rodriguez,
Réunion, and the Seychelles in the one ocean, or of Lord Howe Island,
New Caledonia, Fiji, and Samoa in the other. The student here hesitates
even to raise the question of present plant-dispersal in the face of
such evidence of isolation all over the area of the genus. He is almost
inclined to evade the issue and to place the matter beside that of the
dying or extinct Columbæ that have been found in some of these islands,
as in Mauritius, Rodriguez, Réunion, and Samoa.

For reasons above given in the instance of Fiji and Samoa, it would seem
futile to attempt to connect in their origin the inland with the coast
species; and it may be inferred that, excepting the few dispersed by the
currents, the species are in the main inland in their stations. Those
peculiar to Fiji, for instance, occur in the swampy forests of the lower
regions of the interior, as well as high up towards the mountain
summits. When traversing the Fijian forests I often used to speculate on
the modes of dispersal of the plants familiar to me; but the sight of a
strange Pandanus usually brought my speculations to a close. Many of the
enigmas of insular floras would be solved if we could interpret aright
the 156 species of Pandanus that are enumerated and described by Dr.
Warburg in his monograph. Observers like myself obtain little peeps into
the conditions of existence of these interesting plants; and the
travelled botanist, who becomes a systematist in his later years,
attains to a far more extensive view, yet even he can only penetrate the
mystery for a little way.

It is doubtful whether Pandanus odoratissimus, the shore-tree of the
tropical beaches of the islands of the Pacific and Indian oceans, of
Australia, Malaya, and Southern Asia, can aid us much in any one
locality, since its distribution has no doubt been often assisted by
man. Yet it is probable that the currents have played a predominant part
in its dispersal. Its fruits occur commonly in beach-drift, both in the
Indian and Pacific oceans, and are often incrusted with serpulæ,
polyzoa, and cirripedes. At certain seasons the currents bring them to
Keeling Atoll in abundance. When, however, we come to inquire why it is
that this beach species is the only representative of the genus in
Hawaii and Tahiti, we are met with the possibility of its having been
introduced by the aborigines. The tree is almost as useful to a
Polynesian as the coco-nut palm, and it has been cultivated by him in
some of the atoll-groups, as in the Marshall and in the Radack
archipelagoes. In Chapter VII. good reasons are advanced for regarding
it as an aboriginal introduction into Hawaii. When, therefore, we learn
that in the group just named it extends from the sea-coast to nearly
2,000 feet above the sea, that in Samoa it may at times be found at a
similar elevation though usually restricted to the sea-border, and that
in the same way in Tahiti and in Fiji it may leave the coast-region and
extend into the heart of the islands, we are not inclined to look for
any marked differentiation in its character. This indeed appears to be
the case. Numerous varieties in different regions are referred to by Dr.
Warburg; but the only important one in the Pacific islands here
mentioned is a cultivated form from the Marshall Group. A variety from
Hawaii is distinguished chiefly by the smaller size of its drupes.

Assuming, therefore, that the inland species are as a rule not derived
from littoral species originally brought by the currents, and that no
birds of our own time are in the habit of carrying the drupes of
Pandanus to oceanic islands, in order to explain the distribution of
such species we have to choose between the possibility of the agency of
extinct Columbæ and birds similar in their habits and the alternative of
a continental connection. Dr. Warburg, who says but little of the mode
of dispersal of Pandanus drupes, regards the genus as having now two
centres, one in the East African islands (Madagascar, the Mascarenes,
and the Seychelles), and the other in Papuasia (New Guinea, extending
doubtless to New Caledonia). My readers will recall to their minds that
zoologists have at times felt bound to postulate a continent in both the
centres of the genus Pandanus. There is the well-known Lemuria of the
Indian Ocean, and then we have in the Western Pacific Forbes’ Antipodea
and Hedley’s Melanesian Plateau.

Before, however, we accept the indications of the distribution of
Pandanus as favouring a continental hypothesis for either area it is
essential to exclude the agency of the extinct Aves. In this connection
it is of prime importance to notice that the Mascarene Islands are
remarkable, when contrasted with all other oceanic islands, not only for
the predominance of peculiar species of Pandanus, but also as having
been the home of extinct Columbæ like the dodo and the solitaire. The
dodo’s habit of swallowing pebbles of the size of a nutmeg
(_Encyclopædia Britannica_, vii., 322), and the solitaire’s inclination
for swallowing stones as large as a hen’s egg (_Birds_, by A. H. Evans,
p. 331), doubtless represent, as explained below, a capacity for the
dispersal of large fruits and seeds that would be regarded as
“impossible” for distribution by birds now. It is quite possible that at
some time the ancestors of these birds possessed the powers of flight
now owned by the Nicobar pigeon, in the gizzard of which, in the Solomon
Islands, I found quartz pebbles half an inch across (_Solomon Islands_,
p. 324). In the work just quoted I refer on page 325 to the observation
of Messrs. Chalmers and Gill that the Goura pigeon of New Guinea usually
carries a good-sized pebble in its gizzard. We do not, however, seem to
possess any record of extinct Columbæ in the tropical islands of the
Western Pacific. The nearly extinct Didunculus of Samoa apparently
prefers berries and soft fruits. Dr. Reinecke says that it especially
favours the berries of Cananga odorata, the seeds of which are not over
a third of an inch (8 mm.) in length.

It would appear from Mr. Hamilton’s note in the _Transactions and
Proceedings of the New Zealand Institute_ (vol. 24) that the extinct
Struthious birds of New Zealand, as in the case of the moa, carried
crop-stones sometimes as large as a pigeon’s egg. These pebbles are, of
course, swallowed by birds to enable them to crush the hard seeds, and
“stones” of fleshy fruits, on which they feed. In the Solomon Islands I
noticed that the Nicobar pigeon was able in this way to crack the seeds
of Adenanthera pavonina, which for their fracture require a blow with a
hammer. The implication is that the extinct Columbæ were able to
transport to oceanic groups seeds and “stones” which no existing pigeon
could now carry over a tract of ocean. I am inclined to extend this view
also to extinct Struthious birds, and to suppose that they were able,
like the cassowary (see page 152), to fly across tracts of sea in ages
gone by. Though such an agency would come under discussion in connection
with the floras of New Zealand and Madagascar, we have no evidence to
show that birds of this family ever reached the tropical islands of the
open Pacific.

The Megapodidæ of the Western Pacific are a family of birds that suggest
themselves in this connection. Their distribution corresponds with that
of Pandanus in the Western Pacific, excepting the littoral species; and
like Pandanus the Megapodes have “differentiated” in every group. The
limited powers of flight possessed by existing species would unfit them
for crossing wide tracts of sea; but the parent form or forms of all
these species must have been able to traverse broad tracts of ocean.
These birds subsist on fallen fruits, seeds, &c.; but I have no data
relating to them as seed-dispersers.

It is evident from the endemic character of most of the species of
Pandanus in oceanic islands that, except with a few widely-spread
littoral species, the dispersal of the genus has been for ages
suspended. Whether the explanation is to be found in the isolation and
differentiation of the extinct Columbæ of the Mascarene Islands, where
the endemic species of Pandanus are most numerous, has yet to be
established. It seems to offer the only way out of the difficulty,
unless we accept the old view concerned with the continent of Lemuria.

BARRINGTONIA.

There are two littoral species of this genus in the Pacific, B. speciosa
and B. racemosa, both widely spread over the Old World, but only the
first is generally distributed over the Polynesian region reaching east
to Ducie Island, whilst the second does not extend east of Fiji and
Samoa. With the exception of one or two inland species in Fiji and Samoa
no inland species have been recorded from the groups of the open
Pacific, and the genus is not represented at all in Hawaii. If it were
not for a suspicion that the aborigines may have aided in the
distribution of the inland species, the advocate of the previous
continental connections of the islands of the Western Pacific would
receive from their occurrence in these islands considerable support for
his views. The fruits of the inland Fijian species are large, the
smallest being three inches in length; and the agency of birds seems to
be out of the question.

The fruits of the littoral species possess dry buoyant husks that enable
them to be carried by the currents over wide tracts of ocean. Those of
the Fijian inland species display only a trace of these buoyant
coverings and the floating power is much diminished or absent
altogether. These inland species are two or three in number. One of
them, described as a new species by Seemann under the name of B. edulis,
has edible kernels and is sometimes cultivated. A species that I found
growing in the plantations of the Solomon Islanders in Bougainville
Straits may be near the Fijian tree just named (_Solomon Islands_, pp.
85, 297). Its kernels are edible; and I may add that the Solomon
Islanders cultivate other species with edible fruits. We cannot,
therefore, exclude the agency of the aborigines in the distribution of
the inland species of this genus. Horne found an undescribed species in
Fiji, which may be that which I found on the slopes of Mount Seatura in
Vanua Levu, as described in Note 50; and it is quite possible that it
was originally a cultivated tree, though not necessarily within the
memory of the later generations of the aborigines.

This retrocession to the wild state of cultivated plants and the
resulting production of apparently new species is a point on which Dr.
Beccari lays considerable stress in the English edition of his book on
the Great Forests of Borneo. He takes the case of Nephelium and other
fruit-trees and shows how in old clearings, long since abandoned, they
have undergone singular alteration in characters. For these reasons,
therefore, Barringtonia can scarcely be regarded as offering in its
inland species unequivocal evidence of a previous continental condition
of the islands of the Western Pacific. Nor, as shown in Note 50, should
we be justified in establishing a genetic connection between the inland
and coast species; but a great deal of research is needed before we can
handle the numerous interesting problems connected with the genus; and
indeed it cannot be said that the specific limits of the inland
Polynesian trees have been definitely determined, or the species
themselves diagnosed.

SECTION V.

In this section are included those genera where within the same genus
some inland species have been derived from the coast species whilst
others have been originally brought by birds. Guettarda alone belongs
here. In this genus we find, as is so frequently the case, a littoral
tree (G. speciosa) widely spread in the Old World and ranging over the
whole tropical Pacific as far east as Pitcairn and Elizabeth islands,
but absent from Hawaii. Here also as with Pandanus it is only in the
Western Pacific that we find inland endemic species so distinct in
character from the littoral tree that they may be regarded as of
independent origin.

Since, however, there is an inland form of the coast species in Tahiti
(Guettarda speciosa, var. tahitensis) which, according to Drake del
Castillo, is distinguished only by its more rounded leaves and by the
more marked pubescence of the under leaf-surfaces, we evidently have
there an inland species in process of development from the littoral
species. This inland tree is found at elevations as great as 600 metres
or almost 2,000 feet above the sea; and indeed if we follow Nadeaud the
specific differentiation is complete. However, there is no doubt raised
as to its close affinity to the beach tree; and we are almost compelled
for another reason to regard it as a derivative of the shore species,
because, as pointed out in Chapter XXVII., there are very few inland
plants in the Tahitian flora possessing fruits as large as those of
Guettarda that owe their presence in those islands to frugivorous birds.

Of the two inland species of the genus found in Fiji, G. inconspicua and
G. vitiensis, it may at once be said that, as indicated in Dr. Seemann’s
work, their characters are far from suggesting any connection in origin
with G. speciosa, the shore-species, the inland and littoral plants
belonging to different sections of the genus. In their case we can only
look to the frugivorous bird for the explanation of their existence in
the group. The fruits would be probably small; and in this connection it
is to be noted that Mr. H. N. Ridley in his paper on the flora of
Fernando Noronha evidently looks to birds to account for the presence of
a species of Guettarda on the island, a species not found elsewhere.

But another inland Fijian form of Guettarda found by me in Vanua Levu at
elevations of 1,000 to 1,400 feet above the sea, and dubbed by the
natives with the name of the littoral tree (Mbua-mbua), corresponds in
its close relation to G. speciosa with the inland Tahitian form of that
tree, and is to all appearance a derivative of it. It is chiefly
distinguished by its thinner, more hairy leaves, which taper at each end
and are not subcordate at the base as is often the case with the leaves
of G. speciosa. The coverings of the fruit are less fibrous and the
putamen is not so deeply notched or grooved. The difference also extends
to the buoyancy of the fruits in accordance with the principle laid down
in Chapter II. Whilst those of G. speciosa float for many months and are
of common occurrence amongst the stranded drift of tropical beaches, as
for instance in the Keeling Islands, in the Solomon Group, and in Fiji,
those of the inland species float only for a few weeks, their softer
coverings decaying more rapidly in sea-water.

We seem therefore to have had two principles at work in Fiji in
determining the origin of the inland species of Guettarda. Whilst in one
case the inland species is so sharply distinguished from the coast
species as to require the independent agency of frugivorous birds to
explain its presence, in the other the inland form, as in the instance
also of the Tahitian variety, is so much akin to it that the probability
of derivation from it is very great.

SECTION VI.

In this section are contained genera possessing littoral species
restricted to the Western Pacific islands, and dispersed by birds, but
having little or no capacity for dispersal by the currents. They are
regarded as derived from the inland species of the genus in the western
part of the Pacific, and as distributed from thence over the islands in
that part of the ocean. We are here only concerned with Fiji, Tonga, and
Samoa and the neighbouring islands. The genera Eugenia, Drymispermum,
and Acacia are here comprised.

The genus Eugenia, though essentially inland in its station, is apt to
lend species to the beach-flora in different parts of the tropics. Such
species, being dispersed by frugivorous birds and other animals, and
possessing but slight capacity for distribution by the currents, are
usually restricted in their areas. Thus, Schimper (p. 118) names two or
three species, including E. javanica, as amongst the Indo-Malayan
strand-flora. Ridley notices that E. grandis is a common sea-shore tree
in the Malay peninsula; and the author observed two littoral trees of
the genus in the islands of Bougainville Straits in the Solomon Group,
the fruits of one of them that flourished in the interior of the coral
islets being found in the crops of fruit-pigeons. So also in Fiji, some
of the inland species, as E. rariflora, appear at times amongst the
strand vegetation and in the coral islets. There is, however, one Fijian
species found also in Samoa and Tonga that is a characteristic beach
tree, namely E. richii (Gray), and it is more or less confined to that
station. The fruits will float a fortnight in sea-water, which is nearly
twice as long as most other Eugenia fruits will float; and it is quite
possible that the currents may assist the pigeons in distributing the
species. This genus is dealt with more in detail in Chapter XXVI.

The genus Drymispermum (Thymeleaceæ) comprises in the Western Pacific a
number of species, of which two range over the groups of Fiji, Tonga,
and Samoa, whilst some four or more are peculiar to Fiji. All are inland
plants with the exception of D. Burnettianum, a characteristic littoral
shrub of these three groups. Its bright red drupes float only from five
to ten days, even after some weeks of drying; and like those of the
inland species they are well suited for dispersal by fruit-pigeons. This
beach-plant may be regarded as probably an intruder in the strand-flora
from the interior of one of the islands of the Western Pacific, whence
birds, perhaps assisted a little by currents, have carried it to the
neighbouring groups.

The very remarkable coast tree, Acacia laurifolia, alone represents its
genus in the littoral flora of the Pacific islands. It is confined to
the Western Pacific, having been found in New Caledonia, the New
Hebrides, Fiji, Tonga, and Samoa; but it is doubtful whether it is truly
indigenous in all these localities. Thus, in Samoa, though restricted to
the coast districts, as we learn from Reinecke it seldom flowers, and
according to that botanist it was probably introduced through
cultivation. It is, however, evidently regarded by the Samoans as a tree
of their group, as is shown in a curious legend, given by Dr. George
Turner in his latest book on those islands, which I have quoted in my
book on the Solomon Islands, p. 287. Both in Fiji and Samoa it bears the
name “tatangia” or “tatania,” whilst its hard wood was employed for
various purposes, the leaves being used as spoons. The tree flowers and
seeds freely on the Fijian beaches. The pods dry up on the plant, and do
not dehisce, but are apt to break across between the seeds into
article-like portions, the seeds being ultimately liberated by the decay
of the pod or its fragments. The seeds either sink at once or in the
course of a day or two; whilst the pods or their fragments float at
first in sea-water, but all are at the bottom in a week or less. With
its absence of any apparent means of dispersal this small tree presents
quite an anomaly in the strand-floras of the Western Pacific, and can
only be regarded as a loan from the inland flora, though probably of a
very ancient date, and perhaps going back like Acacia koa, the
forest-tree of Hawaii to some early epoch in the history of these
islands.

_The conclusions to be drawn from the discussion of the relations
between the littoral and inland species of the same genus in the Pacific
islands._ (Chapters XIV., XV., XVI.)

In ten of the twenty-two genera here dealt with (Calophyllum, Hibiscus,
Colubrina, Morinda, Scævola, Cordia, Ipomœa, Vitex, Tacca, Casuarina)
the shore and inland species have their own independent modes of
dispersal, usually by currents in the case of coast plants, and by birds
in that of inland plants; and the relations between the two are not such
as to suggest a derivation of one from the other.

In six genera the inland species are regarded as derived from the
littoral species. In two of them, as in Vigna and Premna, where the
coast and inland species occur in the same group of islands and are
connected by intermediate forms, there is direct evidence in favour of
this conclusion; but such a development of inland species need not have
taken place in every group, since in the instance of Premna it has
apparently occurred only in the Western Pacific, and the inland and
coast species have extended independently to the eastern groups through
the agencies of birds and currents.... In the other four genera
(Canavalia, Erythrina, Sophora, Ochrosia) we have presented the
so-called “Hawaiian difficulty,” that group being alone concerned.
Although these genera have no littoral species in Hawaii, they have
inland species in those islands, which are in three genera endemic.
Since these inland species have non-buoyant seeds or seedvessels, the
transport of which by birds half-way across the Pacific Ocean is in the
case of the first three genera unlikely and in the last impossible, it
is assumed that they are all derived from original coast species with
buoyant seeds or fruits, such as are widely distributed over the Pacific
but are not now existing in Hawaii. This assumption, in the instance of
the Leguminosæ, to which the first three genera belong, derives support
from the singular fact in the distribution of the order pointed out by
Mr. Hemsley, that it is wanting in many oceanic islands where there is
no littoral flora.

In one genus, Guettarda, the inland species are regarded as having been
sometimes developed independently of the coast species, and as at other
times derived from it, both principles having been at work in Fiji and
only the last in Tahiti.

In two genera, Pandanus and Barringtonia, which represent the “Fijian
difficulty,” there is no reason on grounds of affinity to connect the
inland with the coast species; and since the agency of existing birds is
improbable in the first genus and out of the question in the second,
whilst the operation of the currents is excluded for the inland species
of both genera, it is assumed that we must either appeal to the agency
of extinct birds, such as those of the Mascarene Islands, or we must
fall back on the hypothesis of a continental connection. In the instance
of Barringtonia it is also possible that some of the inland species may
have been derived from species spread through cultivation.

Lastly, in three genera (Eugenia, Drymispermum, Acacia) the coast
species are viewed as derivatives of the inland flora in the Western
Pacific, not necessarily in Fiji, but it may be in New Caledonia or in
one of the other large groups. In this case the coast species of all
three genera are either unfitted for dispersal by currents, or display
the capacity only in a small degree.

We thus see that in only seven of these twenty-two genera, containing
both littoral and inland species in the Pacific islands, can it be
argued from the standpoint of dispersal that the inland species are or
may have been derived from the shore species; and in most instances the
evidence is largely presumptive in its character. In three genera the
reverse has been the case, and here the coast has borrowed from the
inland flora. In twelve, or more than half of the genera, the shore and
inland species have been evidently independent in their origin. It is
accordingly apparent that in the Pacific the strand flora has lent more
to the inland flora than it has borrowed from it; but with a large
proportion of these coast genera no interchange has taken place.
Two-thirds of the genera of the beach-plants have no inland species, and
in their case the question of such a connection cannot be raised. With
the remaining genera such a relation can be suggested in only two-fifths
of the cases, or in about one-seventh of the total number of beach
genera. Where a connection can be traced, it points more frequently to
the derivation of the inland from the shore plant. Taking all the
evidence together, the beach flora presents itself in the Pacific as
practically independent of the inland flora as regards its origin. It
has received in these regions but few recruits from inland. It has
yielded, except in Hawaii, but few recruits to the inland flora. In this
ocean it bears the stamp of a high antiquity, though in the mass no
doubt of more recent origin than the mangrove flora.

Yet, as I have remarked in different parts of this work, even with the
beach genera possessing no inland species, considerable variety is
displayed in the behaviour of the strand species. Thus, whilst some,
like Pemphis acidula, Tournefortia argentea, and Triumfetta procumbens,
rarely if ever leave the beach, others, like Heritiera littoralis and
Excæcaria agallocha, find a home on the borders of the mangrove swamps,
and one or two extend inland and take their place in the forests, either
as trees (Afzelia bijuga) or as giant climbers (Entada scandens). Others
again, like Cassytha filiformis, Cerbera Odollam, and Cycas circinalis,
with a number of other beach-plants, may invade the interior of the
island wherever arid plains or exposed scantily wooded districts offer
conditions conformable to the xerophytic habit of the beach-plants.

It will thus be perceived that although the inland and coast floras of
an island are in the mass distinct, the line of separation is by no
means always well defined. Beach-plants are something more than
salt-lovers in their ways. They are in the first place xerophilous, or,
in other words, they will be equally at home in exposed situations away
from the coast where the soil is dry and the rainfall scanty. Whenever
these conditions are presented by the districts backing the coast, as we
find for instance in the plains on the lee or dry sides of many a
Pacific island, the shore-plants will often leave the beach and travel
far inland.

_Summary of Chapters XIV., XV., XVI._

(1) Though littoral floras are as a rule chiefly made up of two sets of
plants, one brought through the agency of the currents from regions
outside, and the other derived from the inland flora of the region
concerned, the proportion of the two varies much amongst temperate and
tropical strand-floras, the current-borne plants forming the majority in
the tropics, and those from the inland flora of the region prevailing in
the temperate zone.

(2) There is, therefore, far greater uniformity as a rule amongst
tropical strand-floras than in the temperate zone, since in temperate
latitudes the prevailing constituents of the strand flora vary with the
inland flora of every region, whilst in the tropics the predominant
plants are those ranging far and wide on the shores of the warm regions
of the globe.

(3) Regarding the tropical strand-flora as comprising two formations,
that of the beach and that of the mangrove swamp, the last, which is the
older of the two, may, it is suggested, be viewed as the remnant of an
ancient flora widely spread over the lower levels and coastal regions of
the globe, during an age when, in a warm atmosphere charged with watery
vapour and heavy with mist and cloud, vivipary or germination on the
plant was not the exception but the rule.

(4) But it is contended that even in the beach formation some of the
plants may date back to this age of vivipary, as is indicated by the
anomalous seed-structures of some of the genera, such as Barringtonia,
which seem to indicate a lost viviparous habit.

(5) Since the beach formation of the islands of the tropical Pacific is
largely formed of plants ranging over great areas in the tropics, there
is no reason to expect that it owes much to recruits from the inland
floras of this region. The discussion, therefore, of the relation
between the littoral and inland floras is mainly concerned with the
possible origin of inland from coast plants in these islands.

(6) Yet there are numerous cases of genera possessing both coast and
inland species that are of peculiar interest in determining the true
relation between the beach and inland floras.

(7) As the result of a detailed discussion of these genera, the
conclusion is formed that the beach and inland floras have been in the
main developed on independent lines, the beach flora receiving from the
inland flora but few recruits, and except in Hawaii yielding but few
plants to the inland flora. Only a third of the genera of the beach
flora have also inland species, and in only a few of these genera, or
about a seventh of the whole beach flora, can any question of a
connection between coast and inland species of the same genus be raised.

(8) Two special difficulties arise in this discussion. The first is the
“Hawaiian difficulty,” which is more particularly concerned with genera
of the orders Leguminosæ and Apocynaceæ. Here are genera which possess
both inland and littoral species, but only the first occur in Hawaii. In
the absence of any likely means of dispersal, whether by currents or by
birds, it is assumed that the inland species are derived from shore
plants, originally brought by the currents, that have since disappeared,
a view supported by the fact that Leguminosæ are wanting in oceanic
islands where there is no littoral flora. The second is the “Fijian
difficulty” which is best represented by Pandanus. From our inability to
regard the inland species as derivatives of the coast species, or to
supply them with a means of dispersal, we are compelled to regard them
either as having been a part of the original continental flora of Fiji
or as owing their existence there to the agency of extinct birds having
the habits of the Nicobar pigeon and of the extinct Columbæ of the
Mascarene Islands. Since the Mascarene Islands are noted not only for
their extinct Columbæ but also for their number of peculiar species of
Pandanus, the implication seems to lie against the continental view. The
subject, however, awaits further investigation. In the Western Pacific
the possible agency of the parent forms of the existing species of
Megapodidæ is worthy of attention. Like the Columbæ and Pandanus in the
Mascarene Islands, the Megapodes and Pandanus have “differentiated”
together in the Western Pacific.

(9) The general view of the independent origin of the beach and inland
floras of the Pacific islands is supported by the large number of genera
in the strand flora that only possess littoral species.

(10) Such shore species, together with other strand plants, sometimes
extend into the interior of an island, but only as a rule where the
requisite conditions for a plant of xerophilous habit exist.

(11) Shore plants, it is pointed out, are xerophytes first and
halophytes afterwards; and under certain conditions the purely
xerophilous inclination prevails and the plants travel far inland.

CHAPTER XVII

THE STORIES OF AFZELIA BIJUGA, ENTADA SCANDENS, AND CÆSALPINIA
BONDUCELLA

IN this chapter we have a study of Leguminous strand plants that are of
great interest. It can be safely said that the student of
plant-dispersal in the Pacific will be brought into contact with the
problems here involved wherever he goes.

AFZELIA BIJUGA (Gray).

This Old World tree, which belongs to the sub-family Cæsalpiniæ, is of
great interest to the student of plant-dispersal. It is one of that
large group of Indo-Malayan plants that extend into the Western Pacific,
and give the prevailing character to the floras of such archipelagoes as
that of Fiji. It is a large tree yielding a valuable timber used by the
Fijians and Samoans for many purposes, such as for canoes, house-posts,
clubs, kava bowls, &c., but it has not been recorded from the Tahitian
region, and is unknown from Hawaii. In the fact of its being a littoral
as well as an inland tree, it possesses a peculiar interest from the
standpoint of plant-dispersal, and especially since this difference in
station is associated with a difference in buoyancy, the seeds of the
inland trees usually sinking, whilst those of the coast trees usually
float, and often for a period of months.

A glance at the distribution of the genus will enable us to appreciate
some of the points that will be touched upon in the following
discussion; and it may be here remarked that the explanation of the
distribution of these Leguminous trees will go far to make clear some of
the most difficult points in plant-geography. Of the eleven species
enumerated in the _Index Kewensis_, five belong to tropical Africa,
occurring on both the east and west coasts as well as in the interior,
three are confined to the mainland of tropical Asia, and two are
peculiar to Malaya. In the last place we have the wide-ranging Afzelia
bijuga, which, if it does not actually occur on the east coast of
Africa, is found at all events in Madagascar and in the Seychelles, and
is to be followed by the way of the Chagos Archipelago to the Malayan
Islands and Queensland, and eastward to Fiji and Samoa.

The most suggestive feature in the distribution of the genus is to be
seen in the frequent station of the species by rivers. We learn from
Oliver’s _Flora of Tropical Africa_ that these trees find a home along
river-courses on both sides of the continent, as on the banks of the
Congo, the Niger, the rivers of Senegambia, and the Zambesi, the Zambesi
species being found also on the shores of Lake Nyassa. Since tropical
Africa possesses about half of the species, it would seem highly
probable that it is the home of the genus, and that from the
rain-forests in the heart of the continent rivers flowing east and west
have borne the buoyant seeds of the wandering species to the coasts of
the Atlantic and Pacific Oceans. The operation that I witnessed on a
miniature scale in the case of a _species_ of Entada (E. scandens) in
the Isthmus of Panama, as described in a later page of this chapter, has
been in progress through the ages with the _genus_ of Afzelia in the
breadth of the African continent. According to the principle illustrated
by Afzelia bijuga in the forests of Fiji, the seeds of the African
forest-trees would, as a rule, possess no floating power; but now and
then in the lapse of long periods of time buoyancy in some species would
be developed, and such species would ultimately, through their buoyant
seeds, find their station along the lower courses of the rivers.

To sustain this view it is not necessary that continuous rain-forests
should now clothe the elevated regions in the interior of tropical
Africa; but it is requisite that there should be sometimes a generic
similarity between the plants of the East African and West African
rain-forests; and it is evident that this is the case. Pechuel-Lösche,
as quoted by Schimper (_Plant-Geography_, p. 299), describes the
rain-forest on the Loango coast as covering the mountain ranges and as
extending to the river-plains. In such a locality the operation would be
rapid. In advancing this hypothesis I am referring to the possibility,
however, of such an operation having effected the distribution of
Afzelia in tropical Africa in the past rather than in the present. I
would suggest that botanists in other habitats of the genus, as for
instance in Queensland, might put it to the test of observation and
experiment.

The interest that attaches itself to the story of the genus in its
African home may be extended to the species that forms its outpost in
the Pacific, and we shall see there a littoral species that doubtless
had its home in the interior of a continent endeavouring, with a
considerable measure of success, to become again an inland plant. Horne
(p. 112), who was familiar with Afzelia bijuga at the two extremes of
its range, namely, in the Mascarene Islands and in Fiji, speaks of it as
characteristic of the shores of tropical regions; and Schimper, who
includes it in the Indo-Malayan strand-flora, implies that it is more or
less exclusively confined to the coast and its immediate vicinity (pages
121, 191-2). In the Seychelles, according to Mr. Button, this tree
attains gigantic dimensions on the sandy flats. Still larger trees occur
in the coral islands of the Chagos Archipelago; but in the atoll of
Diego Garcia, as we learn from Mr. Bourne, it is almost extinct only
some four or five trees existing there about twenty years ago, the
increase of the tree being prevented through the destruction of the
fallen seeds by the rats (_Journ. Linn. Soc. Bot._, vol. 22, 1887).

Afzelia bijuga may, therefore, be safely regarded as a littoral tree. We
shall now see the importance of this conclusion when we come to consider
its station in the Pacific islands, where it grows both inland and at
the coast, and we have to decide to which station we must assign the
priority. Speaking of its occurrence in Fiji, Dr. Seemann says it is
“common in the forests all over Viti,” but makes no allusion to it as a
littoral tree either in Fiji or elsewhere. On the other hand, Mr. Horne
(p. 112) describes it as “generally growing on the shore or sandy
beaches, and in rocky clefts, and by the sides of streams in the
interior of Viti Levu and Vanua Levu.” It was on or near the coast in
Fiji that the present writer was most familiar with this tree, sometimes
bordering the sandy beach, at other times growing behind the
mangrove-belt, or again thriving in the half sandy and half swampy soil
of some low islet off the mouth of the Rewa. Especially is it to be
found on those parts of the coast where the hill-slopes descend rapidly
to the beach, or where some lofty spur from the mountains of the
interior reaches the shore. It is also not uncommon on the banks of
rivers both in their lower and upper courses. But it is as a forest-tree
of the interior that it is most valued by both the white men and the
natives on account of the superior quality of its timber in that
station. There, far removed from stream or river, the Vesi, as the
Fijians name Afzelia bijuga, takes its place amongst the lofty
forest-trees, such as the Ndamanu (Calophyllum), the Ndakua (Dammara),
and the Wathi-wathi (Sterculia). It is not often that one finds a tree
in these islands that, like the Vesi, is able to make its home in almost
any station, excepting, however, the “talasinga” or “sun-burnt” regions
of the plains. Wherever tall trees grow gregariously in Vanua Levu, one
will probably find Afzelia bijuga, whether beside a sandy beach, or
bordering a swamp, or on a river’s bank, or on some rocky declivity, or
on the great forest-clad mountain-slopes and plateaux of the interior.
No doubt the same diversity of station is displayed in Samoa, where,
according to Dr. Reinecke, the tree is most frequent in the
“coast-bush.”

From the variety in station it might be expected that corresponding
variations in character would be found. There are differences, such as
in the quality of the timber and in the size of the seeds between coast
and inland trees; but the most important distinction in connection with
the study of the dispersal of the species is to be found in the
circumstance that whilst the seeds of the coast trees are, as a rule,
buoyant, and often float for months, those of the inland trees usually
sink, even after being kept for three or four years. I made a
considerable number of experiments on the buoyancy of the seeds of this
tree in Vanua Levu, and found that with the coast trees, as a rule,
either all the seeds or the majority of them floated in sea-water,
whilst with the inland trees either all of them or the majority of them
sank. The buoyant seeds are able in most cases to float for a long time.
Thus, in one experiment half were afloat after two months, and in
another half were afloat after five months. It is probable that several
of the exceptions, where inland seeds float, will prove to be connected
with an inland station by a river. (I experimented on eight sets of
seeds of coast trees from eight different localities, and found 70 to be
the mean percentage of buoyant seeds. In the same way, four sets of
seeds from four different inland localities gave 13 as the mean
percentage of buoyant seeds.)

As in the case of Entada scandens, there is a rather fine adjustment
between the mean specific weight of seeds and the density of water. If
we place a number of the buoyant seeds in sea-water and begin to lower
the density, some of the seeds will at once commence to float heavily
and afterwards sink; and when the density has been lowered to
approximately that of fresh water, usually about a third will be found
at the bottom of the vessel. Out of 100 coast seeds, 70 will, as a rule,
float in the sea and about 47 in the river; whilst of the same number of
inland seeds, 13 on the average will float in sea-water and 8 or 9 in
fresh water. The bearing of facts of this kind is especially discussed
in Chapter X.

Coming to the causes of the floating-power of the seeds, we find that
with the buoyant seeds the kernel floats, whilst with the non-buoyant
seeds it sinks, the seed-tests in neither case possessing any
floating-power. In this respect, therefore, the seeds of Afzelia bijuga
belong, with the seeds of some other Leguminous littoral plants of the
Pacific islands, such as Canavalia obtusifolia, Erythrina indica, and
Sophora tomentosa, to the second section of the second non-adaptive
group of buoyant seeds (page 107). But though we can in a measure
explain the cause of the buoyancy, we are still ignorant of the manner
in which the difference in the buoyant behaviour of coast and inland
seeds has been brought about. It is possible that this may be connected
with another difference between the coast and inland seeds, the latter
being markedly smaller, and it is noteworthy that in my experiments the
smaller seeds were generally those that sank. (Whilst the inland seeds
averaged between 8/10 and 1 inch, or 20 to 25 mm., in greatest diameter,
12 to 16 being required to make an ounce, the coast seeds measured 1 to
1-2/10 inch, or 25 to 30 mm., and only 10 or 11 were needed to weigh an
ounce.)

There can be no question that the seeds are at times transported by the
currents over wide tracts of sea, and this no doubt explains the
occurrence of Afzelia bijuga in oceanic islands. They may be usually
seen lying free in numbers on the ground beneath the tree or else still
inclosed in the fallen dehiscing and decaying pods; and they might be
swept sometimes into the sea or washed down into an adjacent stream.
They thus came under my notice amongst the stranded beach drift at the
mouths of estuaries in Fiji. But it is remarkable that the seeds have
not apparently been recorded from the beach drift of other tropical
regions. Penzig does not note them amongst the seeds stranded on the
shores of Krakatoa. They did not occur amongst my collections from the
beaches of Keeling Atoll or of the south coast of Java; nor does
Schimper mention them amongst the drift of the Java Sea. In the _Botany
of the “Challenger” Expedition_ the species is not even referred to in
any connection. Although, however, the capacity of these seeds for
dispersal by currents is for the first time established by me, their
fitness in this respect was surmised by Schimper (p. 191), when he
placed the species in his list of tropical shore plants evidently
distributed by the currents.

It will thus be gathered that we have yet much to learn in this matter;
and I would recommend any resident in the tropics to take up this
subject. When indeed we remember the fine adjustment existing between
the specific weight of the seeds and the density of water, and recall
the unknown factor determining the difference in buoyancy between the
kernels of coast and inland seeds, we can understand how under
particular conditions in certain portions of its range the seeds of
Afzelia bijuga may perhaps never possess any floating power. It would
seem, in fact, that the seeds are much more buoyant in the Western
Pacific than they are in the Java Sea; or it may be that the tree is
much less frequent; or that the stranded seeds are soon destroyed by
crabs, such as is the fate of much seed-drift on the Keeling beaches; or
lastly that, as in Diego Garcia, rats in destroying the fallen seeds are
bringing about the extermination of the species.

_Summary relating to Afzelia bijuga._

(1) Assuming that the genus has its home in the African continent, and
that the species have frequently a riverside station, it is argued that
the distribution of the genus on both sides of that continent can only
be explained by its dispersal by rivers from a centre in the interior.

(2) Afzelia bijuga, a widely distributed shore tree of tropical Asia,
occurs in Fiji both at the coast and in the inland forests.

(3) This double station is associated _inter alia_ with a different
buoyant behaviour of the seeds, those of the coast trees usually
floating for long periods, whilst those from inland generally sink.

(4) There can be no doubt that this widely ranging littoral tree has
been dispersed by the currents; but the specific weight of the coast
seeds is on the average but slightly less than that of sea-water; and it
is to this fine adjustment, always liable to be disturbed by variations
in the environment, that the irregularities in the distribution of the
species are to be attributed.

ENTADA SCANDENS (Benth.).

The story of Entada scandens, a plant familiar to many of my readers
under the name of the Queensland Bean, is a story of three continents,
Africa, Asia, and America. From the point of view of its dispersal two
features at once attract attention in the case of this giant-climber; in
the first place its wide distribution over the tropics of the Old and
New Worlds, and in the second place the great capacity of its large
seeds, often two inches across, for dispersal by the currents. But
before discussing these matters it will be necessary to glance at the
distribution of the genus, since much light will thereby be thrown on
some of the numerous difficult points affecting this extremely
interesting tropical plant. Of the thirteen species enumerated in the
_Index Kewensis_, seven are African, three are American, one is Burmese,
one hails from Madagascar, and, lastly, there is the world-ranging
Entada scandens, concerning whose home botanists are not agreed. Most of
the species would seem to be inland plants, whilst Entada scandens
thrives both inland and at the coast. Africa would thus appear to be, as
with Afzelia, the principal home of the genus, but with America as a
subsidiary centre.

In many points Entada scandens presents a parallel to Cæsalpinia
bonducella, another Leguminous tropical plant which occurs also at the
coast and inland. But since they both owe their wide distribution to
their littoral station, it will be as coast plants that they will be
most properly considered in this and the following chapter. Yet if the
student were to regard the distribution of these two plants in a
continental region as in India, where they extend inland to the
Himalayas, he might fail to discern their true station. To accurately
gauge the matter of their station, it is necessary for him to look at
the plants as they occur in the islands of the Pacific. There he will
first see the stranding of the seeds on a shore by the currents, then
their germination and their development into giant-climbers over the
littoral trees or into straggling bushes on the beach; and afterwards he
will observe the plants of both species extending inland, and in these
three stages he will learn their history in the Pacific; but a history,
it may be observed, that in this region represents their efforts to
return to an inland station, such as they once possessed in their
birthplace in some distant region of the globe.

Dealing first with the station of Entada scandens, it may be remarked,
as Dr. Seemann points out, that in Fiji it is most characteristic of the
mangrove-formation. But it also occurs amongst the trees at the back of
the mangrove swamp, on the beaches, on the banks of the estuaries, and
at the edge of the inland forests where they border on the plains.
Sometimes in the company of Derris uliginosa it grows not as a climber,
but as a prostrate plant on the sandy beaches; and here, not being able
to assume its normal habit of a climber, it does not seed. It is to be
found at times far inland in open-wooded districts. Thus in Vanua Levu I
found it growing in the Mbua district four miles inland, and 1,400 feet
above the sea. Reinecke speaks of it in Samoa only in connection with
the “urwald,” or primeval forest. Cheeseman describes it as most
abundant in the interior of Rarotonga, covering the trees with a
wide-spreading canopy of green. In the Malayan region Schimper refers to
it as a plant of the beach-tree formation. In Ecuador and on the Panama
Isthmus it grows not only at the coast, but also on the hill-slopes in
the rear of the mangrove-belt.

With reference to the distribution of the plant, it may be remarked
that, although it is found all round the tropics and possesses great
capacity for dispersal by currents, there are certain difficulties in
explaining its wide area and in accounting for its very peculiar
distribution in the Pacific islands. It was doubtless in allusion to
some of these difficulties that Mr. Darwin, in a letter to Sir Joseph
Hooker, remarked: “Entada is a beast” (_More Letters_, &c., i, 93).
There is at first the question of the identity of the species in the Old
and New Worlds. It is here assumed that it is the same in both
hemispheres; but it must not be forgotten that the identity is “not
beyond doubt” (_Bot. Chall. Exped._ iv, 147).

Then there is the difficulty connected with its occurrence on both
coasts of tropical America. In this respect it is at one with some other
littoral plants, like Ipomœa pes capræ, as well as with the plants of
the mangrove formation, as is pointed out in Chapter VIII. Whilst with
the mangroves it is necessary to assume that they antedate the land
connection between North and South America, this is not requisite in the
case of Entada scandens, since it grows in the interior of the Panama
Isthmus, and rivers on the north and south sides now carry its seeds
seaward from the same “divide” to the Atlantic and Pacific Oceans, as
described in Chapter XXXII.

But, as I have also shown in Chapter VIII, America forms with the West
Coast of Africa a region characterised by the same tropical littoral
flora. This region, on account of the arrangement of the currents,
stands in a very peculiar relation with the Asiatic region, which
comprises the rest of the tropics, and to a great extent possesses its
own peculiar strand-flora. There are a few littoral plants, like Entada
scandens, Canavalia obtusifolia, Sophora tomentosa, and Ipomœa pes capræ
that occur in both areas; but the large majority are confined to one or
other of them, either to the American region, including the African West
Coast, or to the Old World region, which includes the African East
Coast. The American region gives to the Old World, but it can receive
nothing in return. For this reason, it is argued, we are compelled to
regard most, if not all, of the cosmopolitan tropical shore plants that
are dispersed by the currents, such as those above named, as having
their home in the American region. Entada scandens would, therefore,
from this standpoint have its home in America.

Then, again, there is the difficulty connected with the distribution of
this plant on both sides of tropical Africa. Though Oliver in his _Flora
of Tropical Africa_ mentions this species only in connection with the
West Coast, he says it is probably widely spread in that continent, and
he refers to a pod in the Kew Museum indistinctly labelled “Lake Ngami.”
I have not come upon any reference to its being a littoral plant on the
East Coast, but since numerous littoral plants of tropical Asia are
found on that coast its occurrence there or in the East African islands
would be expected. However, as the genus has a centre in America, and as
this species is regarded as of American birth, we are not called upon to
employ the argument used in assigning to a non-American genus like
Afzelia an African home. Since the African West Coast belongs to the
American region of tropical shore plants dispersed by the currents, the
presence of Entada scandens on that coast of Africa can be readily
explained, whilst if it has reached the Malayan Archipelago from America
by way of the Pacific, it would, by extending like many other Malayan
coast-plants along the shores of the Indian Ocean, almost complete its
circuit of the globe. It is in this fashion, I believe, that the other
littoral plants,, like Cæsalpinia bonducella, Canavalia obtusifolia, and
Ipomœa pes capræ, that are found all round the tropics, have performed
the circuit of the globe with America as their home.

One may remark in passing that the double home of the genus in America
and the Old World, though offering a serious difficulty in plant
geography, has no immediate bearing on the present mode of distribution
of Entada scandens. Questions relating to the distribution of tropical
shore-plants that are dispersed by the currents at first resolve
themselves into considerations of the arrangement of the currents.
Entada is not alone amongst the genera containing littoral species in
having a home both in the Old and in the New World. Carapa is another
instance, and additional cases might be cited.

The next peculiarity in the geographical range of this species is
concerned with its irregular distribution in the archipelagoes of the
tropical Pacific. Notwithstanding its great capacity for dispersal by
the currents, although it occurs in all the groups of the Western
Pacific as well as in the Cook Islands, it has not been recorded from
the Society Islands, the Paumotus, the Marquesas, and Hawaii. Since,
however, its seeds have been gathered by Mr. Arundel on the beaches of
Flint Island, lying about six degrees north of Tahiti (_Bot. Chall._ iv,
302), it is not unlikely that it will be found growing in other parts of
Eastern Polynesia south of the equator. One might have looked for an
explanation of its rarity in Eastern Polynesia to the absence of
mangrove swamps, in which, as in Fiji, it is sometimes most at home; but
this is negatived by its abundance in Rarotonga, where mangrove swamps
do not exist.

_The dispersal of Entada scandens by the currents._—This plant offers
one of the most conspicuous examples of the transport of seeds across
oceans through the agency of the currents. In the pages of many
botanical works, from the close of the 17th century onward, reference is
made to the transport of its beans (often in association with those of
Mucuna urens and Cæsalpinia bonducella) by the Gulf Stream or other
currents across the Atlantic to St. Helena, the Azores, the west coast
of Ireland, the Hebrides, the Orkney Islands, the coasts of Scandinavia,
and even as far north as Nova Zembla (see Hemsley’s _Bot. Chall.
Exped._; Sernander’s _Skand. Veg. Spridningsbiologi_, &c.). That the
seeds of Entada scandens retain their germinating capacity after this
ocean-transport has been demonstrated not only by the germination of
stranded seeds on the shores of St. Helena, but also by the germination
when sown at Kew of seeds drifted to the Azores, as well as by the
results obtained by Lindman, who procured the germination of the seeds
of this plant and of Mucuna urens that had been washed up on the
Scandinavian beaches (see Sernander, pp. 7, 390).

One of the most interesting references to the conveyance by currents of
these seeds to the coasts of Europe is to be found in Dr. Sernander’s
recent work on the modes of dispersal of the Scandinavian flora, where
he sums up the results of Lindman’s investigations respecting the Gulf
Stream drift. The stranded seeds of Entada scandens, it appears, have
been found all along the Norwegian coast, but occur most frequently
north of the Söndmöre district. They have even been found in a
sub-fossil condition in the peat-bogs of Tjörn on the Bohuslän coast in
Sweden, having been originally stranded on a beach in that locality at
some distant, but post-glacial, epoch. Few phenomena in
plant-distribution are more suggestive than this ineffectual transport
through the ages of these large tropical beans to coasts within the
Arctic Circle. The seed, no longer under the care of the mother-plant,
becomes a waif, exposed to the pitiless laws of the physical world which
here prevail. It was not thus that the plant was reared, but it is in
this haphazard fashion that its seeds are spread. The philosopher could
unravel most of the tangled problems connected with present and past
plant-dispersal, if he could follow the clue supplied by this stranded
tropical seed on a Scandinavian beach.

It is a far jump from the North Cape to the coral islands of the Pacific
and Indian Oceans; yet it is within the area covered by the drifting
Entada bean. The stranded seeds occur commonly on the Fijian beaches and
on other islands of the South Pacific; but I never found them in Hawaii.
They were gathered by me on the shores of Keeling Atoll in the Indian
Ocean, and on the south coast of Java. Penzig found on the Krakatoa
beaches, in 1897, not only the stranded seed but the established plant.
They came under my notice in numbers on the beaches of Ecuador and on
the Pacific and Atlantic coasts of the Panama Isthmus; and, as I
learned, they are equally common on the other parts of the coasts of
Central America. Not uncommonly these stranded seeds in various parts of
the world are to be found incrusted with polyzoa and tubicular annelids,
which afford proof of prolonged flotation in the sea. These seeds are
also to be frequently noticed floating in the drift of the tropical
estuaries. Thus they came under my observation afloat in numbers in the
Fijian estuaries, in the Guayaquil river, in the estuary of the Chagres
at Colon, and in the mouth of a river on the Panama side of the isthmus.

The mode of liberation of the seeds is worthy of a passing remark. The
huge pods, often several feet in length, ultimately break up into
separate joints bearing the seeds. The joints may decay on the ground,
and the seeds are thus freed; or not infrequently in a mangrove-swamp
they fall at once into the water, and there they float, as may often be
observed in Fijian rivers, until their decay sets free the seed.

The seeds of Entada scandens are often quoted, and justly so, as
striking examples of the dispersal of seeds by currents. Yet in few
plants could the nature or the structural cause of the buoyancy have so
little claim to be considered as adaptive in its character. Quite half,
and sometimes even the majority, of the seeds freshly liberated from the
plant have no buoyancy at all. The mean specific weight of the seed is
about that of sea-water, but markedly higher than that of fresh water;
whilst the principal determining cause of the buoyancy is, as shown
below, purely mechanical, and one that, whilst favouring the wide
distribution of the species, could not be improved by or come within the
scope of Natural Selection.

From experiments made in Fiji and Ecuador, it appears that at least 50
per cent., and often more than half, of the seeds when first liberated
from the pod have no buoyancy in sea-water. Of those that float in
sea-water, a proportion varying between one-third and one-half sink in
fresh water, so that in the case of plants growing on the banks of a
river only about one-fourth or one-third would be carried down to the
sea. So fine is the adjustment of the specific weight of these seeds to
the density of water, a subject discussed in its general bearings in
Chapter X, that if one gathers a number of drift seeds on a beach, let
it be in Fiji or in Ecuador, although, of course, all will float in the
sea, only one-half or two-thirds will float in the neighbouring
fresh-water stream. Those that float appear to be able to float almost
indefinitely. This is sufficiently established by the transport of the
seeds in a sound condition by the currents across the Atlantic, and by
such evidence as the stranding of seeds incrusted with polyzoa and
serpulæ on the beaches of Keeling Atoll. It has been also proved by the
following experiment. Several years since, I placed a seed in a vessel
of sea-water, where it still floated buoyantly in a perfectly sound
condition twelve months afterwards.

With regard to the cause of the buoyancy, investigation shows that
neither the seed tests nor the seed contents have any floating power,
the buoyancy arising from a large central cavity produced by the
shrinking and bending outward of the cotyledons during the drying and
hardening of the maturing seed (see figure in Chapter XII). With the
seeds that sink, this cavity is, as a rule, reduced to small dimensions,
and may be represented only by a narrow slit. In some cases, however,
where the cotyledons are unusually thick and heavy, even a large central
cavity will not give floating power to the seed. There is an indication
in my experiments that seeds from inland plants that have matured their
pods in the forests sink in a much greater proportion than seeds of
coast plants, or of those growing on the banks of estuaries. This we
might expect, since in the shade of the forests the drying process that
accompanies the setting and final maturation of the seed would be less
complete and the intercotyledonary cavity smaller than with the seeds
matured in more exposed situations. This is a point, however, that
requires further investigation.

It will thus be seen that in respect of buoyancy the seeds of Entada
scandens are to be referred to the mechanical or non-adaptive group of
buoyant seeds, described in Chapter XII, which comprises several other
Leguminous strand-plants, including Cæsalpinia bonducella. I especially
studied the various stages in the development of the buoyancy of seeds
in this mechanical group in the case of the species of Cæsalpinia just
named, and the description of the process as given under that plant will
apply to all.

_Summary relating to Entada scandens_

(1) This plant, which has been distributed by the currents over the
tropics of the globe, has its station in the mangrove swamp, on the
beach, by the estuary, and in the inland forest.

(2) It is regarded as an American plant that has reached the shores of
the Indian Ocean by crossing the Pacific, and the coast of West Africa
by crossing the Atlantic.

(3) Its occurrence on both coasts of America is attributed to its having
a focus of dispersal in the forests of Central America, from which its
seeds have been transported by the rivers to the shores of the Atlantic
and Pacific Oceans.

(4) Its irregular distribution in the Pacific islands, to wit, its
absence from Hawaii and its rarity in the Tahitian region, is not to be
easily explained, but it is more than likely that it will be
subsequently recorded from other localities in Eastern Polynesia.

(5) Although the seeds offer a striking example of dispersal by
currents, since they are to be found stranded on beaches over much of
the globe, from within the Arctic Circle to the Coral Sea, in few plants
could the character of the buoyancy and the structure connected with it
have so little claim to be considered as adaptive in their nature. At
least 50 per cent. of the seeds sink in sea-water, and the cause of the
buoyancy of the other seeds is only to be connected with the large size
of a cavity produced by the shrinking of the embryo within the seed
tests during maturation.

CÆSALPINIA

This genus is represented in the tropics of both the Old and the New
World by some eighty species of trees, shrubs, and climbers, some of
which are noted for their dye-woods, and others for the beauty of their
flowers. In the Pacific islands the botanist is only concerned with
three widely distributed species, all more or less littoral in their
station, and in great part dispersed by the currents, namely, Cæsalpinia
nuga (Ait.), C. bonducella (Flem.), and C. bonduc (Roxb.).

With Cæsalpinia nuga we have little to do, since, although widely
distributed in tropical Asia and the Malayan region, and reaching to
both New Guinea and North Australia, it has not apparently penetrated
into the Pacific further east than the Solomon and New Hebrides groups.
I found it growing on the coasts of the larger islands of the Solomon
group, but no observations were made on its mode of dispersal. However,
as its seeds were identified at Kew (_Bot. Chall. Exped._ iv, 311)
amongst my collections of stranded drift from those islands, it would
appear to be to some degree dispersed by the currents, though since it
does not extend far into the Pacific, its capacity for dispersal by this
agency would seem to be limited. Schimper includes it among the
strand-plants of the Indo-Malayan region.

It is with the other two species, Cæsalpinia bonducella and C. bonduc
that we are especially interested. Their extremely hard, marble-like
seeds at once attract attention, and when pale in colour they look not
unlike quartz pebbles as they lie stranded on a beach. The prickly pods
and the recurved prickles of the leaf-branches often make these plants
provokingly evident to a stranger. Though usually to be characterised
when growing on a beach as straggling shrubs, they will often climb
trees when opportunities occur, and they then display themselves as
stout-stemmed climbers. I have seen one or other of them in the mangrove
swamps of Fiji ascending the Bruguiera trees to a height of 30 feet and
more, the stem quite bare below, but leafing and flowering in the
tree-branches above.

From the standpoint of dispersal there are few more interesting plants
in the Pacific islands; but their discussion raises several difficult
questions, and it will be, therefore, requisite to treat them somewhat
in detail. With regard first to the diagnostic characters between the
species, it may be observed that, as a rule, they are sufficiently
evident, such, for instance, as the number, size, and form of the
leaflets, the presence or absence of foliaceous stipules, and the colour
of the seeds, though, as shown below, the seed-colour in the case of
Fijian plants does not always present a constant distinction. Yet as I
found in Fiji the difference between the two species is not in all cases
well pronounced, and intermediate forms occur, about which it is
sometimes difficult to decide to which of the two species they should be
assigned.

Mr. Hemsley remarks (_Bot. Chall. Exped._ iii, 114, 145, 300) that the
two species have been often confused. I venture to think that this has
been in some cases due to the occurrence of these intermediate forms.
One has only to look at the different “distributions” given by botanists
for C. bonduc, as indicated below, in order to suspect that the cause of
confusion has been at times with the plants themselves. When in Fiji I
paid a good deal of attention to this subject, and the results of the
comparison of the foliage and seeds of the plants obtained from fourteen
different localities in Vanua Levu are given below.

It will be seen in this table that I distinguish in Fiji three littoral
forms and one inland or mountain variety, which may perhaps be a
distinct species. Those of the strand include Cæsalpinia bonducella, C.
bonduc, and an intermediate form. C. bonduc is typically distinguished
by its large leaflets, by the absence of foliaceous stipules, and by its
pale yellow seeds; whilst C. bonducella is similarly characterised by
its small leaflets, its foliaceous stipules, and its lead-coloured or
darkish grey seeds. But in the first species the colour of the seeds may
often be yellow mixed with pale-grey, or almost white; whilst in the
second species the seeds may be stained with brownish-yellow patches.

It seemed to me when examining fresh specimens in Hawaii and Fiji that
the ultimate colour of the seed is a good deal determined by the degree
of alteration of the original olive-green colour of the immature seed.
All gradations may be noticed from the olive-green of immaturity to the
yellow, pale grey, and dirty white hues of the mature seeds of
Cæsalpinia bonduc and to the lead or slate-colour of those of C.
bonducella. It almost appeared as if the changes might be compared to
the bleaching which a dark volcanic rock undergoes in the weathering
process through the hydration and removal of the iron oxides.

CÆSALPINIA IN FIJI, TAHITI, AND HAWAII.

A = Locality.
B = Species.
C = Foliaceous stipules.
D = Pairs of pinnæ.
E = Leaflets—Pairs.
F - Leaflets—Length in inches.
G = Leaflets—Form.
H = Seeds—Size in tenths of inch.
I = Seeds—Colour

+--------+---------------+--------+------+------+-------------+-------------------+-------------+----------------------+
| A | B | C | D | E | F | G | 8 | 9 |
+--------+---------------+--------+------+------+-------------+-------------------+-------------+----------------------+
| {| Bonducella |Present | 8-9 | 9-11 |1-1/4 - 1-1/2| Oblong, obtuse |6-1/2 - 7-1/2| Usually lead- |
| {| | | | | | mucronate: base | | colour with at |
| {| | | | | | rounded and in— | | times brownish- |
| {| | | | | | equilateral | | yellow patches. |
| {| | | | | | | | |
|Coast, {| Bonduc |Absent | 5-6 | 4-6 | 2-1/2 - 5 | Oblong, acuminate | 5-1/2 - 6 | Pale yellow |
|Fiji {| | | | | | mucronate, base | | |
| {| | | | | | rounded or |6-1/2 - 7-1/2| Pale grey, sometimes |
| {| | | | | | subcordate | | mixed with |
| {| | | | | | | | yellow. |
| {| | | | | | | | |
| {| Intermediate |Present | 7-8 | 7-9 | 2-3 | Oblong, obtuse | 6-7 | Lead-colour or |
| | | | | | | mucronate, rounded| | pale grey with |
| | | | | | | at base; upper | | brownish-yellow |
| | | | | | | leaflets may be | | patches |
| | | | | | | elliptical | | |
| | | | | | | | | |
|Inland, | Mountain |Present | 5-6 | 9-10 |1-1/2 - 2-3/4| Lanceolate with | 6 | Yellowish or pale |
|Fiji | species | | | | | long tapering | | grey or mixed. |
| | | | | | | aristate apex | | |
| | | | | | | and rounded | | |
| | | | | | | base | | |
| | | | | | | | | |
|Coast, | Bonducella |Present | | | 1/2 - 1-4/5| Oblong | | |
|Tahiti | | | | | | | | |
| | | | | | | | | |
|Inland, | Bonduc |Absent | | 5-6 | | Oblong | | |
|Tahiti | | | | | | | | |
| | | | | | | | | |
|Inland, | Bonducella | | 4-6 | 6-8 | 1-4/5 - 2 | Oblong, obtuse, | 6-7 | Lead-colour. |
|Hawaii | | | | | | not cordate at | | |
| | | | | | | base | | |
+--------+ --------------+--------+------+------+-------------+-------------------+-------------+----------------------+

_Note._—The characters of the Fijian plants are from my own
observations. Drake del Castillo is quoted for Tahiti, and Hillebrand
for Hawaii. Reinecke observes that the pods of C. bonducella in the
inland forests have no prickles.

In Fiji all three coast forms may be found on the same beach, or they
may exist apart. The large-leaved species (C. bonduc) appears to be much
the most frequent in Vanua Levu; and the intermediate form is common
enough to disturb the serenity of the observer’s mind when he is anxious
to diagnose rather than to collect cumbersome specimens. The mountain
form, which came under my notice as a climber in the forest at an
elevation of 1,700 feet on the slopes of Koro-mbasanga in Vanua Levu,
acquires from the lanceolate shape of its leaflets quite a character of
its own, though it comes nearest to Cæsalpinia bonducella. Mountain
forms also occur, as indicated in a later page, in the forests of Samoa
and in Tahiti; but in the first-named group they are referred by
Reinecke to C. bonducella, and in Tahiti by Drake del Castillo to C.
bonduc. In the Samoan forests the inland plants possess pods deprived of
the prickles that are so characteristic of the beach plants. Before one
can pronounce definitely on the relation between the coast and inland
forms in any of the groups, a thorough investigation of the connections
between the two shore-species is needed. I am inclined to think that
they will prove to belong to a single dimorphic (or perhaps polymorphic)
species.

_The distribution of Cæsalpinia bonducella and C. bonduc._—Botanists
agree in giving C. bonducella a distribution around the tropics of the
globe; but they are not at all unanimous with respect to the other
species. According to Mr. Hemsley this species is by no means so
universally dispersed as C. bonducella. It is unknown from Africa and
Australia; but it is generally characteristic of tropical Asia and the
Malay Archipelago. The same authority alludes to specimens in the Kew
Herbarium from Florida and the West Indies (_Bot. Chall._ iv, 300).
Drake del Castillo gives both species a range through the tropics,
whilst Schimper seems in doubt about the occurrence of C. bonduc in the
New World, and Mr. Burkill makes no allusion to its American habitat in
his paper on the Tongan flora. The cause of this confusion is doubtless
to be mainly attributed to the variation in characters of the plants,
and to the occurrence of intermediate forms.

We should be scarcely consistent if we assumed that of two kindred
shore-species dispersed by the currents one had its home in America and
the other in the Old World. The same home must belong to both. According
to the principle laid down in Chapter VIII, and referred to under Entada
scandens, it is held that a strand-plant, with its home in Asia, on
account of the arrangement of the currents could never reach the
American continent, and that American shore-plants are for the most part
native-born except those hailing from the African West Coast, which,
however, lies within the American province of tropical strand-plants.
From this standpoint Cæsalpinia bonducella would be regarded as now
having its home in the New World, and since it is found on both the
Pacific and Atlantic coasts of that continent (as well as on both coasts
of Africa), it is assumed, as with Entada scandens, that it has reached
the African West Coast by crossing the Atlantic, and the African East
Coast by way of the Pacific and Indian Oceans. The genus, I may remark,
is distributed over the tropics of the eastern and western hemispheres.

As regards the general distribution of the two species in the Pacific
islands, it would appear from the writings of Seemann, Hillebrand,
Hemsley, Drake del Castillo, Reinecke, Cheeseman, and Burkill that with
the exception of Hawaii and Samoa, where Cæsalpinia bonducella alone
occurs, and of Rarotonga where C. bonduc alone is found, they are
generally associated in the larger groups, as in Fiji, Tonga, Tahiti,
and the Marquesas.

_The station of Cæsalpinia bonducella and C. bonduc._—Both the species
are to be regarded as littoral plants likely to stray inland. The
first-named is described in the _Botany of the “Challenger” Expedition_
as essentially a sea-side plant, though flourishing equally well inland,
and in India extending to the Himalayas as far as Kumaon, and up to
elevations of 2,500 feet. Schimper speaks of both species as
characteristic of the Indo-Malayan strand-flora, and he quotes Kurz when
referring to C. bonduc as a constituent of the beach-jungle of Pegu.

In the Pacific islands they are typically littoral in their station; but
they may extend inland, and in one or two groups they are only known in
their inland station. Dr. Seemann speaks of both species only in
connection with the beaches in Fiji, and alludes to Cæsalpinia
bonducella (p. 72) as sometimes climbing over the mangroves. In Vanua
Levu both came under my notice on the beaches, and in their immediate
vicinity, usually as straggling bushes, whilst at times they were to be
observed climbing the mangroves at the borders of the adjacent swamp. In
this island of the Fijis they do not, as a rule, stray far from the
beach, and strange to say are not to be ranked amongst those seashore
plants that frequent the “talasinga” regions or inland plains. Judging
from the mountain form found in the forests of Koro-mbasanga, if they
extend inland in Fiji they prefer the forests and become differentiated
in character. In Tahiti, as we learn from Nadeaud and Drake del
Castillo, C. bonducella occurs on the beach and extends inland to the
mouths of the valleys; whilst C. bonduc is only recorded from the
mountains at elevations of 600 to 700 metres (2,000 to 2,300 feet).
Jouan is quoted by Mr. Hemsley as remarking that C. bonduc is as common
in the Marquesas as brambles are in Europe (_Bot. Chall. Exped._ iii,
145). In Rarotonga, according to Cheeseman, C. bonduc is restricted to
the interior. In Samoa, as we are informed by Reinecke, C. bonducella is
frequent both in the coast districts and in the mountain-forests. In the
Samoan mountains the pods lose their prickles, and from this
circumstance, as well as from the extremely widespread distribution of
the species over the islands, the German botanist concludes that the
plant has been for ages established in the group.

In Hawaii, Cæsalpinia bonducella, which alone occurs, rarely figures as
a beach plant; but it is found, as Hillebrand observes, in the lower
plains of all the islands. In the large island of Hawaii I found it not
on the scanty beaches of the coast, but on the partly vegetated surface
of the old lava-flows at distances varying usually between a hundred
yards and a mile from the sea, but extending at times a few miles
inland, and in one locality reaching an elevation of 2,000 feet above
the sea. It was mostly observed by me on the dry side of the island,
where, associated with Erythrina monosperma, the Cactus, and the
Castor-Oil plant, it thrives in very arid localities, where the rainfall
is only a few inches in the year. Farther inland, where the old
lava-surfaces were more vegetated, it was associated with such shrubs as
Osteomeles anthyllidifolia and Cyathodes tameiameiæ. Dr. Hillebrand,
writing of a generation and more ago, says that in his time the plant
was less common than formerly.

_The Methods of Dispersal of Cæsalpinia bonducella and C. bonduc._—We
come now to the modes of dispersal of these plants; and in so doing we
have to choose between the agencies of birds and of currents. The seeds
of C. bonducella are on the average 7/10 of an inch (18 mm.) in
diameter, whilst those of C. bonduc are rather smaller (6/10 of an inch
or 15 mm.). As far as their size and character go, it would seem
scarcely likely that birds could transport these seeds across an ocean;
but our knowledge of the agency of birds is of a very imperfect nature.
Yet their occasional dispersal by birds is not improbable. When I was in
the Keeling Islands the residents informed me that the seeds of C.
bonducella are sometimes found in the stomachs of sea-birds, such as
frigate-birds and boobies. (See Note 59.)

However, it has long been known that the seeds of one or both of these
species are carried great distances by the currents; but it is to be
gathered that the older botanists, in alluding to this fact, more
usually referred under the synonym of Guilandina bonduc to Cæsalpinia
bonducella. De Candolle, loth to attach much importance to the effective
transport of seeds by currents, was compelled to admit this species in
his scanty list of current-dispersed plants (see Note 33). For more than
two centuries it has been known that the seeds of C. bonducella are
carried in the Gulf Stream drift to the coast of Europe from the
American side of the Atlantic; and ever since they were recorded by
Sloane in 1696 as stranded in a fresh condition on the beaches of the
Orkney Islands, they have been found washed up on other localities, as
on the coasts of Ireland and of Scandinavia and on the shores of the
islands of the Western Atlantic. According to Robert Brown, a plant was
raised from a seed cast up on the west coast of Ireland; and with
respect to Scandinavia, Dr. Sernander informs us that the seeds of
Cæsalpinia bonducella, like those of Entada scandens and of Mucuna
urens, are of frequent occurrence amongst the “Gulf Stream products”
stranded on the Norwegian coasts. The seeds of this species are commonly
washed ashore at St. Helena, and there are specimens in the Kew Museum
that were stranded on Tristan da Cunha. (Those interested in the subject
will find it discussed by Mr. Hemsley in the _Botany of the “Challenger”
Expedition_, and also by Dr. Sernander in his recent work on
Scandinavia.)

The seeds of Cæsalpinia bonducella have been also found stranded on
beaches in other parts of the world. Thus Prof. Schimper found them in
the beach-drift of the south coast of Java. Prof. Penzig noticed them
amongst the stranded seeds of the Krakatoa beaches; but it does not
appear that the plant had established itself up to the date of his visit
in 1897, or fourteen years after the great eruption. They have been
picked up on the other side of the Indian Ocean on the east shores of
Africa (_Bot. Chall. Exped._ iv, 300). They came frequently under my
notice stranded on the beaches of Keeling Atoll in the same ocean; and
seedlings sprouting from the seeds were sometimes to be seen growing
amongst the drift just above the high-tide level. The seeds of both C.
bonducella and C. bonduc have been found also on the shores of Jamaica.
Those of both species are not uncommon amongst the stranded drift of the
Fijian beaches; but notwithstanding a careful search I found only a
solitary seed of C. bonducella in the Hawaiian beach-drift, a
circumstance explained below as arising from the usual non-buoyancy of
Hawaiian seeds.

That the seeds of Cæsalpinia bonducella stranded on the coasts of an
oceanic island are able to germinate and reproduce the plant is, of
course, established by the distribution of the species; and we have just
observed that the process was noticed by the author on Keeling Atoll
where the plant has found a home. It is to be noted that the plant
collected by Darwin in this atoll was identified by Prof. Henslow as C.
bonduc; but the plant observed by me was more like C. bonducella, and
the stranded seeds collected by me were referred at Kew to this species.
Some curious considerations arise from the fact that although, just as
in the Keeling Islands, the plants of C. bonducella have evidently
established themselves from drift seed in one locality in the Bermudas,
they do not seem to have done so either on the shores of Krakatoa, or of
St. Helena, where, although they are frequently washed ashore, Mr.
Melliss never met with an instance of germination (see _Bot. Chall.
Exped._ iv, 300, and Penzig). This is doubtless in part the result of
the destructive efforts of the crabs, which, as I have shown in my paper
on Keeling Atoll, nibble off the shoots of many germinating seeds in
beach drift.

The readiness or non-readiness of seeds to germinate on a beach, and the
nature of the conditions essential for the process, are matters that are
directly concerned with their effective dispersal by currents. On
account of the stony character of the seeds of these two species, it
might be expected that germination would only take place under
exceptional conditions. It should, however, be observed that the fine
transverse striæ on their outer surface represent original fissures or
cracks in the epidermis of the soft immature seed; and as such may be
regarded as lines of weakness in the seed-tests. If a pod is opened
before the seeds are mature, we find the seeds about twice the size of
maturity, and so soft that they can be indented by the nail. The
transverse striæ that mark the mature seed are displayed as indistinct
cracks in the epidermis; and if the immature seed is exposed to the sun,
in a few hours these cracks gape widely, and the seed has the grooved
appearance of a top. If a pod opens prematurely on a plant, as sometimes
happens, the immature seeds will be noticed with the epidermis scaling
off. It is evident that the “setting” or the induration of the
seed-coats and the final great contraction of the seed take place in the
pod before dehiscence. From these remarks it would seem probable that
seeds lying exposed to the fierce rays of the sun on a tropical beach
would be liable to develop cracks along the old fissures, and that such
cracks by permitting the entrance of moisture would favour germination.

My experiments show that high temperature under moist conditions will
not of itself induce germination or in any way affect the seed. Thus in
two sets of experiments, in 1890 and 1902, I failed to induce the
germination of seeds which, after floating a year in sea-water, were
kept in moist soil at a high temperature. In one case a temperature
varying from 80° to 110° F. was sustained for several weeks, and in the
other experiment a temperature of 80° to 90° was kept up for five
months. When, however, an incision was made into the epidermis, or the
seed-coats were partially penetrated with a file, the seeds swelled up
in a day or two, and in a few days began to germinate.

The rapid transformation of the stone-like seed into a softened,
swollen, germinating mass ranks amongst the numerous little wonders of
the plant world. The seed, in fact, assumes again the appearance of
immaturity, and in so doing it suggests to us that the rest-stage
exemplified in the hard, pebble-like seed is but an adaptation to
general climatic conditions, and that in a region of great heat and
humidity, where there are no seasons, and where the sun’s rays are for
ever screened off by mist and cloud, it could be dispensed with
altogether. One of my Hawaiian dreams was to establish vivipary in
Cæsalpinia bonducella by subjecting the maturing pod on the plant to
very warm and humid conditions, my expectation being that the soft,
swollen seed would at once proceed to germinate in the pod, and that the
final process of setting, as indicated by the induration and contraction
of the coats, or in other words the rest-stage, would be done away with.
The dream, however, bore some fruits in enlarging my standpoint in the
matter of vivipary, and I have referred to the subject in Chapter XXXI.

The seed-shell, about 1·5 mm. in thickness, consists of three coats: the
outer skin very tough and waterproof; the inner skin seemingly
permeable; and the intermediate layer of hard prismatic tissue, the
“prismenschicht” of Schimper (p. 164). This middle layer absorbs water
rapidly and in large quantity, so that if a fragment of the shell is
placed in water it will be found after a day’s soaking to be three times
as thick as it was in the dry state. If one files a seed, or makes a
small incision, so as to expose the middle layer without piercing the
inner coat, and then places it in water, it will be noticed that the
middle layer at once begins to absorb water; and within a couple of days
the whole seed will swell and attain the size it possessed in the
so-called immature condition. During the process the outer skin
stretches, usually without rupturing; and all three coats, previously so
hard that a heavy blow with a hammer is required to break the seed,
become in a day or two soft enough to be easily cut with a knife. The
seeds thus treated swell in two days to three times their original size
and increase their weight fourfold. Water finds its way to the nucleus
or embryo partly through the dilated inner opening of the micropylar
passage and partly through the inner skin. The nucleus then swells up
into a fleshy mass, filling the seed-cavity, and in two or three days
more germination begins.

I pass now to the discussion of the buoyancy of the seeds. Considering
that both species occur in oceanic islands, and that the currents are
active agents in transporting the seeds, their behaviour under
experiment appears at first sight to be full of anomalies. Thus, it was
ascertained at Kew (_Bot. Chall. Exped._ iv., 301), both with
comparatively fresh and with older seeds, that those of Cæsalpinia
bonducella floated in salt water, whilst those of C. bonduc sank; but in
the record given of the experiment no mention is made of the original
station of the parent plants; and it will be shown later on that the
station of the plant, whether at the coast or inland, has an important
determining influence on the buoyancy.

In Fiji I found that almost without exception the seeds of littoral
plants of Cæsalpinia bonducella floated both in sea-water and in fresh
water. On the other hand, in Hawaii the seeds of this species, obtained
from three typical localities removed inland from the beach, sank
without exception, even after drying for several months; and the only
buoyant seed noted in these islands was a solitary seed collected from
the beach drift. In Hawaii, however, as before remarked, the species is
not strictly a littoral plant, occurring as it does in the lower levels,
but not necessarily in the vicinity of the coast. In the case of seeds
of littoral plants of C. bonduc in Fiji, I found that sometimes all
floated in sea-water and sometimes only a portion of them, whilst their
specific weight was on the whole rather greater than that of the seeds
of the other species. Thus, in one experiment half the seeds floated in
sea-water and a quarter in fresh water, whilst with seeds from another
locality 90 per cent. of the seeds floated in sea-water and 80 per cent.
in fresh-water; and in a third set of seeds all floated in both waters.

The above experiments on Fijian seeds all relate to littoral plants. In
the instance, however, of the inland species from the mountains of Vanua
Levu, all the seeds sank in sea-water, even after being kept for five
years. If we follow the indications of these several experiments we
shall find that Cæsalpinia presents another illustration of the general
principle established in Chapter II that the seeds of inland plants sink
and those of coast plants float.

My data, therefore, show that with the seeds of Cæsalpinia buoyancy goes
with station and not necessarily with species. It is probable,
therefore, that with the two widespread species, C. bonducella and C.
bonduc, varying results will be obtained with seeds from different
localities, whether insular or continental, according to the original
station. The typically buoyant seeds of the former species may, as we
have seen in Hawaii, lose their floating powers when they grow inland;
and the seeds of an inland species from the mountains of Fiji sink at
once. It is essential in interpreting the results of experiments on the
seeds of these plants to be acquainted with the stations; and in this
respect those of the Tahitian plants may be regarded as probable test
cases. We have seen that in Tahiti, C. bonduc is an inland plant, and C.
bonducella usually a beach plant; and I have no doubt that experiments
in that island on the seeds of these two species from the particular
stations just referred to will give results in agreement with the
principle here laid down.

With reference to the duration of the floating powers of these seeds it
may be observed that a seed of Cæsalpinia bonducella, originally found
stranded on the beaches of Keeling Atoll, floated after a year in
sea-water as buoyantly as at the commencement of the experiment. Seeds
of Fijian littoral plants of both C. bonducella and C. bonduc floated in
my experiments after two and a half years’ immersion in sea-water,
showing no change whatever. Some of the seeds removed at the end of the
first year were filed and placed in soil, when they germinated
healthily. In Chapter IX it is pointed out that some buoyant seeds of
other Leguminous plants, such as Mucuna urens, would be apt to germinate
abortively and to sink in crossing the more heated areas of tropical
seas. The seeds of Cæsalpinia, judging from my experiments and
observations noted on page 84, seem to be quite proof against such
risks. This was well brought out in an experiment where seeds of the two
species of Cæsalpinia were kept afloat for two and a half years in a
vessel of sea-water together with seeds of Mucuna and Strongylodon. None
of the Cæsalpinia seeds attempted to germinate in the sea-water; but
with the other genera some of the seeds began to germinate, and sank in
the course of the first warm season, when the water-temperature ranged
from 75 to 90° Fahr.

The seeds develop their buoyancy during the great contraction that, as
before described, marks the final setting of the seed-coats and the
ultimate maturation, as it is termed, of the seed. During this shrinking
process the kernel also shrinks within the seed-tests, and cavities are
thus produced within the seed-shell, on the relative size of which
depends the buoyancy of the seed, neither the seed-shell nor the kernel
possessing independent floating-power. These cavities, as illustrated in
the figures given in Chapter XII, are of two kinds. That usually
produced, being the one that mainly determines the buoyancy, is a large
central hollow caused by the arching outwards of the cotyledons during
the shrinking process, such as is found also in the seeds of Entada
scandens, Mucuna urens, and some other Leguminous littoral plants. With
such seeds the kernel never rattles when the seed is shaken, since the
cotyledons lie in close contact with the seed-shell. The other kind of
cavity is produced between the seed-shell and the kernel by the general
or partial shrinking of the kernel away from the shell, the cotyledons
remaining in apposition, as shown in the figures. When the shrinking
away from the shell is general, the kernel lies loose within the shell,
and the seed rattles when shaken. When the shrinking is partial the
cavity is on one side of the seed and the kernel is fixed.

Professor Schimper (p. 164) remarks that the buoyant seeds of Cæsalpinia
bonducella all rattle when shaken, and that it is to the incomplete
filling of the seed-cavity, thus indicated by the loose kernels, that
the buoyancy of the seed is due. The rattling of the kernel was,
however, quite exceptional in the seeds handled by me, even in the case
of originally buoyant seeds kept for five years. Seeds with loose
kernels were, in fact, more frequent with non-buoyant seeds than with
those that floated. Thus in Fiji I found that whilst with the buoyant
seeds 17 to 20 per cent. had loose kernels, with non-buoyant seeds the
proportion was as much as 60 per cent.

The normal cause of buoyancy is, therefore, a large intercotyledonary
cavity with the cotyledons lying in close contact with the seed-shell;
but the two kinds of cavity may sometimes be combined. Out of a number
of buoyant seeds of Cæsalpinia bonducella examined by me, 80 per cent.
owed their buoyancy solely to a large central cavity (4 to 5 mm.
across). In 6 per cent. it was due solely to the shrinking of the kernel
away from the seed-shell; whilst in 14 per cent. it was to be attributed
partly to a reduced central cavity (2 to 3 mm. wide), and partly to a
space outside the kernel. The only difference noted in the structure of
the buoyant seeds of C. bonduc was that the two kinds of cavities were
more often combined.

The reason of the absence of floating power was clearly indicated in the
non-buoyant Hawaiian seeds, where there was no central cavity, or it was
represented by a narrow slit. The solitary buoyant seed found in the
beach drift had a typical large central cavity. With the non-buoyant
seeds of the inland species of the mountains of Vanua Levu it was
ascertained that two-thirds had loose kernels with the cotyledons
closely appressed. In the others there was a lateral cavity outside the
kernel, the central cavity being only represented by a slit, a hair’s
width in breadth. In the non-buoyant seeds of C. bonduc, the central
cavity was only 2 to 3 mm. wide, and the lateral cavities were small.

Respecting the influence of “station” in producing the differences in
buoyancy, it cannot be said to be connected with the maturation of the
seeds of inland plants under more humid conditions than those which
prevail at the coast. In Fiji some of the littoral plants with buoyant
seeds grow on the mangrove-trees in the shade and humidity of the
swamps; whilst in Hawaii the inland plants of Cæsalpinia bonducella with
their non-buoyant seeds thrive in exposed arid situations in districts
of little rainfall, such as on scantily vegetated lava-flows. With
non-buoyant seeds, where there is little or no cavity, the cotyledons
are always thicker and moister than in the case of the seeds that float.
Though associated with differences in station, as implied in the terms
“coast” and “inland,” the cause of the difference in buoyancy is not
connected with different degrees of humidity, but with some other cause
or causes acting on the spot which, while they favour the drying of the
kernel in coast plants before the seed-coats finally set, impede it in
the inland plants. That the seed does not subsequently acquire floating
power, even after years of drying, was shown in several of my
experiments.

The light, unopened prickly pods of both species float buoyantly, even
when the inclosed seeds have no floating power. In an experiment on
Cæsalpinia bonduc in Fiji the pods remained afloat after a month in
sea-water. With those of C. bonducella in Hawaii I found that they
floated for several weeks, and in one case a pod was afloat after three
months. The pods dehisce on the plant; but they sometimes do not open
sufficiently to allow the seeds to fall out. The pods, however, have to
be torn off from the plant, and are not likely to occur in the drift.
Indeed, they never came under my notice in any locality in the drift,
and as an effective aid to dispersal they must be disregarded. The
buoyancy of the seeds and their well established distribution by
currents render unnecessary an appeal to the floating pod.

The following is a summary of the foregoing remarks on Cæsalpinia
bonducella and C. bonduc.

(1) The two species in Fiji are not always sharply distinguished, since
intermediate forms occur, and here probably lies the explanation of the
confusion that has sometimes occurred in diagnosing the species.

(2) Both are typical littoral plants, distributed over most of the
tropical zone, and occurring in company in most of the Pacific
archipelagoes; but they at times extend far inland.

(3) Though it is not unlikely that sea-birds may have aided in their
dispersal, the oceanic currents have been the great agencies in their
dispersal, as is indicated by the frequent transport of seeds in the
Gulf Stream drift across the Atlantic, and by their occurrence in beach
drift in various parts of the world.

(4) Having regard to the present arrangement of the currents and the
distribution of the two species, reasons are given for the belief that
their original birthplace was in the interior of the American continent.

(5) Notwithstanding the stony hardness of the seeds, when a notch is
made in the outer skin a seed rapidly takes up water, and in a few days
it becomes a soft and much swollen germinating mass. The author is
inclined to think that this was the original condition of the seed, and
that the rest-stage is an adaptation to secular differentiation of
climate in later epochs.

(6) Unlike the seeds of other Leguminous littoral plants, those of
Cæsalpinia are not likely to germinate abortively when floating in warm
tropical seas, a risk that restricts the distribution of several
littoral species.

(7) As tested by experiment, the seeds of both species are often able to
float unharmed for years; but on the other hand seeds not infrequently
have no floating power.

(8) Observation, however, shows that buoyancy goes with station, and
that the general rule here applies that the seeds of coast plants float
and those of inland plants sink.

(9) The nature of the influence of “station” on the seed-buoyancy is
obscure; but it is evidently not connected with the usual differences
between coast and inland localities, such as those concerned with
exposure or shade, dryness of soil, relative humidity, and similar
contrasts.

(10) The buoyancy of the seed is developed during the final shrinking
process associated with its maturation, a large cavity between the
cotyledons being usually produced.

--------------

_Note._—Since most of the principal conclusions of this work are
involved in my especial study of the littoral species of Afzelia,
Cæsalpinia, and Entada, the reader is advised, if he wishes to form an
opinion of the author’s method of investigation, to read this chapter
carefully through. With most other shore-plants, though in not a few
cases studied with the same detail, the exigencies of space have often
limited me to the employment of the general results in the appropriate
chapters without entering into details. Should he desire to test any
view of his own relating to plant-dispersal, he could not do better than
begin with the materials here provided.

CHAPTER XVIII

THE ENIGMAS OF THE LEGUMINOSÆ OF THE PACIFIC ISLANDS

IT is my intention here to gather up some of the “ends” of the great
tangle presented by the Leguminosæ in the Pacific. When we look at the
indigenous phanerogamic floras of Fiji, Samoa, Tahiti, and Hawaii we
find that the Leguminosæ form 5 or 6 per cent. of the total in each of
the three first-named groups, and only about 2·5 per cent. in Hawaii.
The paucity of Leguminosæ in oceanic floras was long ago pointed out by
Sir Joseph Hooker, whose work forms the foundation of much of our
knowledge of insular plant-life. This is emphasised by Mr. Hemsley in
his volume on the _Botany of the “Challenger” Expedition_ (Introd. p.
25), where he makes the very significant remark that the Leguminosæ are
wanting in a large number of oceanic islands where there is no truly
littoral flora. The islands, however, here more especially referred to,
are those of the southern Atlantic and Indian oceans, such as St.
Helena, Tristan da Cunha, and Amsterdam. It is especially true of New
Zealand, where the Leguminosæ barely make 2 per cent. of the total. Of
the Polynesian islands, as he points out, it is not so correct; and, in
fact, the proportion found in the Fijian, Samoan, and Tahitian floras,
respectively, is much the same as that which characterises the British
flora, namely, 5 to 6 per cent.

When we come to explain the paucity of the Leguminosæ in the Hawaiian
flora we bring to light the singular principle that _Leguminosæ are far
more characteristic of the littoral flora than of the inland flora of a
Pacific island._ About half of the Leguminosæ of Fiji and Tahiti are
coast plants; and about 30 per cent. of the littoral plants of the
islands of the tropical Pacific belong to this order. Since, therefore,
Hawaii possesses much fewer shore-plants (30) than does Tahiti (55) or
Fiji (80), the paucity of its Leguminous plants is readily accounted
for.

We have next to notice a principle, which is, in fact, deducible from
the first, namely, that _buoyant seeds are much more characteristic of
the Pacific Leguminosæ than of any other order_. Three-fourths of the
species have buoyant seeds, and, in fact, about a third of the littoral
Polynesian plants with buoyant seeds or fruits belong to this order.

It may, therefore, be inferred that _the Leguminosæ owe their presence
in the islands of the tropical Pacific mainly to the currents_.

From Mr. Hemsley’s conclusion that the Leguminosæ are wanting in a large
number of islands where there is no truly littoral flora, the
presumptions arise that _when inland species exist that possess no
capacity for dispersal by currents they are to be regarded as
derivatives from the littoral flora, and that they owe their origin to a
strand-plant possessing buoyant seeds originally brought by the
currents_. It has been shown in the case of Afzelia bijuga and of
Cæsalpinia that when Leguminous shore-plants extend inland the seeds
often lose their buoyancy, and it is probable that divergence in other
characters may occur, leading, as in the mountains of Fiji, to the
development of a new species of Cæsalpinia. It is urged that by a
continuation of the same process the inland species, Erythrina
monosperma, has been developed in Tahiti and Hawaii, and the inland
species, Canavalia galeata and Sophora chrysophylla, have been produced
in the last-named group. All these species have non-buoyant seeds, and
in all three cases there is no littoral species in Hawaii, it being
assumed that the parent strand-plant has been driven inland from the
beach. _It is not necessary that the littoral species should be now
represented in the flora._

It is remarkable that _in almost all cases the cause of buoyancy is of
the non-adaptive or mechanical kind, due either to cavities formed by
the shrinking of the seed-nucleus during the setting of the seed or to
the light specific weight of the kernel_. There is but little to show
that the buoyancy of the seeds of Leguminosæ is anything but an
adventitious character of the seed, as far as its relation to dispersal
by currents is concerned. Although this capacity has been the great
factor in the wide distribution of the species, yet it is evident that
Nature here takes advantage of a quality that could never by its aid
become a specific distinction. The upshot of the selecting process would
be the dispersal by the currents of nearly empty seeds or seeds that
have lost their germinating capacity.

The distribution of the Leguminosæ in the Pacific islands, and indeed of
tropical islands generally, is often full of inconsistencies. This is
the only order that sets at nought most of the principles established
for the other plants of the sea-coast, and that defies the application
of the laws of plant-dispersal now most in evidence. Take, for instance,
the inexplicable affinity of Acacia koa, the well-known Koa tree of the
Hawaiian forests, to Acacia heterophylla, a tree restricted to the
Mascarene islands of Mauritius and Bourbon. Mr. Bentham, who placed them
in the same group with three or four Australian species, even doubted
whether the difference between the Hawaiian and Mascarene species
amounted to specific rank. These two closely related Acacia trees of
far-separated islands of the Indian and Pacific Oceans represent
outliers of the great formation of phyllodineous Acacias that have their
home in Australia (_Introd. Chall. Bot._ p. 26). As far as I can gather
Acacia seeds have no known means of dispersal. Not even when the tree
has a littoral station, as in the case of Acacia laurifolia in Fiji,
have the seeds or pods any capacity worth speaking of for dispersal by
currents. We must appeal to the birds; but to what birds we may ask,
unless it be to the extinct Columbæ and their kin, or to the Megapodes.
Some of the other Hawaiian difficulties connected with the inland
Leguminosæ are repeated in the Mascarene Islands. Thus, Bourbon, like
Hawaii, has its inland species of Sophora of the section Edwardsia.

In their irregular distribution the Leguminosæ of the Pacific islands
are often a source of perplexity to the student of plant-dispersal.
Take, for example, the inland Erythrina, E. monosperma, of Hawaii,
Tahiti, and perhaps New Caledonia. Then look at the singular
distribution of the Sophoras of the Edwardsia section in Chile and Peru,
Hawaii, New Zealand, Further India, and Bourbon. The botanist, again,
finds a climber like Strongylodon in the forests of Fiji, Tahiti, and
Hawaii, and he picks up the seeds on the beaches of those islands and
notices that they float unharmed for many months in the sea, yet when he
pays heed to the distribution of the genus he finds that it only
comprises four or five species, and that it occurs outside the Pacific
only in the Philippines, Ceylon, and Madagascar. The extraordinary
distribution of Entada scandens in the Pacific islands has been before
alluded to in these pages. Here we have a plant, the seeds of which are
known to be transported unharmed by currents all round the tropics. Yet
it is absent from Hawaii and from almost all of the islands of Eastern
Polynesia. In many cases an endeavour has been made in this work to
explain these difficulties. But the order in the Pacific teems with such
difficulties. We may ask with astonishment why it is that the genera,
and sometimes even the separate species, of the Leguminosæ seem so often
to follow in each case a principle of their own.

Plants of this order in the Pacific conform to no one rule of dispersal
or distribution, whether we regard a species, a genus, or the whole
order. Take, for instance, the presence in Hawaii of Canavalia galeata,
a plant that, as we know it now, could not possibly have reached there
through the agency of the currents, and the absence from the same group
of Entada scandens that could have been readily transported there by the
currents from America. Or, if we take the whole order and look at the
structures connected with the buoyancy of the seeds, we find two types
of structure and the elements of a third. Then, again, whilst most
littoral plants with buoyant seeds retain the buoyancy of their seeds
when they extend inland, Leguminous shore-plants, like Afzelia bijuga
and Cæsalpinia bonducella, when they extend inland in Fiji and Hawaii,
lose in great part or entirely the floating power of their seeds.

Furthermore, most strand-plants, being typically xerophilous in
character, when they extend inland shun the forests and prefer the dry
soil and sparsely vegetated surface of the open plain; but the
Leguminous genera and species (Mucuna, Afzelia, Entada, &c.) when they
leave the coast take to the forests, growing usually as stout lianes,
but sometimes as tall trees. Here again the Leguminosæ seem to follow a
principle of their own. As far as I know, this is the only order in the
Pacific possessing forest-trees which, as in the case of Afzelia bijuga
in Fiji, are equally at home in the woods of the interior and of the
coast.

Indeed, judging from Professor Schimper’s observations, the littoral
Leguminosæ of the tropics often display a physiological constitution
that seems in some respects out of touch with their surroundings. They
may, as in Sophora tomentosa and in Canavalia, present the xerophytic
character of strand-plants, but frequently they are not halophilous or
“salt-loving,” like other plants associated with them on the same
shore-station. They are often shy of salt in their tissues, though able
to thrive in salt-rich localities. That capacity which strand-plants
usually possess of storing up chlorides in their tissues, and especially
in their leaves, without injury to themselves, is but slightly possessed
by such characteristic shore-plants as Canavalia, Pongamia glabra, and
Sophora tomentosa. This capacity, which, as Professor Schimper
indicates, goes to determine whether or not plants are capable of living
in salt-rich localities, has often no determining influence with the
Leguminosæ. (See Note 60.)

Though the plants of this order form such a large element in the
strand-flora of the Pacific islands and of the tropics generally, they
seem in other respects, besides those just referred to, to act as if
they were strangers to the station. Look, for instance, at the readiness
of the floating beans of Mucuna, Strongylodon, &c., to germinate, as
shown in Chapter IX, in the tepid waters of the warmer areas of the
tropical oceans. This is a great deal more than a disturbing factor of
distribution. It is significant also of the plants being out of touch
with their dispersing agencies.

One may notice in conclusion the fact brought out in Chapter VIII that
nearly all the littoral plants dispersed by the currents that are common
to the Old and the New Worlds belong to the Leguminosæ. This is held to
indicate that their home is in America, since that continent distributes
but does not receive tropical littoral plants dispersed by currents.

_Summary._

The Leguminosæ are far more characteristic of the littoral flora than of
the inland flora of the Pacific islands; and since the greater number of
them have buoyant seeds, it follows that this order mainly owes its
presence in this region to the currents.

As it has been shown that in a large number of islands where there is no
littoral flora the Leguminosæ are wanting, the presumption arises that
when, as in Hawaii, inland species occur which at present have no
capacity for dispersal by currents, they have been derived from
strand-plants originally brought by the currents, even though such shore
species no longer belong to the flora.

As far as its relation to dispersal by currents is concerned, the
buoyancy of the seeds of Leguminosæ is merely an adventitious character,
and the structure connected with it has no specific value.

Plants of this order in the Pacific are a source of much perplexity and
conform to no one rule of dispersal, whether as regards their
disconnected distribution, their means of dispersal, the structural
cause of buoyancy, the loss of buoyancy of inland species, and in other
particulars. Even in their physiological constitution they are often at
variance with the bulk of littoral plants when they grow on the
sea-shore, since typical beach-plants of the order, though thriving in
salt-rich localities, are shy of salt in their tissues.

It is probable that whilst the Pacific islands have derived most of
their littoral plants that are dispersed by currents from the tropics of
the Old World, they have received most of their strand Leguminosæ from
America.

CHAPTER XIX

THE INLAND PLANTS OF THE PACIFIC ISLANDS

PRELIMINARY COMPARISON OF THE PHYSICAL CONDITIONS OF HAWAII, FIJI, AND
TAHITI

INTRODUCTORY REMARKS.

I WILL carry my readers back to that moment when we began to investigate
together the composition of the floras of the islands of the tropical
Pacific from the standpoint of dispersal. It will be remembered that
after collecting all the fruits and seeds of a particular island we
placed them in sea-water, and that some nine-tenths of them went to the
bottom at once or in a few days. We found, speaking generally, that the
buoyant seeds and fruits belonged to coast plants, whilst those at the
bottom of the vessel proved to be obtained from inland plants. Since
that period we have been occupied in following up the clue supplied by
the floating seeds and fruits. In their company we have travelled far
beyond the Pacific islands. We have not only seen their fellows in other
parts of the tropics, both on the coral atoll and on the continental
coast, but we have met their representatives on the beaches of Europe
and of temperate South America. We have followed them in their ocean
traverses round most of the tropical zone, and on the way we have
naturally interested ourselves in the question of the currents. We have
weighed these seeds and fruits and have compared their specific weight
with that of sea-water. We have cut them up and carefully examined them,
and under their guidance we have explored the mangrove-swamps both of
Polynesia and of Ecuador, and have penetrated the mysterious _cul de
sac_ of vivipary. Having formed our opinion of them, we now bid the
subject farewell, and stand once more on the same Pacific beach where,
it seems so long ago, our investigations began.

For the seed and fruits lying at the bottom of the sea-water we have to
appeal to other agencies than to that of the currents if we wish to
inquire into their means of arriving at this island. In imagination we
leave the reef-lined shores for the interior, and exchange the
exhilarating surroundings of a coral beach, where “the sky is always
blue and the wind is always true,” for the arid conditions of an inland
plain, or for the humid conditions of the forest, where the rain is
incessant and the cloud-cap and mist seemingly eternal. When we look at
the motley collection of fruits and seeds obtained in such localities,
we are at a loss to know where to take up the clue. After vainly
endeavouring to obtain some inspiration as to the manner of commencing
the inquiry, we do what all good naturalists in the Pacific islands do
from force of habit when they meet with difficulties of any kind—we sit
down and light our pipes. Then come a flood of old memories and old
trains of thought that came to us years before on some mountain-top or
in a shady gorge or on some river-bank, in regions Pacific and
non-Pacific, and by degrees our ideas shape themselves and we begin to
think the matter over in an orderly fashion.

When the winds first brought the spores of ferns to this Pacific island,
the ocean currents brought the seeds and fruits of littoral plants, and
the birds transported the seeds and “stones” of various inland species.
All three agencies have been working side by side since the earliest
stage in its history. Yet it is only in the work of the wind and the
current that we find any indication of stability in the floral history
of the island. With the work of the bird it has been very different.
Since the first bird carried seeds to this locality all else has been
turmoil and change. Wave after wave of migrant plants has overrun the
interior of the island, and all have left their mark; but the great
distributing factor and disturbing agent has always been the bird.
Genera have been born and have disappeared, and in their place new
genera have arisen. Whole families even have participated in the
revolutions of the plant-world, and species have grown rankly in the
great confusion. Last of all came man with his cultivated plants and his
weeds, introducing new elements of change and discord into the island,
and often upsetting the floral economy altogether. The history of man’s
most troubled epoch would not be more full of catastrophes and great
events than the history of the plants of this Pacific island. Yet
through all these changes the winds and currents have been quietly
carrying on their work, bringing the same plants to beach and hillside
that they did before the age of unrest began.

The monotonous character of an island flora that has been supplied by
the winds and currents can be readily imagined. For their variety the
floras of the Pacific islands are mainly indebted to the bird, the great
disturber of the peace of the plant world. We cannot attach too much
importance to the contrast in the results produced by these several
agencies in stocking a Pacific island with its plants. On the one hand
we have the tranquil working through the ages of the winds and currents.
On the other hand there has been the revolutionary influence of the
bird. One cannot doubt that many of the species of flowering plants now
growing on the beach and many of the ferns on the upper mountain-slopes
have witnessed changes within the forest-zone of the island, such as an
antediluvian might record if he had lived through the ages to the
present time.

Now, what are these changes? How has the bird acted unconsciously such a
determining part? These are questions which I will endeavour in some way
to answer as one picks one’s path slowly through the various epochs in
the plant-history of these islands. We already are fairly well
acquainted with the beginnings of a flora either on a coral atoll or on
an ordinary tropical beach. What we have yet to learn is the subsequent
history of the flora. When Dr. Treub undertook, in 1886, his now
celebrated examination of the new flora of Krakatoa after the great
eruption, he commenced a series of observations which will no doubt be
prolonged into future centuries. Botanists a hundred and two hundred
years hence will complete a long chain of observations which will be
unique as a record of plant-colonisation; and science is deeply indebted
to Prof. Penzig for making, in 1897, the second examination of the new
flora. Though deprived of the valuable record that future generations
will possess for Krakatoa, we yet have at our disposal in the completed
process displayed by many a Pacific island a means of working backward
and in a sense completing the history.

In order to attack this problem I have mainly confined myself to the
Fijian, Tahitian, and Hawaiian floras, taking the three archipelagoes
just named as the centres of the regions in which they occur. These
three groups lie near the three angles of the triangular area of the
Pacific over which the various archipelagoes are scattered. They are
thus geographically well placed for an inquiry into the subject of
plant-dispersal over this ocean, and each of their floras has been
investigated by botanists of various nationalities—American, Austrian,
British, French, German, and Italian. The Fijian area may be regarded as
including the adjacent Samoan and Tongan groups, though the individual
group or the whole area will always be in this work particularised. In
the same way Tahiti will be viewed as usually representative of the
larger islands of the surrounding groups of the Cook and Austral Islands
and of the Marquesas; and under the designation of the Tahitian area or
Tahitian region there will be generally included the Paumotu
archipelago.

COMPARISON OF THE AREAS AND ALTITUDES OF HAWAII, FIJI, AND TAHITI.

Since differences in physical conditions have played an important part
in plant distribution in these groups—such, for instance, as in
determining the development of a mountain flora or in favouring the
relative abundance of particular types of plants—it is at first
essential to obtain a general idea, in the case of the larger islands of
each group, of their size and elevation, and of the more conspicuous
differences in their climates.

Hawaii, the largest island of the Hawaiian archipelago, has an area of
4,210 square miles. All the other islands of the group are considerably
smaller—Maui, the second in size, having a surface of 760 square miles;
Oahu coming next; and after it Kauai, with an area of 590 square miles.
The area of Viti Levu, the largest island of the Fijis, is 4,112 square
miles, being thus closely similar to that of the island of Hawaii; Vanua
Levu, the second in size, is 2,433 square miles in extent; whilst the
other important islands of the group are much smaller, Taviuni, the
third in size, having an area of 218, and Kandavu an area of 125 square
miles. Tahiti, the largest and loftiest island of Eastern Polynesia, has
a surface of about 400 square miles; whilst most of the other elevated
islands of the groups around are considerably smaller.

In respect of elevation above the sea, there is a great contrast between
the islands of these three regions. Taking the Hawaiian Group first, we
notice that the three principal mountains of the large island of Hawaii
rise in the cases of Mauna Kea and Mauna Loa to between 13,000 and
14,000 feet, and in that of Hualalai to rather over 8,000 feet. Situated
between these three mountains there is an extensive tableland or
plateau, known as the Cattle Plains, which is elevated between 4,000 and
6,000 feet, and has an area of not less than 200 square miles. At least
a third of the whole area of the island exceeds 4,000 feet in altitude.
In the eastern portion of Maui the huge mass of Haleakala rises to
rather over 10,000 feet; whilst Mount Eeka, in West Maui, rises in bulk
to some 6,000 feet. The island of Kauai, which is elevated between 5,000
and 6,000 feet, possesses in its interior an elevated tableland 40
square miles in extent and 4,000 feet in altitude. Oahu attains in Mount
Kaala a maximum elevation of 4,000 feet, but 3,000 feet is the limit of
the other peaks, and much of the island is low in elevation.

On the other hand, in the two largest islands of Fiji, namely, Viti Levu
and Vanua Levu, we find in the first-named only two or three of the
highest mountain peaks rising to between 4,000 and 5,000 feet; whilst
the highest peak of Vanua Levu reaches only to about 3,500 feet. Amongst
the lesser islands, Taviuni just reaches the level of 4,000 feet, and
Kandavu, the next in height, about 2,750 feet. The area of the
land-surface in this group that is above a level of 4,000 feet is very
scanty, and for the botanist a negligible quantity, so that for purposes
of comparison the Fijian Islands, as far as elevation is concerned,
correspond to the lower levels of the Hawaiian Islands, that is, to the
areas below 4,000 feet. The same may be said of the Samoan Islands with
the exception of a limited area in the centre of Savaii, where a peak
rises to 5,400 feet above the sea.

Coming to the Tahitian region, we find that Tahiti, the most elevated
island, attains an extreme height of about 7,300 feet; but from its
surface-configuration it is evident that not one-tenth of the area
exceeds 5,000 feet; yet since its total extent is about 400 square miles
there must be an elevated region of some 30 square miles in amount
comparable in some degree with the uplands of Hawaii. The Marquesas,
next in order in size and height, attain a maximum elevation of about
4,000 feet; whilst, amongst the Cook and Austral Groups, Rarotonga
reaches a height, according to Mr. Cheeseman, of 2,250 feet. Excepting
the limited elevated area of the uplands of Tahiti, there is nothing in
Eastern Polynesia corresponding to the higher levels of the Hawaiian
Islands over 4,000 feet. We formed the same conclusion for Fiji, and I
may add that it applies to the whole area of Fiji, Samoa, and Tonga,
since the solitary peak of Savaii in the second-named group, which
reaches 5,400 feet, alone represents a high-level area. The uplands of
Hawaii—that is to say, the elevated region between 4,000 or 5,000 feet
and 14,000 feet (strictly speaking 13,800 feet)—are therefore almost
unrepresented amongst the Oceanic groups of the South Pacific; and it is
only in the peak of Savaii and in the limited high levels of Tahiti that
we would expect to find their conditions reproduced. The great effect
that this contrast implies in determining differences between the floras
of the Hawaiian, Fijian, and Tahitian regions will become apparent as we
proceed in this discussion.

COMPARISON OF THE CLIMATES OF HAWAII, FIJI, AND TAHITI.

Before comparing the climatic conditions in the three groups, it may
first be remarked that since they lie, roughly speaking, at not very
dissimilar distances north and south of the equator a great contrast is
not to be expected in so far as they agree in elevation. The mean
latitudes do not differ greatly, that of Hawaii being 20° to 21° N., and
those of Fiji and Tahiti both about 18° S. The climate of both groups is
tempered by the north-east trade in the one region and by the south-east
trade in the other. Still there is a difference in the temperature and
dryness of the air which noticeably distinguishes Hawaii from Fiji, and
to a less extent from Tahiti. The mean temperature of the Hawaiian
Islands would be 74° or 75°; whilst that of Tahiti is placed at 76° to
77°, and that of Fiji at 79°. But it is to be observed that to a person
residing in Fiji after a residence in Hawaii the climate is perceptibly
warmer, more humid, and more enervating. No doubt this is in part
connected with the greater dryness of the air in Hawaii, where the
average relative humidity at Honolulu is placed at 72 per cent., and it
must be much less on the Kona coast on the dry side of the largest
island. It is, however, probable that the Hawaiian climate was less dry
before the destruction of the forests, and that the contrast with the
Fijian climate was then less pronounced.

The great distinguishing feature, however, of the Hawaiian Islands is to
be found in their mountain climate. This is not represented in Fiji, but
slightly in Samoa, and to a small extent in Tahiti; and I will now refer
more particularly to this important subject.

In the uplands of the large island of Hawaii, on the tops of the lofty
mountains 10,000 to 14,000 feet above the sea, we have a mean
temperature only found far north. Snow lies often on these barren
summits in winter, more particularly on Mauna Kea, which thus derives
its native name of the White Mountain. The details of my meteorological
observations on Mauna Loa will be found in Note 61; and only some of the
general results will be referred to here.

The mean temperature for the period of twenty-three days passed by me on
the summit of Mauna Loa (13,600 feet) between August 9th and 31st, 1897,
was 38·5° F. The mean temperature for a period of twenty days from
December 24th, 1840, to January 12th, 1841, during which Commodore
Wilkes and his party were making pendulum observations on the summit of
the same mountain, was approximately 33·5° (see Note 61). From these
results, which are tabulated below, it will be seen that the mean annual
temperature would be probably about 36°, which is scarcely comparable
with any continental climate, since only a difference of a few degrees
is indicated between the mean temperatures of August and of a similar
period in mid-winter. I may add that although it was in the summer month
of August, water froze inside my tent during twenty out of the
twenty-three nights passed on the top. We may, therefore, infer that the
temperature falls below the freezing point at night practically
throughout the year. It will be seen from the table that the mean annual
temperature for the summit of Mauna Loa, as here computed from the
observations of Commodore Wilkes and myself, comes very near to that
which might be estimated by employing Hann’s tables of variation in
temperature with altitude on tropical mountains (see Schimper’s
_Plant-Geography_, iv. 691).

WINTER AND SUMMER TEMPERATURES ON THE SUMMIT OF MAUNA LOA
(13,600 FEET), IN DEGREES FAHRENHEIT.

+---------+---------------+--------------+-------+--------+--------+------------+
|Observer.| Period. | Mean daily |Lowest.|Highest.|Mean for|Approximate |
| | | range. | | |period. |yearly mean.|
+---------+---------------+--------------+-------+--------+--------+------------+
|Wilkes |Dec. 24, 1840- |17°-50°=33° | 13° | 55°? | 33·5° | } |
| |Jan. 12, 1841 | | | | | } |
| | | | | | | } 36° |
|Guppy |Aug. 9-31, 1897|23·2-53·8=30·6| 15 | 61·2 | 38·5 | } |
+---------+---------------+--------------+-------+--------+--------+------------+

Estimated mean annual temperature of the summit of Mauna Loa, taking
that of the coast at 75°, would be 34° if the rate of increase was the
same as on Mount Pangerango in Java (1° per 328 feet).

The great daily range of temperature is one of the most striking
features of the climate of the summit of Mauna Loa. The extreme recorded
by me was 38·7°, whilst Wilkes registered as much as 42°. As on most
lofty mountains the dryness of the air, as indicated by the relative
humidity, was usually great. The average percentage during my stay
between 8 and 9 A.M. was 44, at midday 43, and between 5 and 6 P.M. 56.
This may be contrasted with 72, the average for the year at Honolulu. In
the tropics the mean for the year in the lower levels often rises to 80
and over; and it can scarcely be doubted that the Hawaiian climate is
generally drier than it was before the destruction of the forests. The
lowest relative humidity recorded by me on the summit of Mauna Loa was
20 per cent. Junghuhn on the summits of two mountains in Java, 10,500
and 11,500 feet in height, recorded percentages as low as 5 and 13.
Further details relating to this subject are given in Note 61. The
rainfall on the top of Mauna Loa is probably very slight. During my
sojourn rain was noted on six days, but on only two could it be
measured, and the total fall could not have amounted to over a third of
an inch.

The mean annual temperature of the great forest-zone at the elevations
where it displays the greatest luxuriance of growth on the island of
Hawaii, that is, between 4,000 and 6,000 feet, would be estimated at 63°
and 57° F., if we take the rate of decrease before employed of about
three degrees per 1,000 feet. But remembering the heavy rainfall in this
region and the usual occurrence of a protecting belt of cloud during the
day, this might seem to be too high. According, however, to a table
given by Mr. Jared G. Smith in his annual report of the Hawaii
Agricultural Experiment Station for 1902, the average temperature at
4,000 feet would be 65°. I cannot help thinking this is excessive as an
average for the island. In the latter part of May, 1897, the mean
temperature during my sojourn of twelve days at elevations between 6,000
and 6,700 feet around the slopes of Mauna Kea was 51·2°; whilst for
eight days in the first part of June in the same region the mean
temperature was 58·2° at an altitude of 4,000 to 4,300 feet.

It is possible, as I have pointed out on a later page, to recognise in
the different zones of vegetation the floras of a variety of latitudes;
and these zones are to a large extent controlled by temperature as well
as by other conditions. Thus the Fijian would be amongst familiar
vegetation on the lower slopes of Mauna Kea, whilst the Maori would be
at home halfway up the mountain-slopes, and the African from the upper
forests of Kilima Njaro and Ruwenzori would find in the higher levels
much to remind him of his native land.

The upper woods extend usually to 8,000 or 9,000 feet above the sea, and
vegetation of a scrubby character occurs as high generally as 10,000 or
11,000 feet. The highest regions present only a barren rocky waste.

THE RAINFALL.

_The Hawaiian Islands._—Although on account of the extensive deforesting
of the Hawaiian Islands since their discovery the contrast between this
group and that of Fiji is now, as regards rainfall, somewhat emphasised,
it is almost certain that in early times the contrast was much less
marked. In the lower levels the natives and sandal wood traders in the
past, and the agriculturists in the present, have accomplished much in
this direction. Between 1,000 and 3,000 feet, whole forests were in my
time disappearing under fire and axe for the coffee plantations. Above
those levels up to the higher limits of the woods, cattle were
destroying the forests in a wholesale fashion; whilst foreign insects
were proving themselves almost as great enemies to the vegetation. I
remember an enterprising agriculturist explaining to me how he cleared
the land of forest around his station. A large tract having been fenced
in, the cattle were introduced. After destroying the undergrowth and the
young trees, the animals attacked the bark of the trees, and in a year
or two, without fire or axe, the land was cleared. The consequence of
this unchecked destruction of the forests was in my time becoming only
too evident. When I passed through Ookala, on the Hamakua coast, at the
end of May, 1897, there was a water famine. Water was sold at a quarter
of a dollar a bucket, and the allowance for a family was three oil-cans
a week. Stealing water was a crime and punished by the plantation
authorities by dismissal or a five-dollar fine.

If we could look back for fifty or sixty years—I am now quoting from the
reports of Prof. Koebele and Dr. Stubbs—we should see large forests
where we now see barren slopes and plains. Originally forests covered
the upland plateaux and mountain slopes of all the islands. Now much of
the original forests has been removed, and large areas of naked soils
and bare rocks remain. The present forest area, writes Mr. Giffard, the
editor of the _Hawaiian Forester_ (August, 1904), is about 20 per cent.
of the islands, a small fraction of what it was a hundred years ago. It
is, however, very satisfactory to learn that American energy is now
combating this evil. Already in the January number of the same journal
is to be found a report by Mr. W. L. Hall, of the Bureau of Forestry, on
“The Forests of Hawaii”; and now, under the charge of Mr. Jared G.
Smith, institutions have been formed and experiment stations have been
established for “the intelligent and skilful cultivation of the soil.”
Hawaii owes much to the United States Department of Agriculture. May we
in England take the cue in the case of our own Crown colonies!

Under these circumstances the comparison of the present rainfall of
Hawaii must be carried out with discrimination. But it may be at once
observed that to make a contrast in detail between the rainfalls of
these three groups is quite beyond the province of this work; and this
remark applies also to the other observations on the climatic
conditions. I can only treat the subject in an illustrative fashion in
connection with the general subject of their floras.

Thanks to Professor Lyons, the Government meteorologist, the rainfall
has long been systematically investigated. It may be said to range
anywhere between 10 and 300 inches. As in most groups within the
trade-wind belts, there is a great contrast in the rainfall between the
weather and leeward sides of the islands, which is well exhibited in the
large island of Hawaii. Whilst in the Hilo district on the wet side of
the island the annual rainfall near the coast is about 120 inches, on
the Kona coast of the dry side of the island it may be anything between
20 and 50 inches and it may fall to less than 10. The effect of
elevation is, however, evident on both the weather and lee sides of the
island. Thus at a height of 1,650 feet in the Hilo district it is as
much as 180 inches, and at a greater elevation 210 inches. At a height
of about 1,600 feet at Kealakekua, on the dry side of the island the
average yearly rainfall, according to the results kindly supplied to me
by the Rev. S. H. Davis, was for the six years, 1891-6, 60 inches. On
the beach, as he says, it is “very much less,” probably not 30 inches.
Dr. Maxwell, in his report on “Irrigation in Hawaii,” mentions a
locality in Maui where the rainfall at the sea-shore was 28 inches, and
at a height of 2,800 feet up the mountain side as much as 179 inches. In
the region of the cloud-belt, which coincides with that of the
forest-zone on the slopes of the great mountains of Hawaii and extends
up from about 3,000 to 7,000 or 8,000 feet above the sea, the average
annual rainfall would probably be rarely under 200 inches, and in some
localities it might approach 300 inches. There are some particularly wet
mountains, and amongst these may be placed the high table-land of Kauai
(4,000 feet) and the flat summit of Mount Eeka (6,000 feet) in West
Maui. Here in a region almost of eternal mist we have developed a
special bog-flora.

Hillebrand describes the flat top of Mount Eeka as “wrapt in a cloud of
mist nearly the whole year.” Whilst descending this mountain I was
overtaken by the darkness at a little under 5,000 feet above the sea.
Through the night there was a continuous soft rain, or rather a heavy
wet mist, and I passed it under conditions suggestive of living in a
sponge. Everything was reeking with moisture. The air was saturated with
it, and water dripped from every leaf and branch, whilst the ground on
which I stood was soft and yielding and soaked with water like a sponge.
The surface was cut up by numerous narrow water-channels ten to twenty
feet deep and only a couple of feet wide, their very existence almost
concealed by ferns, whilst torrents rushed along at the bottom and kept
up a strange music through the night. This was the longest night I have
ever experienced, as my standing-ground was very limited, and with a
water-channel a foot or two away on either side I had to keep on my legs
until the dawn.

Above the cloud-belt, at elevations of 10,000 feet and over, the
rainfall is evidently very small. I have before remarked that during my
stay of twenty-three days (August 9-31) on the summit of Mauna Loa
(13,600 feet) the rain did not exceed one-third of an inch in amount. I
have by my side the report to the Weather Bureau, compiled by Prof.
Lyons, on the rainfall of this large island of Hawaii for the entire
month (August, 1897); and it enables one to make a comparison, in some
respects unique, of the distribution of the August rainfall on Mauna
Loa, from its base to its summit, where it occupies the breadth of the
island. Whilst on the east or wet side from the coast up to 1,500 feet
amounts ranging from 11 to 15 inches were measured, on the west or dry
side between one and two inches were registered at the coast, and 10
inches at Kealakekua, about 1,600 feet above the sea. But the level of
maximum precipitation would lie much further up the mountain slopes on
either side, probably at an altitude of 4,000 or 5,000 feet, and here
the rainfall for the month could not have been less in either case than
20 inches. Above this line of greatest rainfall the amount of
atmospheric precipitation would become less and less until beyond the
upper forest zone above 10,000 feet to the summit (13,600 feet) the
quantity would be very small; and judging from my observations, that
covered three-fourths of the month, the rainfall on the top of the
mountain for August would not have far exceeded half an inch.

The dry climate of the summits of Mauna Kea and Mauna Loa is reproduced
on the tops of the Java mountains and on the summits of the Owen Stanley
Range in New Guinea. Sir W. Macgregor found a fine and dry climate on
the top of the mountains last named, beyond the limits of the forests,
which extend to 12,000 feet above the sea. Below lay the cloud belt, a
zone of moss and fog, where at an elevation of 7,000 to 8,000 feet
everything was reeking with moisture (_Journ. Roy. Geogr. Soc._ 1890).
Observers at the coast often little imagine, when looking at a
cloud-concealed mountain peak, that although the cloud-belt from below
looks black and lowering and rain is falling heavily in the gloomy
forests, there is on the upper side a region of bright sunshine, and
that the peak stands out, unseen by them, above a sea of clouds
sparkling brilliantly in the sun and dazzling in their whiteness. It
will be seen from the table given in Note 61, that during my sojourn on
the summit of Mauna Loa the sky was cloudless or almost free from cloud
during nearly half the time. The mean cloudiness in the forenoon for
twenty-two days was 1·3 and for the afternoon 3·5, whilst the nights
were cloudless.

_The Rainfall of Fiji._—The rainfall of Fiji is known to be very large.
In illustration I will take Vanua Levu, the second largest island,
partly because of my familiar acquaintance with it, and partly because I
have at my disposal measurements for both the lee and weather sides of
the island—the first dry and characterised by a scanty and peculiar
vegetation, the second humid and densely forested. At Davutu, near the
sea-level on the weather or wet side of the island, the average yearly
fall for a period of sixteen years up to 1898 was 160 inches (these
observations were made in the grounds of the manager’s house and I am
indebted to Mr. Barratt for allowing me to inspect them). The
mountainous backbone of the island, which has an elevation ranging
usually from 2,000 to 3,000 feet, is generally in the rain-clouds.
During the months I was occupied in examining the geology of these
mountains, it was a common experience to be drenched to the skin all day
long, and I cannot doubt that the annual rainfall in the higher levels
must often reach 300 inches. Those familiar with the “sun-burnt” lands
or “talasinga” plains that mainly form the north or lee side of the
island, would expect a great difference in rainfall as compared with the
south or weather side. There is a marked difference, it is true, but it
is far less than we might have looked for. At Delanasau on the north
coast, less than a hundred feet above the sea, the mean rainfall for
seven years (1871-77), according to the observations of Mr Holmes, was
113 inches, and the range 80 to 159 inches (see Horne’s _Year in Fiji_).
In discussing the origin of the arid-looking plains on the north or lee
side of the island in Note 22, I have shown that the explanation is to
be found not so much in the rainfall as in the dryness of the air as
indicated by the relative humidity.

The rainfall varies greatly in and around Vanua Levu, but there is
little doubt that by far the greatest bulk of the rain is precipitated
on the upper weather slopes of the mountainous backbone of the island.
Taviuni, which lies off its weather coast, is probably the wettest among
the smaller islands of the group. In 1877, when 80 inches were recorded
by Mr. Holmes at Delanasau on the north side of Vanua Levu and 73 inches
at Levuka in the island of Ovalau, 251 inches were measured in Taviuni
at Ngara Walu 564 feet above the sea; and in 1875 the rainfall recorded
at Taviuni was 212 inches, and at Delanasau 126 inches (Horne).

Fortunately, the Fijian islands have not been long enough occupied by
the whites to produce much effect on the rainfall through the
destruction of the forests. A significant warning, however, has been
given in the vicinity of Levuka. The woods of the hills around the town,
as we learn from Mr. Horne, were cut down to prevent them from affording
shelter to the unfriendly natives of the interior, the result being to
reduce the number of rainy days in a few years from 256 to 149 per
annum.

_The Tahitian rainfall._—The annual rainfall of the coast districts of
Tahiti is placed at about 50 inches (_Encycl. Brit._ vol. 23); but, as
is observed by Nadeaud and Drake del Castillo, the rain-clouds gather
round the peaks, and the precipitation is much greater in the interior
than at the “littoral,” with a corresponding result in a striking
difference between the vegetation of the two regions. Probably,
therefore, the rainfall for the year on the wooded mountain slopes and
at the heads of valleys where the vegetation is most luxuriant would be
over 100, and perhaps as much as 150 inches in places. (The annual
rainfall in Rarotonga is, according to Cheeseman, about 90 inches.)

It is evident that in the three groups of Hawaii, Fiji, and Tahiti, the
rainfall varies greatly with situation and with elevation; but the
contrast is much greater in Hawaii than in Fiji. Thus there would be
scarcely any place on the lee side of Vanua Levu where the average
annual fall would be less than 80 or 90 inches, except perhaps in the
Undu Promontory, whilst on the lava-bound coast of the west or lee side
of Hawaii, it may be reduced to 20 inches and less. There is no doubt
that this was to some extent the case in pre-European times, since Fiji
must have possessed for ages, on the northern sides of the larger
islands, its arid “talasinga” or “sun-burnt” plains; and in the island
of Hawaii there must have always been vast, scantily vegetated lava
fields at the sea-border. It is probable, however, that it is in the
older islands of the Hawaiian group, those where the volcanic forces
have been long extinct, that the rainfall has been chiefly affected by
deforestation. Speaking generally, in pre-European times the climatic
conditions of the lower levels of the group, that is below 4,000 feet,
which are alone comparable with Fiji, were less contrasted with the
climatic conditions of the Fijian islands than they are at present. By
reason of their great elevation, the Hawaiian islands present a mountain
climate not found in Fiji, and scantily represented in Tahiti. It is,
therefore, in the flora of the Hawaiian uplands that we should expect to
find the great distinguishing feature between that group and Fiji.

_Summary of the Chapter._

(1) Whilst the winds and the currents have been working tranquilly
through the ages, bringing always the same vascular cryptogams and
shore-plants to the Pacific islands, the bird has ever been a disturbing
factor in the inland flora, and changes often of a revolutionary
character have taken place from time to time within the forest-zone.

(2) In the discussion of the inland plants of these islands, the Fijian,
Tahitian and Hawaiian areas are taken as centres of development and
dispersal, and as including the groups around.

(3) On account of the contrast in physical conditions presented by these
archipelagoes, differences with which some of the most distinctive
features of the floras are to be connected, a comparison of the islands
from this standpoint is first necessary.

(4) Since the largest islands of the Fijian and Hawaiian areas are from
five to ten times the size of Tahiti, the largest island of the Tahitian
region, we would expect to find in the two first-named groups a much
more varied flora.

(5) There are three huge mountain-masses in the Hawaiian group which
rise to between 10,000 and 14,000 feet, and there is in the aggregate a
large area elevated more than 4,000 feet above the sea. These elevated
regions are almost unrepresented in the southern groups, the Fijian
islands being only comparable with the lower levels of the Hawaiian
islands below 4,000 feet, and the same is true of all the groups with
the exception of a limited area in Tahiti, where the mountains reach a
height of 7,300 feet, and of the solitary peak of Savaii in Samoa, which
attains an altitude of 5,400 feet. Thus the conditions for a high-level
or mountain flora which exist in Hawaii are not to be found in Fiji, but
slightly in Samoa, and to a limited extent in Tahiti.

(6) From their position with regard to the equator and with reference to
the trade-winds a great contrast between the climates of these three
regions—the Fijian, the Tahitian, and the Hawaiian—is, as far as the
islands agree in elevation, not to be expected, and in fact does not
exist. The Fijian climate, however, is now warmer and more humid, and
the general rainfall is greater than in the case of Hawaii, but it is
probable that these differences were much less pronounced before the
destruction of the Hawaiian forests, which has been in progress since
the discovery of the group.

(7) Anywhere around the coasts of the larger Fijian islands we might
expect an annual rainfall of not less than 80 or 100 inches. In the
Hawaiian group the rainfall at the coast may be anything between 10 and
100 inches, but is generally less than 50 inches. In Tahiti, at the
coast, it is 50 inches. In all cases the rainfall increases greatly with
elevation. In the Fijian mountains the rainfall probably varies between
200 and 300 inches. In the Hawaiian forest-zone it would range probably
between 100 and 200 inches, though this is probably exceeded in a few
localities. In the Tahitian uplands it would doubtless exceed 100 inches
and approach 150 inches.

(8) Quite a different climate prevails on the lofty summits of Hawaii
13,000 to 14,000 feet above the sea. Here the snow lies in winter, and
the mean annual temperature is only a few degrees above the freezing
point, probably about 36° F. The difference between the mean summer and
winter temperatures is very small, and does not exceed five or six
degrees. Water freezes here during nearly every night of the year. The
daily variation of temperature is very large, the average being probably
about thirty degrees. Great dryness of the air prevails, the average
relative humidity in August, 1897, being about 43 per cent. There is but
little rain. The sun shines fiercely, and the sky is usually clear.

(9) All Pacific climates are represented in the Hawaiian mountains, that
of Fiji on the lower slopes, that of New Zealand half way up, and that
of the Antarctic islands on the summits.

(10) When contrasting the floras of Fiji, Tahiti, and Hawaii, it will be
necessary to restrict our comparison in the case of Hawaii to the lower
slopes below 4,000 or 5,000 feet; and we should expect the Hawaiian
mountain flora to be scantily represented in Tahiti, and scarcely at all
in Fiji and Samoa.

CHAPTER XX

THE EPOCHS IN THE FLORAL HISTORY OF THE PACIFIC ISLANDS

THE AGE OF FERNS

The epochs in the plant-stocking.—The age of ferns and lycopods.—The
relative proportion of vascular cryptogams in Hawaii, Fiji, and
Tahiti.—The large number of peculiar species in Hawaii.—The mountain
ferns of Hawaii.—The origin of peculiar species.—Dr. Hillebrand’s
views.—Their origin connected not with greater variety of climate in
Hawaii, but with isolation.—Summary.

_Introductory Remarks_

IN the endeavour to follow the various stages in the floral history of
the Pacific islands from the standpoint of plant-dispersal, a method is
here adopted which is not often employed. The usual mode of making a
general description of a flora is not intended to bring out its genesis
in point of time. We describe the result of a long series of changes
dating back to some unknown period, much as one might describe the
present condition of a people without reference to their history; and
for obvious reasons rarely is an effort made to differentiate the epochs
of the stocking of the region with its plants. The difficulties
investing such a task in the case of a region situated within a
continental area would be almost insuperable. With the oceanic groups of
the Pacific such difficulties, though still very numerous, would at all
events be fewer in number and less formidable in appearance.

Taking my cue from the well-known instance of Krakatoa, it is here
assumed that the earliest epoch is connected with the arrival of the
cryptogamic flora (ferns, mosses, lichens, &c.) through the agency of
the winds, and with the arrival of the littoral plants through the
agency of the currents. The next era is represented by the genera now
peculiar to each group, since it is implied that they have descended
from the earliest phanerogams that established themselves in the group.
The following epoch, which ends only with the arrival of man, is
characterised by the genera found outside the group; and here different
degrees of antiquity are indicated according as the genus is represented
wholly or in part by peculiar species, or contains only species found in
other regions. The following eight chapters will be devoted to the
development of the method here briefly indicated.

THE AGE OF FERNS.

It was established by Dr. Treub in the case of Krakatoa that ferns and
algæ formed the earliest vegetation of this island after it had been
completely stripped of all its plants in the great eruption of 1883. It
is, therefore, but natural that the vascular cryptogams should first be
dealt with in any discussion relating to the historical aspects of these
floras.

It has been before remarked that the epoch of ferns and lycopods, which
began with the earliest stage in the island’s floral history, may be
regarded as extending to our own day. It is thus implied that the
vascular cryptogams of those early times are yet brought there, and
that, alike with the littoral plants, these ferns and lycopods have
witnessed almost unchanged the great revolutions that have marked the
history of the inland flowering plants, more particularly those of the
forest flora. This, as I will show, is true in Hawaii, though only in a
partial sense in comparison with the other island-groups of Fiji and
Tahiti, since in Hawaii nearly half the ferns and lycopods are peculiar
to that group, whilst in Fiji and Tahiti not more than 8 or 9 per cent.
appear to be endemic. (Rarotonga, according to Cheeseman, possesses one
new species amongst its seventy-two ferns and lycopods, and probably in
this it is typical of the smaller elevated islands of Eastern
Polynesia.)

The large proportion of peculiar Hawaiian species is the central fact in
the distribution of vascular cryptogams in the Hawaiian, Fijian, and
Tahitian archipelagoes, and indeed in the Pacific islands; and it is
around this fact that much of the following discussion will lie. (For
the data relating to the Tahitian region, I have almost exclusively
followed Drake del Castillo.)

On looking at the table given below, it will be noticed that whilst
there are about the same number of species of ferns and lycopods in the
Tahitian and Hawaiian islands there are at least half as many again in
Fiji. When we reflect that the total areas of the Fijian and Hawaiian
groups are in each case about 7,000 square miles and that the extent of
the whole Tahitian region does not amount to 2,000 square miles, these
facts acquire a fresh significance. Ferns and lycopods might, therefore,
be expected to figure more largely in the Tahitian flora than in those
of Fiji and Hawaii; and this is indeed the case. When we examine the
relative proportion of the vascular cryptogams to the indigenous
flowering plants in each area we find that whilst in Hawaii they form
about 18 per cent. of the total flora and in Fiji not much more than
this (see Note 62), in Tahiti they constitute just a third. This excess
of vascular cryptogams is reflected in the flora of the outlying groups,
the proportion in Rarotonga being, according to Cheeseman, 30 per cent.
It is, therefore, evident that in comparison with the other groups
Tahiti possesses a marked preponderance in ferns and lycopods. In this
respect the Tahitian islands resemble those of Juan Fernandez, where
judging from the data relating to the indigenous flora given in
Hemsley’s _Botany of the Challenger Expedition_, the proportion of
vascular cryptogams amounts to between 30 and 38 per cent.

But it has been already implied that the proportion of endemic species
of ferns and lycopods is from four to five times as large in Hawaii as
it is in Tahiti or Fiji. In Hawaii, therefore, there has been a
production of many new species, whilst in Fiji and Tahiti there has been
a great rush of immigrants. “Formative energy” in Hawaii (to adopt an
expression of Dr. Hillebrand) and “active colonisation” in Fiji and
Tahiti, such would appear to be the most conspicuous features in the
history of the vascular cryptogams of these three archipelagoes.

In these floras it is, therefore, apparent that respecting the vascular
cryptogams the average number of species in a genus does not supply a
means of contrasting them. As indicated in the table, the fern and
lycopod floras of Fiji and Hawaii are similar in this respect. Yet in
each the average number of species to a genus has a separate
significance. A genus may acquire its species through immigration, or
they may arise from its formative energy within the particular area. The
first principle has been largely dominant in Fiji, the last in Hawaii,
and the resemblance between the average number of species in a genus in
these two groups is to a large extent accidental. Between the vascular
cryptogams of Fiji and Tahiti, however, such a comparison is legitimate;
and since the average formative energy is in these groups about the
same, the difference is to be attributed to a lessened number of
immigrants into the Tahitian area.

TABLE OF VASCULAR CRYPTOGAMS (FERNS AND LYCOPODS) IN THE GROUPS OF
TAHITI, HAWAII, AND FIJI. (See note 63.)

+--------------------+------+------+----+
| Group. |Tahiti|Hawaii|Fiji|
+--------------------+------+------+----+
| Number of genera. | 38 | 29 | 40 |
+--------------------+------+------+----+
| Number of species. | 154 | 155 |237 |
+--------------------+------+------+----+
|Species to a genus. | 4·1 | 5·4 |5·9 |
+--------------------+------+------+----+
| Number of | 13 | 70 | 20 |
| endemic species. | | | |
+--------------------+------+------+----+
| Percentage of | 8 | 45 | 8 |
| endemic species. | | | |
+--------------------+------+------+----+
|Percentage of ferns | | | |
| and lycopods among | 33 | 18 | 21 |
|the vascular plants.| | | |
+--------------------+------+------+----+

The results, so far mentioned, are in the main consistent with the
geographical position and the degree of isolation of these three areas.
From their proximity to the large continental islands of the Western
Pacific, the Fijian islands would have readily received a great number
of immigrants from the west, since the intervening sea is not over 500
miles in breadth. They lie in the track of the main line of migration
into and across the South Pacific, a track which has been followed by
flowering plants and animals as well as by aboriginal man. Assuming that
the migration of the vascular cryptogams extended from Fiji eastward to
Tahiti, fewer of the immigrants would reach the last-named group. Fewer
still would reach the Hawaiian islands, which excluding the groups of
low coral islands to the southward are cut off on all sides, whether
from the Fiji-Samoan and Tahitian areas, from the coasts of North
America, or from the regions north and west, by a breadth of ocean that
is never less than 1,500 miles.

That the main track of the ferns and lycopods across the South Pacific
to Tahiti has been eastward there can be little doubt. This is indicated
in the tables given by Drake del Castillo for Eastern Polynesia, and
also by an analysis I have prepared of the distributions that he gives
for the species of the Tahitian region (see Note 64). Out of the 154
species there are only two that belong exclusively to the American side
of the Pacific; whilst 58 are derived exclusively from the Asiatic side,
and mainly from Indo-Malaya. The drift of the ferns and lycopods
eastward from Fiji is also brought out in the number of Tahitian species
common to Hawaii and Fiji. Of these about 76 per cent. are common to
Fiji or to the groups around, and only 30 per cent. occur in Hawaii. The
Tahitian species found in Hawaii occur also in Fiji with the exception
of two or three mountain species which have doubtless failed to find a
suitable elevation in Fiji. These two or three mountain ferns and
lycopods are probably the only vascular cryptogams possessed in common
by Hawaii and Tahiti to the exclusion of other groups. (See Note 64.)

The prevailing Indo-Malayan origin of the ferns and lycopods of the
archipelagoes of the Fijian area (Fiji, Tonga, Samoa) is so well
established in the writings of Seemann, Baker, Hemsley, Christ, and
Burkill that there is no necessity to enter into details here. That the
stream of vascular cryptogams to Hawaii has proceeded mainly from the
Old World side of the Pacific is shown in the circumstance that of the
eighty and odd species found outside the group nearly half are from the
Asiatic side exclusively and only three from America alone, whilst about
a fourth occur in both continents, and a fourth are confined to
Polynesia. One point, says Dr. Hillebrand, comes out in strong relief,
and that is “the great number of ferns scattered over the long track
which leads from the Hawaiian islands through Polynesia and Malaysia to
the east coast of tropical Africa.” But he adds significantly that “it
cannot be inferred from this fact that all the species in question have
travelled eastward to find the terminus of their long migration on this
group, unless the principle be established, that the formative energy of
a species or genus be greatest at the circumference or farthest
extremity of its area” (p. 542).

Though evidently prepared to admit the general eastward trend of plants
in the Pacific, Dr. Hillebrand (p. xxviii) puts forward in the case of
the ferns the startling view that originally spores of a few simple
species have been diffused over various countries and that they have
there evolved on parallel lines “predetermined by the structure of the
original immigrant” a series of higher forms, so that the same form
might have been produced in two widely distant localities, as, for
instance, in Ceylon and Hawaii. The editor, Mr. W. F. Hillebrand, gives
good reasons for his belief that this does not represent the matured
opinion of the author. It is, however, worth noting in this connection
that Dr. Karl Mueller has advanced a similar view with respect to the
lower orders of plants. (See a translation of his paper in _Trans. and
Proc. N. Z. Inst._ Vol. 25.) Bearing in mind the known capacity of ferns
for dispersal by the winds, an hypothesis of this kind, even if
established, seems scarcely needed in the study of fern-dispersal.

It is probable that many of the ferns and lycopods reached Hawaii
directly and not through South Polynesia. The mountain-ferns of this
group could hardly have been received by that route, since, as is shown
below, they do not as a rule occur in that region.

Some other interesting relations present themselves in connection with
the Hawaiian ferns and lycopods when we consider the distribution of its
non-endemic species in the other two groups of Fiji and Tahiti. Out of
these species, some eighty in all, not more than half are common to all
three groups, and about two dozen have not been found either in Fiji or
in Tahiti. Of these last quite half are mountain species in Hawaii,
having their station at elevations exceeding those of the highest
districts of Fiji and of the several islands of the Tahitian area,
excepting the limited region comprised in the uplands of Tahiti itself.

A glance at the list, given in Note 65 of some of the mountain ferns of
Hawaii not recorded from Fiji and Tahiti will show that these species
are very widely distributed. Ferns and lycopods found in the Himalayas
and in the Andes meet on the higher slopes of the lofty mountains of
Hawaii and in no other of the less elevated island-groups of the open
Pacific. This distribution of the vascular cryptogams thus foreshadows a
principle that will come into prominence in the case of the flowering
plants, namely, that difference in elevation has been an important
factor in determining some of the contrasts between the Hawaiian,
Fijian, and Tahitian floras. The contrasts here implied are those
connected with the climatic conditions of station, since several plants
of temperate regions, such as Aspidium filix mas, Asplenium trichomanes,
Asplenium adiantum nigrum, &c., that are at home in the highlands of
Hawaii, do not occur in either Fiji or Tahiti. We can infer that widely
ranging ferns and lycopods have been dispersed over the oceanic groups
of the tropical Pacific with a fair degree of uniformity, and that any
marked contrasts in their distribution may be attributed to considerable
differences in the altitude of the islands.

In appreciating such a conclusion, and in dealing with apparent
exceptions to the rule, the relation between the vertical range of a
species and its lateral distribution has to be considered. We find, for
instance, that whilst the Common Bracken (Pteris aquilina) is a mountain
plant in Hawaii, it occurs also in Fiji and Tahiti. Since, however, it
is found all over the temperate and tropical regions, and has a vertical
range in Hawaii of from 800 to 8,000 feet, any difficulty in this
respect is thus explained. Aspidium aculeatum, a characteristic fern of
temperate latitudes, seems at first to present a difficulty, which,
however, proves to be more apparent than real. Whilst it has been
recorded from Hawaii at heights of 6,000 to 9,000 feet, and from Tahiti
at 4,000 feet, it has also been found in Fiji and Samoa; but since it
was not collected by Seemann in Fiji, it can scarcely be common, and
Horne seems only to have obtained it from the tops of mountains in Vanua
Levu at an elevation of 1,800 feet.

Up to this point the non-endemic ferns and lycopods have been chiefly
discussed. We will now briefly deal with the probable cause of the
relative preponderance of peculiar or endemic species in Hawaii as
contrasted with Fiji and Tahiti. In this respect the Hawaiian islands,
as remarked at the commencement of this chapter, come into sharp
contrast with the other two groups; but it would seem that the
differentiation has rarely acquired a generic value (see Note 66). In
this respect the age of ferns is markedly distinguished from the
succeeding era, the age of the arborescent Compositæ and of
Tree-Lobelias, to which a large number of peculiar genera belong. This,
according to my view, is to be attributed to the circumstance that
whilst the dispersion of spores by the wind is probably as active in our
own time as it was in the earliest stage of the floral history of
Hawaii, the dispersion of seeds by birds, to which the flowering plants
in the main originally owe their presence in this group, has been
greatly influenced by the various changes that have affected the
migration of birds over the Pacific, a subject discussed in later pages.

Respecting the origin of the species of ferns and lycopods peculiar to
Hawaii, it is first of importance to quote the remarks of Dr. Hillebrand
on the subject. Speaking of the whole flora (p. xxv), but evidently with
the ferns more especially in his mind, he says:—“Nature here luxuriates
in formative energy. Is it because the islands offer a great range of
conditions of life? Or is it because the leading genera are in their age
of manhood, of greatest vigour? Or is it because the number of types
which here come into play is limited, and, therefore, the area offered
to their development comparatively great and varied?” It is deeply to be
regretted that sickness and death intervened before the author was able
to give to the world his matured views on the very important points here
raised. Yet they are much the same questions that man is ever putting to
the life around him. There is the same querulous note that we find in
all, the question that begins, the question that ends, and the reply
that never comes.

“The evolution theory (writes Dr. Hillebrand, p. xxix) could hardly find
a more favourable field for observation than an isolated island-group in
mid-ocean, large enough to have produced a number of original forms, and
at the same time so diversified in conditions of temperature, humidity,
and atmospheric currents as to admit an extraordinary development in
nearly every direction of vegetable morphology, uninfluenced by
intercrossing with foreign elements.” Isolation thus admittedly offers
the preliminary determining or favouring conditions. This is directly
indicated by the fact that Hawaii possesses fewer genera of ferns and
lycopods than either Fiji or Tahiti, notwithstanding that it has the
same area as Fiji, and is in extent three or four times the size of the
whole Tahitian area. One effect of isolation in Hawaii has, therefore,
been greater room for the development of new forms. It has, however,
already been remarked that the islands of the Fijian area are much less
isolated than those of the Hawaiian group, and that in consequence the
free immigration possible in the one group has been checked in the
other. Fiji possesses in respect to vascular cryptogams at least half as
many species again as Hawaii, but Hawaii has three or four times the
number of peculiar species. Yet before this great contrast can be
ascribed to different degrees of isolation, it is necessary to exclude
another possible cause presented by the greater range of life-conditions
in Hawaii. It is possible that all the Hawaiian peculiar species may
belong to the higher levels, elevations, as before shown, not
represented in the Fijian islands, which correspond only to the lowlands
of Hawaii, that is, to levels below 4,000 feet. If this is the case, the
contrast between Fiji and Hawaii would be connected mainly with a
difference in life-conditions, and, however potent the isolating
influences might have been in Hawaii, they could hardly have been
concerned with this striking difference.

In order to determine this point, I went carefully through the account
given by Hillebrand of the Hawaiian ferns and lycopods, noting the
altitudes there given, and making use of the maps and of my own local
knowledge of the islands of Oahu and Hawaii, where the elevation is
neither directly nor indirectly implied. As a result, I found that out
of sixty-six endemic species available for my purpose, forty-seven had
their stations at levels below 4,000 feet, that is in the region
corresponding to Fiji, and nineteen at elevations exceeding this height.
This, however, did not finally decide the question, since the proportion
of endemic species may be much smaller in the region below 4,000 feet
than in that above it. I, therefore, went over the ground again, and
found, as shown in the table below, that the percentages of peculiar
species amongst the total available for my use were not very far apart,
58 per cent. for the upper region and 43 per cent. for the lower region.

Distribution of the Hawaiian ferns and lycopods above and below 4,000
feet.

+------------------------+-------+--------+----------------+
| |Number.|Endemic.| Percentage of |
| | | |endemic species.|
+------------------------+-------+--------+----------------+
|Species below 4,000 feet| 110 | 47 | 43 |
+------------------------+-------+--------+----------------+
|Species above 4,000 feet| 33 | 19 | 58 |
+------------------------+-------+--------+----------------+

From the above it would appear that although the process of
species-production in the Hawaiian islands has seemingly been rather
more active above than below 4,000 feet, if we were to compare the
entire vascular cryptogamic flora of Fiji with that of the corresponding
lower levels of the Hawaiian group we should obtain much the same
contrast in the proportion of peculiar species that we obtained when
comparing all the ferns and lycopods of both groups. In other words, if
we were to restrict our comparison with Fiji, and I may add Tahiti, to
that lower portion of Hawaii that corresponds in elevation, we should
not get results very different from those to be obtained by including
the Hawaiian upland regions as well.

We are, I think, on these grounds justified in assuming that the
relatively great development of new species of ferns and lycopods in
Hawaii as contrasted with Fiji is not to be connected with the greater
elevation of those islands. The only thing that we have been able to
associate with the greater altitude of the Hawaiian Islands, and the
consequent greater range of climatic conditions, when contrasting the
Fijian and Hawaiian vascular cryptogams, is the occurrence of a number
of peculiar mountain species and of wide-ranging temperate species that
are found in the uplands of Hawaii, but not in the less elevated islands
of Fiji.

On the whole, therefore, it is to be inferred that the greater display
of formative power among the ferns and lycopods of the Hawaiian Islands
is in great part to be associated with the isolation of this group as
compared with those of Fiji and Tahiti. The indications supplied by the
vascular cryptogams resemble in kind those we shall obtain from the
study of the flowering plants, but there is this important distinction.
In formative power, as shown in the development of new specific and
generic types, the Hawaiian vascular cryptogams are far exceeded by the
flowering plants where the proportion of endemic species amounts to 80
per cent. We have no reason to believe that the winds, to which the
ferns and lycopods chiefly owe their dispersal, are less effective now
in carrying their spores than they were in the earliest era of the
floral history of Hawaii or in the intervening periods. In the course of
ages the winds have been more uniform in their action as
plant-dispersers even than the currents, and certainly far more than
birds.

On the other hand, in the case of the Hawaiian flowering plants that
depend on the varying influence of the migrant bird, the agency of
dispersal has often been suspended altogether, and far greater
differentiation or departure from the original type has resulted, the
amount of change often reaching to the value of a generic distinction.
It is a question, however, whether the isolation of the Hawaiian Islands
is to be entirely connected with their mid-oceanic position. It will be
shown in Chapter XXXIII. that effects almost as great have been produced
in continental regions and in continental islands, and that the isolated
situation of Hawaii has not induced but has intensified these results.
In the later eras of plant-life a process of segregation has been ever
active throughout the tropical world whether in the case of an elevated
oceanic island or of a mid-continental mountain.

The following are some of the principal points that have been emphasised
in the foregoing discussion of the ferns and lycopods of the Hawaiian,
Fijian, and Tahitian Islands:—

(_a_) In all three groups the vascular cryptogams (ferns and lycopods)
have been largely supplied from the warmer regions of the Old World. But
whilst in the South Pacific the migration has been mainly from Fiji
eastward to Tahiti, it is probable that Hawaii in the North Pacific has
been in part independently stocked.

(_b_) Whilst in Hawaii many peculiar species of ferns and lycopods have
been developed, in Fiji and Tahiti there have been comparatively few.

(_c_) Whilst there has been more or less free immigration into Fiji and
Tahiti there has been comparative isolation in Hawaii. Though the areas
of the Fijian and Hawaiian archipelagoes are about the same, Fiji
possesses at least half as many species again as Hawaii; but Hawaii owns
three or four times the number of peculiar species.

(_d_) Though the land-area of the Tahitian region does not exceed a
fourth part of that of Hawaii, it has the same number of species. The
Tahitian islands therefore display a predominance of ferns and lycopods.

(_e_) The non-effective influence of the greater elevation of the
Hawaiian Islands on its preponderance of peculiar species is shown by
comparing all the ferns and lycopods of the Fijian and Tahitian Islands
with those of the corresponding lower levels of the Hawaiian Islands,
when we find much the same contrast exhibited in the number of peculiar
species.

(_f_) Whilst a large proportion of the ferns and lycopods are common to
all three groups, Hawaii possesses a number of mountain species, widely
distributed in temperate regions and on the higher levels of mountainous
areas in the tropics, that are not found either in Fiji or in Tahiti.
Their absence from these two groups is due to the insufficient elevation
of the islands and to the non-existence there of extensive areas of any
altitude.

(_g_) The agency of the winds in dispersing the spores of ferns and
lycopods has been relatively uniform through the ages when compared with
the varying agency of the migrant bird, to which the flowering plants
mainly owe their distribution. Thus it is that in the Pacific islands
the vascular cryptogams have experienced much less differentiation than
the flowering plants, though as a rule far older denizens of the
islands. Yet we cannot doubt that the same principle has been at work in
both cases, the difference arising in the instance of the flowering
plants from the interrupted and often suspended agency of birds in the
work of dispersal.

(_h_) It is a question whether there is not something more concerned in
the isolation of the Hawaiian group than its mid-oceanic position, since
effects almost as great have been produced in continental regions.

CHAPTER XXI

THE ERAS OF THE FLOWERING PLANTS

THE AGE OF COMPOSITÆ.

_The Endemism of the Pacific Island Floras._

AS far as the production of new species is concerned, the Hawaiian group
presents the same contrast with the Fijian and Tahitian groups in
respect of the flowering plants that it does as regards the ferns and
lycopods. The proportion of endemic species, after excluding all
introduced plants, is in Hawaii 80 per cent., in Fiji about 50 per
cent., and in Tahiti 35 per cent. (see Table A). The same contrast is
also displayed in the number of peculiar genera. In Hawaii there are,
according to Dr. Hillebrand, 37 or 38, and in Fiji Dr. Seemann
discovered 16; whilst, as we learn from Drake del Castillo, there are
only 3 or 4 in the Tahitian Islands. (As will be pointed out later on,
these numbers for Fiji and Hawaii have to be reduced, but the general
inference to be drawn from them is not materially affected; see Table
B.)

But if we look at the accompanying table (Table B) we notice that the
flora of Hawaii is sharply contrasted with those of Fiji and Tahiti not
only in the large proportion of endemic genera, but also in the large
number of non-endemic genera with peculiar species, and in the small
proportion of genera possessing no peculiar species. There is an endemic
element of greater or less degree in about 70 per cent. of the Hawaiian
genera, whilst in Fiji only about 53 per cent. and in Tahiti as few as
34 per cent. of the genera contain to a varying extent peculiar species.
Another feature brought out in this table is the relative poverty of
genera in the Hawaiian Islands. Fiji, though about the same size as
Hawaii, contains nearly half as many genera again, whilst the islands of
the Tahitian region, which in the aggregate amount to only one-third or
one-fourth of the area of the islands of Hawaii, possess nearly as many
genera.

TABLE A (FLOWERING PLANTS).

_Proportions of Endemic Species in the Hawaiian, Fijian, and Tahitian
floras, with those
for Samoa, Tonga, and Rarotonga added._

+---------+------------------+----------------+----------------+
| Groups. |Number of species.| Number of | Percentage of |
| | |endemic species.|endemic species.|
+---------+------------------+----------------+----------------+
| Hawaii | 686 | 546 | 80 |
| | | | |
| Fiji | {S. 617 | {288 | {47 |
| | {H.1086 | {620 | {57 |
| | | | |
| Tahiti | 315 | 112 | 35 |
+---------+------------------+----------------+----------------+
| Samoa | 326 | 110 | 34 |
| | | | |
| Tonga | 285 | 17 | 6 |
| | | | |
|Rarotonga| 140 | 17 | 12 |
| Island | | | |
+---------+------------------+----------------+----------------+

_Remarks._—The materials for this table have been obtained from the
works of Hillebrand for Hawaii, Seemann and Horne for Fiji, Drake del
Castillo for Tahiti, Reinecke for Samoa, Hemsley and Burkill for Tonga,
and Cheeseman for Rarotonga. The two estimates for Fiji are marked S.
for Seemann and H. for Horne, the last being a rough preliminary
computation made by Horne himself.

The results given are only to be considered as approximations liable to
emendation, but as regards the proportion of endemic species in the
several groups they no doubt illustrate fairly well the relative degree
of endemism in the various archipelagoes. The results for Samoa, Tonga,
and Rarotonga are merely added in order to enable a comparison to be
made with sub-groups of a region and with solitary islands, the
Hawaiian, Fijian, and Tahitian groups being regarded as the three
principal centres of plant-life in the open Pacific.

All plants introduced by the aborigines and the white man are excluded.
In so doing, I have mainly followed Seemann, a safe guide in all matters
relating to weeds and to cultivated plants. The flora of a Pacific
island thus treated undergoes serious diminution in its extent. In the
case of the Rarotonga flora, for example, which according to Cheeseman
includes about 260 flowering plants, the number of truly indigenous
plants, in the sense here implied, is only 140. Though this is an
extreme case, it will serve to illustrate the principle here followed.

TABLE B (FLOWERING PLANTS).

_Comparison of the Hawaiian, Fijian, and Tahitian genera._ (_All genera
containing
introduced plants entirely are excluded._)

+----------+------------------------------------------+---------+-----------+
| | Non-endemic genera. | | |
| +------------+---------------+-------------+ | |
| Group. | No endemic | Some species | All species | Endemic | Total. |
| | species. | endemic, some | endemic. | genera. | |
| | | not. | | | |
+----------+------------+---------------+-------------+---------+-----------+
|Hawaii | 70(31)| 30(13)| 95(43)| 28(13)| 223(100)|
| | | | | | |
|Fiji {| S. 150(47)| S. 74(23)| S. 87(27)| S. 10(3)|S. 321(100)|
| {| H. 162(47)| H. 80(23)| H. 94(27)| H. 10(3)|H. 346(100)|
| | | | | | |
|Tahiti | 125(66)| 21(11)| 40(21)| 4(2)| 190(100)|
|(Eastern | | | | | |
|Polynesia)| | | | | |
+----------+------------+---------------+-------------+---------+-----------+

_Remarks._—The figures in brackets are percentages. S. = Seemann, H. =
Horne and Seemann.

In the construction of this table, Hillebrand, Seemann, and Drake del
Castillo have been mainly followed, except with regard to the endemic
genera for Hawaii and Fiji. In this respect the _Index Kewensis_ has
been largely consulted as well as Engler’s publications, as indicated in
the text. Hillebrand’s total of nearly forty Hawaiian peculiar genera
and Seemann’s total of sixteen for Fiji have thus been considerably
reduced. The two results given for Fiji are those of Seemann alone and
with Horne superadded. Horne discovered, according to Hemsley, no new
genera, but several genera from outside regions were added to the Fijian
flora. Taking them as twenty-five (two-thirds of his own computation), I
have apportioned them as in Seemann’s results. The Tahitian region here
includes Eastern Polynesia.

It is necessary before proceeding further to obtain a correct idea of
the significance of a large endemic element in the phanerogamic flora of
a Pacific archipelago. We have therefore at the outset to inquire
whether it is indicative of isolation or of antiquity. If the number of
peculiar genera is to be regarded as the test of the relative antiquity
of different Pacific floras and, by implication, of the islands to which
they belong, these three groups, as shown in Table B, would arrange
themselves in the following order, namely, Hawaii, Fiji, Tahiti. This
test might be reliable if the several groups were in the same condition
of isolation. Since, however, as we have previously seen, the Fijian
Islands still enjoy a fairly free communication with the islands
westward, whilst the Hawaiian group is largely cut off, it is apparent
that the tendency to generic differentiation in Fiji might have been
often swamped by immigration, and that Fiji with its much smaller number
of endemic genera may even be older than Hawaii. This objection does not
apply quite as forcibly to a comparison between Hawaii and Tahiti, yet
for reasons before given it may be regarded as sufficient to negative
any inferences concerned with relative antiquity.

On account, therefore, of the great differences in the degree of
isolation of these three groups, we cannot be guided in our estimation
of the relative antiquity of their floras by their number of peculiar
genera. With the evidence at our disposal we are compelled to accept the
view, which indeed a single glance at a map would suggest, that the
number or proportion of endemic genera is to be connected with the
degree of isolation. Whether a parallelism can be traced in the original
stocking of these groups with their earliest flowering-plants is a
matter that can only be elucidated by a further analysis of the peculiar
genera.

SYNOPSIS OF THE ERAS OF THE FLOWERING PLANTS IN THE TROPICAL PACIFIC.

A. _The Era of the Endemic Genera._—Mostly American in their affinities.
Represented particularly by Compositæ and Lobeliaceæ.

B. _The Era of Non-Endemic Genera._

(1) The mountain genera, either cosmopolitan in temperate latitudes or
derived from the New Zealand or the Antarctic flora. Mostly
represented in Hawaii.

(2) The genera forming the low-level flora of Hawaii below 4,000 or
5,000 feet and composing almost the entire floras of the Fijian and
Tahitian regions. Predominantly Indo-Malayan.

(_a_) The age of general dispersal over the tropical Pacific, the
genera with only peculiar species being first treated, and afterwards
those possessing a non-endemic element.

(_b_) The age of local dispersal over the tropical Pacific.

THE FIRST ERA OF THE FLOWERING PLANTS, BEING THE AGE OF THE ENDEMIC
GENERA.

With the above preliminary remarks I pass on to the next stage in the
history of the stocking of these islands with their plants. The age of
the ferns and lycopods is left behind, and it is assumed that the next
era is mainly indicated by those genera of phanerogams that are now
peculiar to their respective groups. In this connection by far the most
interesting of the three regions, the Hawaiian, the Tahitian or East
Polynesian, and the Fijian, is that of Hawaii, which, as before
observed, is distinguished from the groups of the Fijian and Tahitian
regions, or, in other words, from all the oceanic archipelagoes of the
tropical Pacific, by its large number of endemic genera.

Peculiar genera of shrubby and arborescent Compositæ and of arborescent
Lobeliaceæ form the most striking characteristics of the endemic genera,
and therefore of the ancient flora of Hawaii. It is in this connection
of singular interest to remark that of the three endemic genera of the
Tahitian flora one is an arborescent genus of the Compositæ, and the
other two are shrubby genera of the Lobeliaceæ. There are, therefore,
indications here of an ancient insular flora of the Pacific,
characterised mainly by the prevalence of Compositæ and Lobeliaceæ. It
is, however, remarkable that not only are no endemic genera of these
orders known from Fiji or from the adjacent groups of Samoa and Tonga,
but that the Lobeliaceæ are not represented at all, whilst amongst the
Fijian Compositæ, with the exception of Lagenophora, the genera display
no endemic element as far as the data at my disposal indicate.

The problem we are brought face to face with is clearly stated by Mr.
Hemsley in the _Introduction to the Botany of the Challenger Expedition_
(p. 68). “In Polynesia as elsewhere,” he remarks, “the Compositæ more
particularly are perplexing to the botanical geographer, for although
they have their greatest affinities in America, as well as the
sub-arboreous Lobeliaceæ, so numerous in the Sandwich Islands, yet the
bulk of the vegetation seems to have been derived from the
Australo-Asiatic region.”

In attempting to approach this problem I do so from the standpoint of
dispersal. There are so many intricate questions bound up with the
systematic position of these genera that in dealing with them the
student of plant-distribution would require the capacities and
opportunities of the eminent botanist who dealt with the distribution of
ten thousand species of Compositæ. On such ground, therefore, and only
under the guidance of others, I will lightly tread.

THE ENDEMIC GENERA OF COMPOSITÆ.

On account of their endemic character the peculiar genera of Compositæ
are regarded as belonging to the oldest era of the flowering plants of
the island-groups lying in the tropical latitudes of the open Pacific.
This is the view of Bentham, but it is, of course, the opinion that most
botanists would arrive at with the facts before them. With the exception
of the solitary Tahitian genus Fitchia, they are all restricted to the
Hawaiian Islands, and nearly all are either shrubby or arborescent, the
greatest height of 25 to 30 feet being attained in the Tahitian genus
and in Hesperomannia of Hawaii.

Nine Hawaiian genera are included in this era, though, strictly
speaking, we ought only to concern ourselves with the six genera, Remya,
Argyroxiphium, Wilkesia, Dubautia, Raillardia, and Hesperomannia, since
the other three, Tetramolopium, Lipochæta, and Campylotheca, are only on
the borderland of generic distinction. It is, however, necessary that we
should include these three genera in our treatment of the Hawaiian
endemic genera, more especially because they appear to have been the
last arrivals of the early Compositæ. They still display, as shown
below, a very suggestive connection with the land of their birth, a
circumstance that is of much importance in finally determining the
source of the other strictly endemic genera, where the links with their
original homes have been in most cases largely severed.

It would, however, be quite out of place here to enter into any details
into the affinities of these Hawaiian genera of Compositæ, and I will
limit myself here to such general conclusions as may be derived from the
pages of Bentham, Hillebrand, Hemsley, and other writers, and such as
are in accordance with the facts of distribution given in the _Index
Kewensis_. Most ancient of all are the genera Remya, Argyroxiphium,
Wilkesia, and Hesperomannia, which, although belonging to tribes that
only occur on the American continent, as in the Mexican region, stand
quite isolated, and, as Dr. Hillebrand remarks, probably belong to the
oldest denizens of the Hawaiian Islands. It is noteworthy that these
four ancient genera only contain two species apiece, a circumstance that
favours their priority in point of age.

The American affinities, however, are not always of the character that
we might have expected. Thus, it was remarked by Mr. Bentham that
although the tribe Mutisiaceæ attains a great development in South
America, and especially in Chile, its only representative in the Pacific
islands is the very rare arboreous Hesperomannia of Hawaii.

Rather less isolated in character, and we would presume therefore of
somewhat less antiquity, are the two closely allied genera of Raillardia
and Dubautia, which have a close relative in Raillardella of the Sierra
Nevada in California. Then we come to the three genera, Tetramolopium,
Lipochæta, and Campylotheca, that, being still in touch with the world
outside, may be regarded as the latest arrivals of the early genera of
the Compositæ. Tetramolopium, concerning which botanists were unable to
agree, would seem, according to the _Index Kewensis_, to possess Mexican
and Ecuadorian as well as Hawaiian species. Lipochæta, nearly related to
other American genera, contains a dozen species, of which eleven are
found only in Hawaii, whilst the twelfth occurs, according to the _Index
Kewensis_, in California, and, according to Dr. Hillebrand, in the
Galapagos group. Of the generic value of Campylotheca there seems a
doubt, and its distinctness is scarcely recognised in the _Index
Kewensis_. It is, however, closely allied to Coreopsis, an American
genus represented, according to Drake del Castillo, in the Marquesas.

In the Tahitian region, that is to say in Eastern Polynesia, the genus
Fitchia alone belongs to the early age of the Compositæ, so
characteristic of Hawaii. Indications of the former widespread range of
the genus over this region of the South Pacific are afforded by its
being now represented by two species in Tahiti and by one species in
Rarotonga, localities nearly 700 miles apart. It was thus regarded by
Bentham, who saw in it a solitary remnant of the ancient South Pacific
flora. Like the Hawaiian genera, as shown below, it is often restricted
to the higher levels. Botanists differ about its affinities, and a
discussion of the subject will be found on pages 20 and 66 of the
_Introduction to the Botany of the Challenger Expedition_.

The restriction of these ancient genera of the Polynesian Compositæ to
the upland regions is of some interest. “The preponderance of Compositæ
among the high-level plants obtains almost throughout the world.” This
observation was made by Mr. Hemsley in connection with the flora of the
highlands of Tibet (_Journ. Linn. Soc. Bot._ vol. 35, 1902), where the
Compositæ constitute about 19 per cent. of the flowering plants; and I
may remark in passing that, according to Mr. Ball, one of the most
conspicuous elements in point of frequency in the higher flora of the
Great Atlas is presented by the Compositæ which make up between 12 and
13 per cent. of the whole flora (Hooker and Ball’s _Marocco and the
Great Atlas_). This feature of alpine floras is brought into great
prominence in Schimper’s recent book on Plant Geography.

Some of the most lasting reminiscences that the naturalist will bear
away with him from the highlands of Hawaii are connected with the
Compositæ. Those who have ascended the mountains of Mauna Kea and Mauna
Loa, will remember that amongst the last plants occurring above the
forest zone, and scattered about on the ancient lava fields at
elevations exceeding 10,000 feet above the sea, are species of
Raillardia and the beautiful “Ahinahina” (Argyroxiphium). It is,
however, in the open, scantily wooded region, elevated 6,000 to 9,000
feet, and lying between the true forest zone below and the bare lava
slopes above, that the shrubby and arborescent Compositæ of the large
island of Hawaii are most at home. Such regions, as Hillebrand well
describes (p. xxiv), are characterised by stunted trees, chiefly
Sophora, Cyathodes, Myoporum, and others, associated with arborescent
Raillardiæ of the order of Compositæ. Between them luxuriate other
shrubby Compositæ of the genera Raillardia, Dubautia, Campylotheca, and
Artemisia, together with Strawberries, Raspberries, and species of
Vaccinium.

Botanists have not given us much account of the associates of the
interesting genus Fitchia on the uplands of Tahiti. We learn, however,
from Nadeaud that in his time these Composite trees and shrubs were
spread over the higher region of the island of Tahiti above 800 and
1,000 metres. Cheeseman, to whom we are indebted for the discovery and
the description of the Rarotongan species, tells us that this tree,
which attains a height of 25 feet in the sheltered valleys, and is much
dwarfed on the exposed ridges and hill-tops, often forms the greater
part of the forest above 500 feet, and reaches the highest peaks of the
island (2,250 feet).

In discussing the probable mode of dispersal of these early Composite
plants of the Pacific we shall be treading on somewhat debatable ground.
We will, however, point out that the mere possession of structures that
could be utilised for dispersal of the seeds is not the only important
question here involved. If we could demonstrate that all these genera
possess exceptional capacities for distribution over the ocean, we
should prove too much, since the process has been in the main suspended
for ages. If, on the other side, it could be shown that their fruits are
not at all suited for such dispersal, we should prove too little, since
the ancestors of these genera must have been transported to these
islands in some fashion or other. This clearly indicates that other
important factors have also come into play in determining the
distribution of the early Compositæ of the Pacific islands.

It was long ago pointed out by De Candolle that the possession of a
pappus does not, as a rule, increase the area of a Composite plant,
although as regards hooks and barbed appendages, such as occur in
Bidens, the greater areas of the plants thus provided may be, as he
thought, in some measure explained. Even in respect to hooks and barbs
it would be easy to point to cases where, as Bentham remarks, unusual
powers of adherence are by no means indicative of wide dispersal in all
cases. In any event it will be also incumbent on us to explain why these
genera no longer possess facilities for distribution. This suspension of
the means of dispersal is not, however, peculiar to the age of the
endemic genera of the Pacific islands. It is a character but in a less
degree of the succeeding age, the age of genera found outside the group,
but represented within it by endemic species; and from this we may
suspect that we have had in operation in the Pacific an influence,
far-reaching both in time and space, to which the agencies of dispersal
have been compelled to adapt themselves, an influence which has acted as
a distributor of the distributing agencies.

Coming to the fitness for dispersal of the achenes of the early
Composite genera of the Pacific islands, it will be assumed that they
have been, as a general rule, transported in birds’ plumage. The fruits
are usually 2·5 to 12 millimetres (1/16 to 1/2 inch) in length, and are
provided either with a pappus of soft or stiff bristles, or with awns or
teeth, but these appendages vary much in size in the different genera
and in different species of the same genus. The instance of Lipochæta is
especially significant as indicating the alterations which the
appendages of the achene may have undergone in the cases of other
genera. With most species there are usually two or three teeth or short
awns, but in some species these are obsolete, and in others they are
long and stout.

Bearing these facts in mind we should hesitate to rely too much on the
present condition of the achenes in the other genera as an indication of
the fitness for dispersal of the fruits of their ancestors. In one
genus, Campylotheca, which may be regarded as among the youngest of the
genera, the achenes are provided with barbed or hooked awns which cause
them to adhere as tenaciously to one’s clothes as in the case of those
of Bidens, an allied genus. In Fitchia, the Tahitian genus, which may be
looked upon as one of the oldest of the Pacific genera of Compositæ, the
achene is furnished with two long awns or setæ, which, as Drake del
Castillo observes, recall those of Bidens. The achenes of the other
Hawaiian genera, as regards their fitness for dispersal in plumage, may
be said to give less definite indications. In some, as in Dubautia and
Raillardia, there is a typical pappus of ten to twenty long hair-like
bristles. In others again, as in Wilkesia and Argyroxiphium, the pappus
is much reduced, and in some species of Lipochæta it is, as above
remarked, quite obsolete.

The chances of the achenes of the parent plants having in some cases
been originally transported to the islands in the plumage of birds would
be increased by a bird making its nest of the plant-materials or amongst
the plants themselves, or by its pecking at the fruit-heads. In our own
time different species of the grouse family on the slopes of the
Californian and Columbian mountains make their nests on the ground under
the shade of Artemisia bushes and find a portion of their sustenance in
their fruits. Artemisias also form one of the features of the vegetation
of the Hawaiian uplands; but since they present only specific
differentiation they are referred to a later era. Yet it will be on the
slopes of the Rocky Mountains and of the Californian Sierra Nevada,
amongst the “sage-brush” and the grouse, that we may have to stand when
we look in thought across the Pacific towards far distant Hawaii and ask
ourselves whence came its tree-like Raillardias, its shrubby Dubautias,
its tall Wilkesias, and the silvery Ahinahinas (Argyroxiphium).

It is possible that in some genera the achenes have, or had, a means of
adhering to plumage through a “sticky” secretion, such as is sometimes
found with Lagenophora, an Hawaiian genus of the next era, and also with
the weed-plant Adenostemma viscosum; but this is a point that has not
yet been investigated. Nor can we altogether exclude the chance of the
achenes having in some cases been transported unharmed to Hawaii in a
bird’s stomach. The possibility of this has been above implied in the
case of Artemisia; and it is pointed out in Chapter XXXIII. that pigeons
in Hawaii feed sometimes on the achenes of Compositæ. The Hawaiian goose
(Bernicla sandwicensis) lives, according to Mr. Dole, on Sonchus asper,
an introduced plant, as well as on berries (Wilson’s _Aves
Hawaiiensis_). There are numerous references of this nature in books
about birds, and it should always be remembered that birds in pecking at
the fruit-heads scatter the seeds on their feathers. (See Note 67.)

From the foregoing remarks it may, I think, be inferred that the achenes
of the ancestors of the original Composite genera of the Pacific islands
were in all probability not unfitted for transport by birds, more
especially in their plumage. Some of my readers, however, may express a
doubt as to whether birds likely to disperse seeds would be found in any
numbers at the great heights where some of the continental Compositæ
occur. But it is well known that birds of the grouse and partridge
family frequent high levels in continental regions over much of the
globe. Arborescent Compositæ are found at heights of 10,000 to 14,000
feet on the mountains of Central Africa; and it should be noticed that
Sir Harry Johnston observed “francolins” on the slopes of Ruwenzori up
to 13,000 feet (_Uganda Protectorate_, vol. 1; _Trans. Linn. Soc. Bot._,
Ser. II. vol. 2). Sir Martin Conway in the Bolivian Andes found geese,
ducks, gulls, snipe, &c., numerous in suitable places up to 17,000 feet
(_Journ. Roy. Geogr. Soc._, 1899); whilst geese and teal were noticed by
Sir Joseph Hooker and others at elevations of 17,000 feet in the
mountains of Tibet (Hooker’s _Himalayan Journals_; _Journ. Linn. Soc.
Bot._, vol. 35, p. 147). These are all birds, as shown in Chapter
XXXIII., that are likely to disperse plants, and probably none more
effectually than the goose, of which Hawaii possesses a particular
variety or species. It may be remarked that geese, ducks, gulls, and
other birds use Cotula plumosa in Kerguelen for making their nests (Dr.
Kidder quoted by Mr. Dixon in his book on Birds’ Nests).

Sea-birds were probably the principal agents in carrying the achenes of
the early genera of the Compositæ to Hawaii. Dr. Hillebrand attached
importance to the tropic-bird (Phaethon) in the distribution of species
(Introd., p. 30); and since these birds breed at the crater of Kilauea
in Hawaii, 4,000 feet above the sea, and also high up in Tahiti
(Moseley), its agency is not unlikely, I am inclined to think, however,
that birds like the petrels and puffins, that in nesting burrow in the
ground, choosing places where the vegetation is thickest, and where they
would be likely to get seeds on their feathers, would be more efficient
agents. This is the view expressed by Prof. Moseley in Wallace’s _Island
Life_, p. 250. He considered that albatrosses, petrels, and puffins have
played a great part in the distribution of plants, and to some degree
especially account for the otherwise difficult fact that widely distant
islands in tropical seas have similar mountain plants. Birds, he says,
that in high latitudes, as at Tristan da Cunha and Kerguelen, often
burrow near the sea-level, in the tropics choose the mountains for their
nesting-place; and he refers to a puffin that nests on the top of one of
the high mountains of Viti Levu at an altitude of 4,000 feet, to a
petrel nesting among ferns at Tahiti at an elevation of 4,400 feet, and
to another petrel breeding in like manner in the high mountains of
Jamaica at a height of several thousand feet above the sea. He gives
point to these interesting remarks, which might be supplemented by data
from other parts of the world, by observing that it is not necessary
that the same species should now cover the range of the plants
concerned. The ancestor of the species might have carried the seeds, and
the range of the genus is alone sufficient. It may be added that, as I
have shown in Chapter XXXIII., sea-birds have been far more active
agents in the distribution of plants than many people might imagine. The
more recent observations of Ekstam in Spitzbergen have thrown
considerable light on this subject.

Having in the first place formed the opinion that the achenes of the
early Hawaiian Compositæ are suited for dispersal by birds, and then
shown that sea-birds were probably the principal agents, we are met with
the curious difficulty that in the case of the early Hawaiian genera of
Compositæ the complete suspension for ages of the means of dispersal is
involved in the circumstances that these genera are confined to the
Hawaiian group. We can attribute to the agency of existing sea-birds the
occurrence of the genus Lagenophora in the uplands of Hawaii, on the
mountain-tops of Fiji, and in Australia and New Zealand; but the agency
of birds as at present in operation does not assist us except indirectly
in the case of the genera restricted to Hawaii or to Tahiti. Is it
possible, we may inquire, to penetrate this mystery? Why, we may ask
with Mr. Hemsley, has the agency ceased acting, and why have its
operations been confined to the conveyance of seeds _to_ the islands and
not _from_ the islands as well (_Intr. Bot. Chall. Exped._, p. 66)? I
need scarcely add that the same question presents itself with all the
other peculiar genera of these islands, and in fact with endemic genera
all over the world. What can be stranger, it may be remarked, than the
limited distribution of the Pandanaceous genus Sararanga in the Western
Pacific, although suited for dispersal by frugivorous birds. This is
not, indeed, a special difficulty connected with oceanic islands; it
applies to the whole plant-world; yet it is possible that, as it is
exhibited by the Compositæ in these islands, we may be in a better
position to grapple with the problem. But before doing so it will be
requisite to look a little closer at these early Hawaiian genera of the
Compositæ.

The distribution within the archipelago of the genera and species of the
early Compositæ of Hawaii is worthy of notice from the light it throws,
not only on the relative antiquity of the genera, but also on the
subsequent conditions of isolation. Of the nine genera here referred to
five are distributed over most of the islands of the group. These
include all the genera possessing a number of species, namely,
Tetramolopium with seven species, Lipochæta with eleven, Campylotheca
with twelve, Dubautia with six, and Raillardia with twelve species. Of
the four genera remaining all have only two species, and are restricted
to two or three islands, Remya and Wilkesia being in both cases found in
Kauai and Maui, whilst Argyroxiphium is confined to the adjacent islands
of Maui and Hawaii, and Hesperomannia to those of Oahu, Lanai, and Maui.
These four genera that are restricted to only two or three islands are
the same before referred to as regarded by Hillebrand as the oldest,
partly on account of their isolated generic position, and partly because
in each case they only possess two species.

Although the early Hawaiian Compositæ were evidently originally
transported to most of the islands of the group, it is noteworthy that
their subsequent isolation from the rest of the world has in the later
ages been repeated within the limits of the archipelago. Of the 56
species, all of which are now endemic, 28, or just half, as shown in the
table on the following page, are confined to a single island. Of the
remainder, almost all are restricted to two or three adjacent islands.
Hillebrand gives only a solitary species, Lipochæta connata, as
occurring in all the islands. This suspension, to a great extent, of the
means of dispersal between the islands is also strikingly illustrated by
the Lobeliaceæ.

We have only to mention the flora of Fiji and those of the adjacent
groups of Samoa and Tonga to exclude them from any share in the early
era of the Compositæ in the Pacific. The prevailing adventitious
character of the Fijian Compositæ is indicated in the fact that the
species of the majority of the genera are included by Seemann in his
list of Fijian weeds. There are only one or two Fijian Compositæ, such
as the mountain species of Lagenophora and the littoral species of
Wedelia, that merit the special attention of the student of dispersal.
So also with Samoa, Reinecke enumerates eight species, of which six are
weeds either of aboriginal or of European introduction, the others being
the littoral Wedelia above alluded to, and a species of Blumea found
also in Fiji.

DISTRIBUTION OF THE ENDEMIC GENERA OF COMPOSITÆ IN THE HAWAIIAN ISLANDS.

+--------------+-----------------------------------------------------+-------+
| | Distribution of the Species. | |
| +---------+----------+----------+----------+----------+ |
| Genus. | One | Two | Three | Four | | |
| | island. | islands. | islands. | islands. | General. | Total.|
+--------------+---------+----------+----------+----------+----------+-------+
|Remya | 2 | — | — | — | — | 2 |
|Tetramolopium | 1 | 4 | 2 | — | — | 7 |
|Lipochæta | 3 | 4 | 3 | — | 1 | 11 |
|Campylotheca | 5 | 4 | 3 | — | — | 12 |
|Argyroxiphium | 1 | 1 | — | — | — | 2 |
|Wilkesia | 2 | — | — | — | — | 2 |
|Dubautia | 4 | — | 2 | — | — | 6 |
|Raillardia | 9 | 1 | — | 2 | — | 12 |
|Hesperomannia | 1 | 1 | — | — | — | 2 |
+--------------+---------+----------+----------+----------+----------+-------+
| | 28 | 15 | 10 | 2 | 1 | 56 |
+--------------+---------+----------+----------+----------+----------+-------+

We have now, I venture to think, gone far to establish the existence of
an early “Composite” flora with mainly American affinities in the
Pacific islands, an ancient flora of which only the remnants now occur
in the uplands of Hawaii, Tahiti, and Rarotonga. That the achenes were
originally transported in birds’ plumage is, as we have seen, probable;
but we are still quite in the dark as to the causes of the subsequent
suspension of the means of dispersal and of the resulting period of
isolation, during which the original immigrant plants acquired their
endemic characters. In our uncertainty, therefore, we will look to Fiji
in the hope that in the absence of the early Compositæ from that group
we may find a clue that will enable us to divest this problem of some of
its difficulties.

It might be at first considered that since these peculiar genera of
Compositæ occur in the higher levels of Hawaii and Tahiti their absence
from Fiji might be connected with the relatively low altitude of those
islands, a character that is concerned with the exclusion from the
Fijian flora of many Hawaiian and Tahitian mountain plants (see Chapters
XXIII. and XXIV.). But this view is at once negatived by the fact that
Fitchia thrives in Rarotonga, an island which does not far exceed 2,000
feet in elevation. It is negatived also by the extensive development of
shrubby and arborescent Compositæ in the Galapagos Islands, on the
equator, in St. Helena in 16° South latitude, and in other tropical
islands, which are less than, or do not exceed, the Fijian Islands in
their altitude.

During the age of the Compositæ it is reasonable to suppose that the
dispersal was general over the Pacific. The absence of genera indicating
this era from the islands of the Fijian region, that is, from Fiji,
Tonga, and Samoa, would become intelligible if these groups were
submerged during this age of the general dispersal of the order over
this ocean. In my volume on the geology of Vanua Levu in Fiji, I have
shown that these island-groups of the Western Pacific emerged from the
sea towards the close of the Tertiary period, a conclusion that would
enable us to assign the age of the general dispersal of the Compositæ
over the tropical Pacific to an earlier portion of the same period.

In order, however, to make further progress in the discussion of this
difficult problem we are obliged to approach it from the outside. We
must in fact regard these genera from the standpoint of their position
as members of the vast and ancient order of the Compositæ. It is now
more than thirty years since Mr. Bentham completed his remarkable memoir
on the classification, history, and geographical distribution of the
Compositæ (_Journal Linnean Society, Botany_, London, Vol. 13, 1873).
Like De Candolle, when dealing with the facts of distribution, he
handled thousands of species, and as a result he drew certain inferences
which are of prime importance to students of plant-dispersal. In his
time the order included nearly 10,000 known species, and although this
number has since no doubt been considerably increased, it is not likely
that his main conclusions, in so far as they are free from purely
hypothetical considerations, will be materially affected by the later
discoveries.

Accepting the antiquity of the order, and regarding it as probably
dating far back in geological time, he observes that the evidence points
to a very wide dispersion of its original stock at an early period.
Africa, West America, and possibly Australia, possessed the order at the
earliest recognisable stage. There must have existed, he contends, at
this early period some means of reciprocal interchange of races between
these regions. Then followed a stoppage of communication, or a
suspension of the means of dispersal, between the tropical regions of
the Old and New Worlds; but long after communication was broken off in
the warmer regions, it still existed, as he holds, between the alpine
heights in those regions and also between the high northern latitudes of
both hemispheres. Referring particularly to the Hawaiian Group, he
considers that the large endemic element among the Compositæ indicates
that the ancient connection, whether with America or with Australasia,
has been so long severed as not to have left a single unmodified common
form. Fitchia, the Tahitian genus, as we have already remarked, is
regarded as the only remnant of an ancient Composite flora in the
tropical islands of the South Pacific.

In the light of these reflections it will be interesting to glance at
the general distribution of the shrubby and arborescent or woody
Compositæ. Mr. Hemsley, having generally discussed the subject, arrived
at the conclusion that, “although they form so large a proportion of the
floras of St. Helena, Juan Fernandez, the Sandwich Islands, and some
other islands, they are not specially insular.” There are scores of
them, he goes on to say, in South America, Africa, Madagascar, India,
Australia, and New Zealand from twenty to forty feet high, and more
truly arboreous than the insular ones; whilst nearly every sub-order has
its arboreous representatives. He was, however, unable to form any
definite opinion of the method of distribution of the woody Compositæ.
Taking those of St. Helena and Juan Fernandez, he observes that they are
not more closely allied to the Compositæ of the nearest continents than
they are to those of more distant regions. The occurrence of arboreous
Compositæ, belonging in each case to different tribes, in so many remote
oceanic islands, coupled with the distribution of the genera to which
they bear the greatest affinity, seems, he observes, to indicate that
they are the remains of very ancient types (_Introd. Bot. Chall.
Exped._, pp. 19-24, 66, 68; also Parts ii. p. 61, and iii. p. 23).

The further discussion of this subject would lead us into a wide field
of inquiry, quite beyond the scope of this work. There is, however, an
inference that I think we may legitimately draw from geological evidence
in this region. With respect to the antiquity of the woody Compositæ of
the Pacific as illustrated by the endemic genera, both Mr. Bentham and
Mr. Hemsley view them as belonging to ancient types. Mr. Wallace, in his
_Island Life_, a book that becomes more and more indispensable for the
student of dispersal as years progress, dwells on the importance of
these ancient Compositæ in the floral history of the Pacific islands. We
may look upon the Hawaiian Compositæ, he remarks, as representing the
most ancient portion of the existing flora, carrying us back to a very
remote period when the facilities for communication with America were
greater than they are now. The date of this period of oceanic dispersal
of the Compositæ we can now approximately determine, since these plants
are absent from the Fijian region, an area of submergence during the
Tertiary era. Before the island-groups of the Fijian region had emerged
towards the close of the Tertiary period the achenes of the early
Compositæ had been dispersed far and wide over the tropical Pacific.

But this is not all that we can infer from the convergence of these
independent lines of botanical and geological investigation. Mr. Bentham
observes that the tribes of the Compositæ had acquired the essential
characters now employed in classification before the dispersion of the
order over the Pacific. Since this general dispersion took place, as we
hold, during the Tertiary submergence of the island-groups of West
Polynesia (Fiji, Tonga, Samoa), it follows that the birth of the tribes
of the Compositæ antedates that period. If this interesting order could
supply us with a “datum-mark” in the history of the Pacific floras, it
would be stated in terms of the development of specific and generic
characters, but not of those of a tribe.

_Summary of Chapter._

(1) The Hawaiian Islands present the same contrast with the Fijian and
Tahitian groups as regards the development of new species in the case of
the flowering plants that they offer in the case of the vascular
cryptogams (ferns and lycopods). But the contrast is intensified, and it
is further emphasised as respecting the flowering plants by the
evolution of a large number of endemic genera.

(2) This great preponderance of peculiar species and genera in Hawaii is
not to be connected with the relative antiquity of the group but with
its degree of isolation.

(3) The earliest stage of the flowering plants of the islands of Hawaii
and of Eastern Polynesia (the Tahitian region) is indicated by the
endemic genera, particularly those of the Compositæ and Lobeliaceæ. Such
genera are numerous in Hawaii, and occur also in the Tahitian region, as
in Tahiti and Rarotonga; but do not exist in the groups of the Fijian
region (Fiji, Tonga, and Samoa).

(4) The endemic genera of the Hawaiian Compositæ are mainly American in
their affinities. The relationship of the solitary Tahitian genus
(Fitchia) is still a subject of discussion.

(5) In the Hawaiian Islands, as well as in Tahiti and Rarotonga, the
plants of the endemic genera of Compositæ are, as a rule, arborescent or
shrubby; and in the first two localities they are mainly restricted to
the higher levels.

(6) In discussing the mode of dispersal of the achenes of the original
genera we have also to explain why the process of dispersal has been in
the main suspended.

(7) It is shown that the achenes of these early Compositæ were in all
probability suited for dispersal in birds’ plumage.

(8) Yet the isolating influence that cut off these genera from the
outside world has, in later ages, been active within the limits of the
Hawaiian archipelago, with the result that half the species are not
found in more than a single island. Inter-island dispersal has,
therefore, been also largely suspended.

(9) The absence of endemic genera of Compositæ from Fiji, Tonga, and
Samoa cannot be attributed to unsuitable climatic conditions connected
with the relatively low elevation of those islands as contrasted with
those of Hawaii, since a species of Fitchia abounds in Rarotonga, which
is not far over 2,000 feet in elevation. Shrubby and arborescent
Compositæ of peculiar types also occur in the Galapagos and other
tropical islands not more elevated than the Fijis.

(10) These endemic genera are the remains of an ancient Composite flora
in the islands of the tropical Pacific, and ages have elapsed since the
severance of their connections with regions outside.

(11) According to Mr. Bentham the Compositæ were distributed over
Africa, West America, and possibly Australia, at an early period, but
subsequent to the differentiation of the tribes of the order. Some means
of reciprocal interchange of races between these regions then existed.
Then followed a suspension of the means of dispersal between the
tropical regions of the Old and New Worlds except between the alpine
heights of those latitudes.

(12) It is inferred by the author of this volume that the general
dispersion of the early Compositæ over the Pacific took place during the
Tertiary submergence of the island-groups of the Fijian region (Fiji,
Tonga, and Samoa), and that their absence from that region may be thus
explained. At the time of this general dispersion, as above pointed out,
the tribes of the Compositæ had been already differentiated.

CHAPTER XXII

THE ERA OF THE ENDEMIC GENERA (_continued_)

THE COMPOSITÆ AND LOBELIACEÆ (_continued_)

THE AGE OF THE TREE-LOBELIAS

THE Lobeliaceæ rank with the Compositæ in the prominence of their
position in the early Pacific floras. Though absent, as far as is known,
from Fiji, they are represented in Hawaii by 58 species, all endemic and
belonging to six genera, of which five are not found elsewhere. All
possess, as Hillebrand remarks, a woody stem, by far the greater number
being either tall shrubs, 5 or 6 feet high, or small trees, 10 to 20
feet or more in height. In the East Polynesian or Tahitian region, the
order is represented by two genera containing in all five known species
and restricted to those islands. One genus is common to the islands of
Tahiti and Rarotonga, and the other is confined to Raiatea. The species
may be shrubby or arborescent.

It was for some time considered that the oceanic archipelagoes of the
Pacific were the exclusive centres of these singular arborescent
Lobeliaceæ (I am here quoting Baillon in his _Natural History of
Plants_). And indeed this idea would receive some support from the
circumstance that Dr. Hillebrand, in his work on Hawaii, says little or
nothing about the affinities or general relations of plants which he
enthusiastically termed “the pride of our flora.” His death in 1886
deprived his work of its crowning piece, a discussion of “the
interesting questions of the origin and development of the Hawaiian
flora” (see the Editor’s Introduction, p. ix.). In no group of plants is
this want more keenly felt than with the Lobeliaceæ. Yet in his time the
explorations had yet to be made that could set the student of
plant-distribution on the road to investigate this problem.

It was true, no doubt, that types analogous to those of the Hawaiian
Lobeliaceæ were known from the American and African continents. Thus
Oliver in his _Flora of Tropical Africa_, published in 1877, gives an
account of the species of Lobelia then known from the mountains of this
region. The genus was, however, not entirely confined to mountainous
districts, but it would almost seem that most of the high mountains of
Equatorial Africa had their peculiar species, some of them being
tree-like and others shrubby. Two mountain species were recorded from
Abyssinia, one of them from an elevation of 11,000 to 13,000 feet and
growing to a height of 12 to 15 feet, the other from an altitude of
about 8,000 feet; another, Lobelia Deckenii, attaining a height of 4
feet, was recorded from the uplands of Kilimanjaro, 12,000 to 13,000
feet above the sea, and yet another from the mountains of Fernando Po,
at an altitude of 9,000 feet. So again, in the case of the American
continent, Hemsley, writing in 1885 (_Intr. Bot. Chall. Exped._, p. 32),
speaks of arborescent species of the American genera Centropogon,
Siphocampylus, &c.; and Baillon in his _Natural History of Plants_
(Engl. edit. viii. 350) refers to the similar Tupas and Haynaldias from
South America. But what the student of plant-distribution looked for was
not merely the occurrence of “tree-lobelias” in other parts of the
world, but also the reproduction of these wonderful plants under the
same conditions and on the same scale as those familiar to him on the
Hawaiian mountains. He has accordingly had to wait for the results of
the more recent explorations of the mountains of Central Africa in order
to obtain his wish.

On the upper flanks of Ruwenzori, Kilimanjaro, and Kenya, at elevations
of 9,000 to 13,000 feet and reaching to the snow-line, there flourish in
boggy portions of the forest arborescent Lobeliaceæ that attain a height
of 15 or 20 feet. They have the habit sometimes of a Dracæna and
sometimes of an Aloe, and do not exhibit the branching trunks so
characteristic of the Hawaiian genus of Clermontia. They all belong,
however, to the genus Lobelia, and thus do not display the extensive
differentiation of the endemic genera of Hawaii. Nor, apparently, has
there been the same degree of formative energy in the development of
species, since only about half a dozen species are hitherto known. We
find, however, produced on these lofty mountains of Equatorial Africa
the same climatic conditions under which the arborescent Lobeliaceæ
flourish in Hawaii, namely, the very humid atmosphere, the heavy
rainfall, and the mild temperature; and if there are important contrasts
in their character and in the amount of differentiation which they have
undergone in the two regions, the one a continental and the other an
insular region, it will be from such contrasts that some of the most
interesting results of this comparison of a mountain of Central Africa
with an island of the open Pacific will be ultimately derived (see Sir
H. Johnston’s _Uganda Protectorate_, 1902, and _Kilimanjaro Expedition_,
1886; also _Trans. Linn. Soc. Bot._, ser. 2, vol. 2, p. 341.)

THE LOBELIACEÆ OF THE HAWAIIAN AND OF THE EAST POLYNESIAN OR TAHITIAN
ISLANDS.[1]

HAWAIIAN ISLANDS.

EAST POLYNESIAN OR TAHITIAN ISLANDS.

+-----------+-----+-------------+-------------+------------+-------------+-----------+
|Sclerotheca| 4 | Endemic in |{Tahiti, | 6 to 25 | 1,500 to | Humid |
| | | E.Polynesia.|{Rarotonga. | feet. | 3,000 feet. | wooded |
| | | | | | | slopes. |
| | | | | | | |
|Apetahia. | 1 | Endemic. | Raiatea. | 3 to 6 | In the | |
| | | | | feet. | mountains. | |
| | | | | | Elevation of| |
| | | | | | island 3,400| |
| | | | | | feet. | |
+-----------+-----+-------------+-------------+------------+-------------+-----------+
Footnote 1:

The materials are nearly all derived from the works of Hillebrand and
Drake del Castillo. Some of those relating to the elevations in Hawaii
are supplemented from my notes. All the genera are endemic except
Lobelia, of which all the species are apparently endemic, excepting
perhaps one, which, according to Hillebrand, resembles greatly a
species from the Liukiu Islands.

Footnote 2:

The range of the heights of different species of Clermontia is from 5
or 6 feet for shrubs to 25 feet for trees.

Footnote 3:

The heights attained by different species of Cyanea range from 3 or 4
feet to between 30 and 40 feet, thus:—

In 8 species 3 to 6 feet.
In 9 species 6 to 10 feet.
In 7 species 10 to 15 feet.
In 3 species 15 to 25 feet.
In 1 species 30 to 40 feet.

THE LOBELIACEÆ OF THE HAWAIIAN ISLANDS.

Having thus prepared the way, I will proceed to the discussion of the
Hawaiian Lobeliaceæ, dealing first with their “station.” Their vertical
distribution is well illustrated in the large and lofty island of
Hawaii. Whilst the woody Compositæ, as before described, are most at
home on the open-wooded and often scantily-forested slopes between 5,000
and 9,000 feet, the Lobeliaceæ are most characteristic of the middle or
true forest zone that extends from 2,000 or 3,000 feet to between 5,000
and 6,000 feet above the sea. This lies within the region of clouds and
mists, and it is here that the rain-belt or area of greatest rainfall is
situated, the annual amount averaging probably 150 to 200 inches. It is
in such humid conditions that, as Hillebrand observes, trees and jungle
are developed in greatest luxuriance; and it is here that “the
Lobeliaceæ exhibit their most striking forms.” The traveller, as he
ascends the mountains, finds the Tree-Lobelias in the region of mist and
rain-cloud; and he is lucky if he escapes the usual downpour and
encounters only a fine drizzling rain.

The mild climate of this region is indicated by a mean annual
temperature ranging probably with elevation from 65° to 55° F. It is
secure from the frosts of the upper slopes of the mountain; whilst at
the same time it is above the regions of tropical heat. There is,
however, no doubt that when the forests extended to the coasts, as they
occasionally do now on the north side of Hawaii, the Lobeliaceæ occurred
much lower down than they do at present, though still only attaining
their greatest development in size and number in the higher levels.
Thus, at rare intervals, I noticed in the forests of Hamakua and Kohala,
where they descended to the coasts, species of Clermontia at an
elevation of only 500 or 600 feet above the sea.

Probably in no part of the Hawaiian Islands are the conditions under
which the “Tree-Lobelias” thrive better illustrated than on the higher
slopes of Mount Eeka, a bulky mountain mass about 6,000 feet in height,
forming the western portion of Maui. Its flat top, as Hillebrand
observes, is wrapped in a cloud of mist nearly the whole year. On the
boggy surface of the summit, where Acæna exigua gives a tussocky
appearance, and Sphagnum or bog-moss abounds, flourish Cyperaceæ,
Lycopods, and Selaginellæ; and here Drosera longifolia and a peculiar
species of marsh violet (Viola mauiensis) find a home. The upper slopes,
down to 4,000 feet, present similar moist conditions, and here in an
open-wooded district, associated with Cyrtandræ, Marattias, and true
Tree-Ferns, the ground being covered with Lycopods, the “Tree-Lobelias”
abound. I noted four kinds within two hundred yards. Of the humidity of
the upper slopes of Mount Eeka I have a very vivid recollection, and my
experience of passing a night on that mountain is described in Chapter
XIX.

The Lobeliaceæ, as Hillebrand remarks, occur invariably as isolated
individuals. I was often struck, however, with the preference the genera
showed for particular localities. Thus, Clermontia is well represented
on the western slopes of Mount Eeka, Delissea on the northern slopes of
Hualalai (3,800 to 4,500 feet), Cyanea on the Hamakua slopes of Mauna
Kea (2,300 to 4,100), and Lobelia on the southern slopes of Mauna Loa
behind Punaluu (2,000 to 3,500 feet).

To the student of geographical distribution the Hawaiian Lobeliaceæ are
of especial interest. Mr. Hemsley observes that they have their greatest
affinities in America (_Intr. Bot. Chall. Exped._, p. 68). M. Drake del
Castillo, in his “Mémoire couronné par l’Académie des Sciences” (Paris,
1890), remarks that these plants connect Hawaii with America just as the
Goodeniaceæ link the same group with Australia. This is what we might
have expected since the centre of the order is in America, principally
in the Mexican and Andine regions (Drake del Castillo, _Flore Polyn.
Franc._, xi.).

Though five out of the six genera are endemic, the sixth, that of
Lobelia, has a world-wide distribution. Here then, we have a genus that
belongs strictly to the next or second stage of the plant-stocking of
the Hawaiian Group, namely, when the non-endemic genera now containing
endemic species were introduced. As with the Composite genera,
Campylotheca and Lipochæta, Lobelia marks the beginning of the new or
the close of the old era. It is, however, necessary to point out that
many of the conditions favouring luxuriant and rank vegetable growth are
pre-eminently represented in the zone of the Lobeliaceæ. In these
soft-stemmed plants with their copious milky sap and large fleshy
flowers, sometimes two or three inches long, the very redundancy of
growth would tend both to exaggerate and to disguise the generic
distinctions. To the ordinary observer these “Tree-Lobelias” call up
vague notions of a flora of a bygone age, and by their _bizarre_
appearance he might with some excuse be led to give play to his
imagination when describing them; but the systematic botanist, seeing
through their disguise, frames rather more prosaic notions of their
antiquity and degree of differentiation. According to my view, the first
Hawaiian Lobeliaceæ occupied open, exposed localities such as are held
by the decadent genus Brighamia now, and acquired their monstrous form
in the humid forests of a later age. (See Perkins in Note 80.)

In his monograph on the Campanulaceæ (Engler’s _Nat. Pflanz. Fam._, teil
4, abth. 5, 1894), S. Schönland, speaking of the sub-family Lobelioideæ,
places the seven endemic Hawaiian and Tahitian genera in a group by
themselves. Though, as he observes, the Hawaiian tree-forms appear at
first sight to constitute a natural group, they cannot be sharply
distinguished from other forms, and even in habit come near some Indian
and Abyssinian types of Lobelia. In their treatment, he says, they
should all go together, and he does not approve of the endeavours of
some botanists to isolate one of them (Brighamia) from the rest and to
connect it with the Australian genus Isotoma.

It is also to be noted that whilst four of the Hawaiian genera are more
or less dispersed over the group, one (Brighamia) with only one species
is confined to the islands of Molokai and Niihau, the double habitat
being suggestive of its approaching extinction. Another (Rollandia) with
six species is restricted to the island Oahu. Cyanea, which possesses
twenty-eight out of the total of fifty-eight species, may, from the
point of view of its formative energy, be regarded as in its prime. It
is thus apparent that, as with the Compositæ, the early Lobeliaceous
immigrants were not all contemporaneous arrivals.

DISTRIBUTION OF THE LOBELIACEÆ IN THE HAWAIIAN ISLANDS.[4]

Column headings:

A: Brighamia.
B: Lobelia.
C: Clermontia.
D: Rollandia.
E: Delissea.
F: Cyanea.
G: Total.

+----------------------------------+----+----+----+----+----+----+----+
| Hawaiian Lobeliaceæ. | A | B | C | D | E | F | G |
+----------------------------------+----+----+----+----+----+----+----+
| Species confined to one island | — | — | 6 | 6 | 4 | 22 | 38 |
| Species confined to two islands | 1 | 2 | 2 | — | 2 | 5 | 12 |
| Species confined to three islands| — | 1 | 2 | — | 1 | 1 | 5 |
| Species generally distributed, | | | | | | | |
| but still endemic | — | 2 | 1 | — | — | — | 3 |
| +----+----+----+----+----+----+----+
| | 1 | 5 | 11 | 6 | 7 | 28 | 58 |
+----------------------------------+----+----+----+----+----+----+----+
Footnote 4:

All the species are endemic.

Another interesting fact of distribution, brought out by an analysis of
Hillebrand’s materials and illustrated in the subjoined table, is that
out of the fifty-eight Hawaiian species, all of which are endemic,
thirty-eight, or 66 per cent., are recorded from only one island. In
most of the other cases they are recorded from two or three islands,
usually adjacent, like Maui and Molokai; and except in the instance of
two species of Lobelia and one species of Clermontia they never range
over the length of the group.

These facts speak eloquently of the suspension to a great extent of the
agencies of dispersal in recent times within the group. Some corrections
of the figures will be rendered necessary by future investigations, but
the main conclusion will not be materially affected. Such facts are
paralleled in the distribution of the Hawaiian insects, mollusca, &c.;
but these matters need only be mentioned here. We might, indeed, have
expected, apart from other considerations, that the isolation of the
Hawaiian Lobeliaceæ from their kindred in other parts of the world would
not have been reproduced within the group itself. This, however, is not
the case; and we now see that not only have they been deprived for ages
of their means of distribution over the Pacific, but that even within
the archipelago their transportal from island to island has been largely
suspended. We have before arrived at similar conclusions with regard to
the early Compositæ, when we saw that about half the species were not
found in more than one island. It is therefore evident that the same
great principle regulating the operations of the distributing agencies
has influenced to a similar extent both the Compositæ and the Lobeliaceæ
of the Hawaiian Group.

THE LOBELIACEÆ OF THE TAHITIAN OR EAST POLYNESIAN REGION.

The order is represented in this region by two endemic genera,
Sclerotheca of Tahiti and Rarotonga, and Apetahia of Raiatea. These
islands are, however, not sufficiently large for the extensive
development of the arborescent Lobeliaceæ, such as we find in Hawaii.
The species in both genera are either arborescent or shrubby; but I do
not gather that they give any character to the floras of these islands.
According to the data given by Drake del Castillo for one of the two
peculiar species of Sclerotheca occurring in Tahiti, these plants grow
on the humid wooded slopes of the mountains at elevations of 2,000 to
3,000 feet. Whilst in one species the plants attain a height of 10 to 25
feet, in the other they do not exceed 10 feet. Rarotonga possesses a
peculiar species of Sclerotheca, 4 to 6 feet high, which was discovered
by Cheeseman growing plentifully on the upper slopes of the highest
mountain of the island at altitudes of 1,500 to 2,200 feet. The same
botanist also came upon a second species of the genus on another
mountain in Rarotonga at elevations of 1,000 to 1,500 feet, but it was
rare and has not yet been described. The other genus, Apetahia, has only
been recorded from Raiatea, where it is represented by a solitary
species (6 feet high) growing, according to Nadeaud, in the mountains of
that island.

It is apparent that the dispersal of these genera of the Lobeliaceæ
amongst the groups of Eastern Polynesia ceased long ago. From the
circumstance that Sclerotheca exists in Tahiti and in Rarotonga, which
are about 650 miles apart, it may be inferred either that the genus was
introduced into this region from outside, or else, which is perhaps more
probable, that it was developed in Tahiti whence it was transported to
Rarotonga. Hemsley speaks of this Tahitian genus as seemingly marking a
former wide extension of the Hawaiian arborescent type of the Lobeliaceæ
(_Introd. Bot. Chall. Exped._, p. 68). This is the view that will be
adopted in this chapter, and it is precisely the view advocated by
Bentham and followed here, in the case of the early Compositæ of the
Pacific.

With regard to the absence of these arborescent Lobeliaceæ from the
island-groups of the Western Pacific, and notably from Fiji and Samoa,
where no members of the order seem to occur, it is probable that, as in
the case of the similar distribution of the early Compositæ described in
the preceding chapter, this is to be attributed to the fact that the
Western Pacific archipelagoes were more or less submerged during the
general dispersion of the Compositæ and Lobeliaceæ over the Pacific in
the earliest age of the floral history of these islands. The occurrence
of the early Compositæ and Lobeliaceæ in Rarotonga, which is almost
half-way between Tahiti and Tonga on the outskirts of the Fijian region,
sufficiently indicates that they are not lacking in that region from
inability to reach there in the past. During the age of general
dispersal of these two orders over the Pacific, probably only a few
rocky islets, tenanted perhaps by Conifers, marked the situation in the
Tertiary period of the present archipelagoes of Fiji and Samoa.

One may note in passing the general absence of these arborescent types
of the Lobeliaceæ from Malaya, since they do not seem to have been
recorded either from the Owen Stanley Range in New Guinea or from
Kinabalu in North Borneo, the highest mountain in the Malayan Islands,
or from the mountains of Java.

The consideration of the occurrence of these plants in other tropical or
subtropical oceanic islands need not detain us long, since, with the
exception of the solitary Lobelia scævolifolia of St. Helena, they seem
rarely to be found. This species, which is endemic, is a shrub growing
on the upper slopes and summit of the island at elevations of 2,000 to
2,700 feet (_Introd. Bot. Chall. Exped._, p. 40, and Part ii. pp. 54,
76).

There are two herbaceous species of Lobelia in Juan Fernandez, of which
one only, according to Hemsley, could be regarded as indigenous. This is
a showy Chilian and Peruvian species (Lobelia tupa) noticed by Bertero
as very common in 1829 (_Bot. Chall. Exped._, Part iii.). Since,
however, it would belong to the present age of plant-dispersal in the
Pacific, it does not require further mention here; and indeed it would
almost appear, when we bear in mind the geographical position and the
history of this island since its discovery in 1563, that even as a truly
indigenous plant it is not above suspicion. Lobelias of this type are
now amongst the commonest plants of the coast regions of northern Chile,
where I noticed some as much as 9 or 10 feet high.

_On the Capacities of Dispersal of the Lobeliaceæ of the Pacific._—Of
actual observations, with the exception of the instance of birds pecking
at the capsules of our garden Lobelias, I have come upon few that bear
directly on this point. When writing of the flora of the Kermadec Group,
many years ago, Sir Joseph Hooker referred (_Journ. Linn. Soc. Bot._,
vol. i.) to the minute seeds of Lobelia as not adapted for transport
unless their minuteness and number fit them for it; but since he
associates in this connection the tiny seeds of Metrosideros, which is
now represented by a species found all over the Pacific, it would seem
that the difficulty in the case of Lobelia is not connected so much with
the nature as with the suspension of these means of distribution during
the later stages of the plant-stocking of the oceanic islands of the
tropical Pacific. It will be gathered from the following remarks that
the descendants of the early Pacific Lobeliaceæ are probably as well
fitted for dispersal as their ancestors, and that the break in the
communication is the ultimate subject for inquiry.

The fruits of the Hawaiian endemic genera are in four out of five cases
baccate, with usually fleshy or pulpy contents. Such berries, which are
generally yellow, but sometimes bluish in colour, vary in size from
about half an inch in Rollandia and Delissea to an inch in Cyanea, and
not infrequently to more than an inch in Clermontia. The fruits of
Lobelia and Brighamia are capsular and dehiscent. With regard to the two
genera of the Society Islands and Rarotonga, the fruits of Sclerotheca
are hard-walled capsules, opening by two pores; whilst those of Apetahia
are seemingly dry and indehiscent. I do not imagine, therefore, that the
character of the fruit has determined to any important degree the
distribution of these plants.

Nor is there reason to suppose that the fruits have acquired their
baccate character in Hawaii, and that they were originally dry and
capsular. Both types of fruit are found among the arborescent Lobeliaceæ
of America, with which the Hawaiian genera have their affinities.
Centropogon, for instance, which occurs in Central America and in the
warm parts of North and South America, has, according to Baillon, a
somewhat fleshy berry. It is noteworthy that a similar question is
raised with respect to Cyrtandra as to the relation between fleshy
fruits in the Pacific islands and dry or capsular fruits in the
continental home of the genus (see Chapter XXV.).

The berries of the Tree-Lobelias would attract birds. We learn from Mr.
Perkins that one of the Hawaiian Drepanids, the Ou, is very partial to
the berries of some of the Tree-Lobelias and especially those of
Clermontia, the seeds passing unharmed in the droppings. The mode of
dispersal of the seeds of the dry-capsular fruits is not so apparent;
but the fruits could scarcely be less inviting to birds than the dry
capsules of Metrosideros, the small seeds of which have in some way or
other been carried to almost every island-group of the Pacific. I have
beside me the dark brown, smooth crustaceous seeds of a species of
Clermontia. They measure 1/42 of an inch or 0·6 of a millimetre, and
about 500 go to a grain. Mr. Wallace, in his book on Darwinism,
advocates the paramount influence of winds over birds for carrying small
seeds, like those of Orchis and Sagina, over tracts of ocean a thousand
miles across. I am, however, not inclined to think that, except as
regards the spores of cryptogams, winds have done very much for Hawaii.
For small seeds we can appeal not only to the agency of birds and bats
but also to insects (see Chapter XXXIII.).

Observations of this kind, however, merely indicate that these early
Lobeliaceæ possessed the same capacities for dispersal that in the
succeeding stages of the plant-stocking of the Pacific islands have
belonged to Metrosideros, Cyrtandra, Ophiorrhiza, Freycinetia, and many
other small-seeded genera. They go no way to explain why the same
agencies which transported the minute seeds in a later age could not
have been available for continuing the dispersal of the early
Lobeliaceæ. To find an explanation we are compelled to go behind the
mere capacities for dispersal and to appeal to the general laws of
distribution in so far as our facts enable us to interpret them.

We have seen that the two principal components of the early Pacific
flora, the Compositæ and the Lobeliaceæ, have American affinities. The
plants of the later ages are mainly Old World in their connections.
Though containing often endemic species in the various groups, the
genera occur also outside each group. The stream of migration that came
from America during the early age of the Compositæ and the Lobeliaceæ,
when the islands of the Western Pacific were more or less submerged, was
during the later ages (after these islands had re-emerged) suspended or
diverted, giving place to a stream that brought plants in numbers from
tropical Asia, Malaya, and Australia. The general dispersion of the
Compositæ and Lobeliaceæ took place during the Tertiary submergence of
the islands of the Western Pacific, including the island-groups of Fiji,
Samoa, and Tonga. The migration from the west, mainly Indo-Malayan in
character, occurred after the re-emergence of those archipelagoes. Thus
we get to understand how genera like those of the early Lobeliaceæ and
Cyrtandra, which possess, as regards the minute size of their seeds,
closely similar capacities for dispersal, have such different
distributions, the first confined to Hawaii and Tahiti and American in
their affinities, the second widely spread over the Pacific with its
home in Malaya.

We have yet to inquire whether this suspension of the means of transport
in the later ages of the Pacific Lobeliaceæ is confined to the tropics
or whether it extends to the colder latitudes in the southern
hemisphere. The indications of the Lobeliaceæ of the “antarctic flora”
go to establish that the dispersal of the order is still, or was very
recently, in operation in these high latitudes. It is well illustrated,
among other plants, by Lobelia anceps, which is found in extra-tropical
South America, Australia and New Zealand, and South Africa. This,
indeed, recalls Bentham’s view concerning the Compositæ, that whilst
communication was broken off in the tropics, it was kept up in higher
latitudes.

Here ends, therefore, our consideration of the Tree-Lobelias of the
Pacific islands; but as it is not quite complete without a discussion of
the remaining endemic genera of other orders than the Compositæ and
Lobeliaceæ which also belong to the same early age of the Pacific
floras, I will proceed at once to their consideration.

THE HAWAIIAN ENDEMIC GENERA EXCEPTING THOSE OF THE COMPOSITÆ AND
LOBELIACEÆ.

It will not be possible for me to do more than point out a few general
indications that can legitimately be drawn from these genera. The
subject bristles with difficulties for the systematist; but on one point
there can be but little danger of going astray, namely, in imputing to
them a high antiquity in the floral history of Hawaii. This can be said
of all of them, whether or not the generic distinction adopted in Dr.
Hillebrand’s work is always adopted by botanists. It is therefore in
this general sense that they may be regarded as belonging to the early
age of the Hawaiian flora.

Although the genera of Compositæ and Lobeliaceæ are prominent amongst
the representatives of the original flora of the Hawaiian Islands,
forming about two-fifths of the whole, the genera of other orders are by
no means inconspicuous, and their variety is shown in the fact that
though twenty-three in number they belong to twelve orders. It is
possible to divide these genera into two groups—one the older and
perhaps more or less contemporaneous with the Lobeliaceæ and Compositæ,
the affinities when apparent being American; the other the more recent
and marking the close of the first era of the plant-stocking of these
islands, the affinities being all with the Old World, and especially
with Malaysia. This grouping is indicated in the list subjoined; and it
may be here remarked that whilst shrubs, undershrubs, and perennial
herbs of the Caryophyllaceæ, Labiatæ, and Urticaceæ form the features of
the earlier group, trees of the Rubiaceæ and Araliaceæ are the most
conspicuous members of the later group. At the close of the earliest era
known to us of the floral history of the Hawaiian Islands we observe the
commencement of those forests that now throughout Polynesia as well as
in Hawaii betray their Asiatic origin.

In making this distinction I am proceeding on the assumption that the
stream of migration, at first chiefly American in its source, came
ultimately in the main from the Asiatic side of the Pacific. The change
commenced, as I hold, in the latter portion of the first era of
plant-stocking, an era characterised by the arrival of those early
plants that are now represented by the endemic genera of the
archipelago. The genera of this early period that belong neither to the
Compositæ nor to the Lobeliaceæ are, as above observed, arranged by me
in two groups, one regarded as contemporaneous with, the other as of
later origin than, the genera of these two orders. To the first belong
the shrubby, highly differentiated genera of the Caryophyllaceæ,
Schiedea and Alsinidendron, and the Labiate genera, similarly
differentiated, of Phyllostegia and Stenogyne. To the second belong the
Rubiaceous genera Kadua, Gouldia, Bobea, and Straussia, the Araliads
Cheirodendron, Pterotropia, and Triplasandra, and the Loganiaceous
Labordea.

In the earlier group the fruits are dry in half the genera, and in such
cases granivorous birds probably were usually the transporting agents.
Only in one case (Nothocestrum) is the fruit a berry, and in the other
cases we have fruits like the fleshy nucules of Phyllostegia and
Stenogyne which would probably attract birds. In the later group
two-thirds or three-fourths of the genera have moist fruits such as
would be eaten by frugivorous birds. Of these most are drupes,
possessing not a single stone, but two or more pyrenes. This is the
first appearance of the drupe in the plant-history of the archipelago.
The Rubiaceous type of drupe inclosing two or more pyrenes plays a very
conspicuous part in the distribution of plants over the Pacific in the
succeeding eras.

I would here lay stress on an important characteristic of all the fruits
of the endemic genera of the Hawaiian Islands. There are no “impossible”
fruits of this era in Hawaii, such as we occasionally find in the
succeeding eras. I mean by this term, fruits that defy the efforts of
the student of distribution to explain their transport in their present
condition. The discovery of a new inland genus possessing dry
indehiscent fruits three or four inches long, or even of a single
species of the coniferous Dammara, would play havoc with all our views
respecting the stocking of these islands with their plants. The finding
here of a large marsupial would scarcely produce more astonishment. The
fruits indeed of this early era are very modest in their size, the dry
indehiscent fruits and the stone-fruits rarely exceeding half an inch
(12 mm.) in size.

There is another interesting point which is connected with the
deterioration of some of the fruits in their capacity for dispersal.
Some of the species of Phyllostegia, and a few also of the Araliads, as
well as those of Nototrichium, are ill fitted for dispersal by birds
now, the coverings of the seeds being not sufficiently hard to protect
them from injury in a bird’s stomach. At the same time there are in some
cases other species of the same genera that are better suited for this
mode of transport. The effect of dispersal by frugivorous birds is that
only the hard-coated seeds propagate the plant in a new locality. When,
however, as has occurred in the Hawaiian Islands, bird-agency largely
ceases to act, this selective influence is removed (see Note 68).

ENDEMIC HAWAIIAN GENERA, EXCLUDING THOSE OF THE COMPOSITÆ AND
LOBELIACEÆ, AS GIVEN IN HILLEBRAND’S “FLORA OF THE HAWAIIAN ISLANDS.”

[Those preceded by * are not usually regarded now by botanists as
endemic, though they nearly
take that rank.]

THE EARLIER GROUP.

+-----------------+---------------+---------+------------+-----------------+-----------------------------+
| Genus. | Order. |Number of|Character. | Fruit. | Affinities. |
| | | species.| | | |
+-----------------+---------------+---------+------------+-----------------+-----------------------------+
|Isodendrion | Violaceæ. | 3 |Shrubs. |Capsule. |American (H). |
| | | | | | |
|Schiedea |Caryophyllaceæ.| 17 |Undershrubs,|Capsule. {|Near Colobanthus of the |
| | | | &c. | {| Antarctic islands, temperate|
|Alsinidendron |Caryophyllaceæ.| 1 |Undershrubs.|Capsule, with {| South America, |
| | | | | fleshy calyx. {| and Australia (C). |
|Platydesma | Rutaceæ. | 4 |Small trees |Capsule. | — |
| | | | or shrubs. | | |
|Hillebrandia | Begoniaceæ. | 1 |Herbs. |Capsule. | — |
|Nothocestrum | Solanaceæ. | 4 |Small trees.|Berry. |South American (H). |
|*Haplostachys | Labiatæ. | 3 |Herbs. |Dry nucules. |Regarded by Gray as a |
| | | | | | section of Phyllostegia. |
|*Phyllostegia | Labiatæ. | 16 |Undershrubs.|Fleshy nucules. {|Belong to the tribe Prasiæ, |
| | | | | {| which is mostly Asiatic. |
|Stenogyne | Labiatæ. | 17 |Trailers or |Fleshy nucules. {| Two other species of |
| | | | climbers. | {| Phyllostegia recorded |
| | | | | {| from Tahiti and Paumotu |
| | | | | {| Islands. |
|Charpentiera | Amarantaceæ | 2 |Trees. |Utricle. |American (H). |
|Touchardia | Urticaceæ. | 1 |Shrubs. |Achene with | — |
| | | | | fleshy perigone.| |
|Neraudia | Urticaceæ. | 2 |Shrubs. |Achene with |Allied to Bœhmeria, a |
| | | | | fleshy perigone.| genus of Old and New |
| | Worlds. |
| THE LATER GROUP. |
| | | | | | |
|*Pelea | Rutaceæ. | 20 |Trees. |Capsular. |Belongs to Melicope, an |
| | | | | | Old World genus (IK). |
|Broussaisia | Saxifragaceæ. | 2 |Small trees.|Berry. |Malayan (H). |
|*Cheirodendron | Araliaceæ. | 2 |Trees. |Drupe. |Referred to Panax, an |
| | | | | | Old World genus (IK). |
|*Pterotropia | Araliaceæ. | 3 |Trees. |Drupe. {|Malayan (H). |
| | | | | {| Pterotropia referred to |
|Triplasandra | Araliaceæ. | 4 |Trees or |Drupe. {| Heptapleurum of Old |
| | | | shrubs. | {| World (IK). |
|Kadua | Rubiaceæ. | 16 |Shrubs, &c. |Capsular |Approaches both Asiatic |
| | | | | | and American types (C). |
|Gouldia | Rubiaceæ. | 5 |Small trees |Drupaceous |American (C). |
| | | | or shrubs. | berry. | |
|*Bobea | Rubiaceæ. | 5 |Small trees.|Drupe. |Malayan (H). Genus |
| | | | | | also in Malaya (IK). |
|Straussia | Rubiaceæ. | 5 |Trees. |Drupe. |Near Psychotria, a genus |
| | | | | | of Asia and America (H). |
|Labordea | Loganiaceæ. | 9 |Small trees |Capsule with |Malayan (H). |
| | | | or shrubs. | pulp. | |
|*Nototrichium | Amarantaceæ. | 3 |Trees or |Utricle. |Referred to the Australian |
| | | | shrubs. | | Ptilotus (IK). |
+-----------------+---------------+---------+------------+-----------------+-----------------------------+

(H) = Hillebrand’s _Flora of the Hawaiian Islands_.
(C) = Drake del Castillo’s _Remarques sur la Flore de la Polynésie_.
(IK) = _Index Kewensis._

NOTE.—Probably Schumann’s genus, Pteralyxia, should be placed in the
later group (see p. 154).

Another feature of interest is to be found in the distribution within
the archipelago of the species of the peculiar genera. As in the case of
the Compositæ and Lobeliaceæ, but few of the species are generally
distributed, most being restricted to one island or to two or three
adjacent islands. The suspension of the dispersal among the islands is,
however, not so marked as with the species of the two orders just named.

NOTE.—Some further remarks on some of these genera are given in Note 68.

THE ENDEMIC GENERA OF THE FIJIAN ISLANDS.

The interest that is associated with the endemic genera of Hawaii fails
to attach itself to those of Fiji. For this there are several reasons.
In the first place, our acquaintance with the Fijian flora is much less
complete. In the next place, the group holds a much less isolated
position, and the history of an endemic genus may have a significance
quite different from that connected with it in Hawaii. Fiji also lacks,
on account of its submergence in the Tertiary period, those highly
interesting genera of the Compositæ and Lobeliaceæ that form the chief
feature in the early history of the flowering plants of Hawaii. Then,
again, on account of our imperfect knowledge of the floras of the
neighbouring groups of continental islands to the westward, the New
Hebrides, Santa Cruz, and Solomon Groups, we can never feel quite
confident that any particular genus is really peculiar to the Fijian
archipelago. This is well brought out in the later history of the genera
designated by Dr. Seemann in his _Flora Vitiensis_ as peculiar to Fiji.

Of the sixteen genera enumerated by Dr. Seemann, and given in the table
below, only about half now retain their character of being restricted to
Fiji. Nor does it seem likely that future investigations will increase
this number, since, judging from a remark made by Mr. Hemsley in his
paper on the botany of the Tongan Group, explorations subsequent to
those of Dr. Seemann, more especially those of Mr. Horne, have not
apparently added a single new endemic genus to the Fijian flora. It will
be seen from the list that at least four of the sixteen genera have
since been found in the Malayan region, and in one case (Smythea
pacifica) the same species occurs in both regions; whilst a fifth genus
(Haplopetalon) has been recorded from New Caledonia.

There are, however, some peculiarities about the Fijian endemic genera
that will attract our attention from the standpoint of dispersal. One
remarkable feature is the paucity of species. Almost all the genera are
monotypic, that is to say, they are only known by a single species.
Amongst the twenty-eight Hawaiian genera that are strictly endemic, only
four or five are monotypic, and they are mostly regarded by Hillebrand
as worn-out, decadent types found in only one or two islands. In Hawaii
there are on the average six species to each endemic genus; and it is
thus apparent that in the display of formative energy Nature has worked
on very different lines in these two groups. Since the nine Fijian
endemic genera belong to nearly as many different orders, the
composition of this endemic generic flora is by no means homogeneous. It
is, I venture to think, such a motley collection as one might expect in
a region that has been exposed to wave after wave of migration from the
west, with no lofty mountains, as in Hawaii, to afford a refuge against
extinction. It by no means follows that all these endemic genera have
been produced in Fiji. Some of them may represent genera that have
become extinct in the large continental groups to the westward.

SEEMANN’S SIXTEEN FIJIAN ENDEMIC GENERA.

Those genera marked * have since been found outside the group.

The authorities are thus indicated: (C)=Drake del Castillo; (H)=Horne;
(IK)=_Index Kewensis_ (S)=Seemann; (Sc)=Schimper; (So)=Solereder in
Engler’s _Nat. Pflanz. Fam._

The fact that several of them are fitted for dispersal by frugivorous
birds is very suggestive of the lack of means of transport in later
times. In the instance of Couthovia corynocarpa the drupes are known to
be the food of fruit-pigeons at the present time (Seemann), whilst this
is also true of Oncocarpus vitiensis, though this genus has since been
found in New Guinea. Since, as will be pointed out in a later chapter,
birds must still be fairly active in carrying seeds to Fiji from regions
westward, it would seem that genera only become peculiar to Fiji when
they fail at their source, and it is indeed doubtful whether any of the
Fijian peculiar genera are home productions. One may instance in this
connection the genus Pimia, the fruits of which are especially well
suited for attachment to a bird’s plumage, yet it is only known from
Fiji.

It should be here observed that no peculiar generic types have been
recorded from the adjacent Tongan Group, and scarcely any from Samoa.
Except perhaps with the Palmaceæ, no peculiar genera seem to be
mentioned in Dr. Reinecke’s memoir on Samoa.

_Summary._

(1) The Lobeliaceæ, like the Compositæ, take a prominent place in the
early Pacific flora, being represented, more particularly in Hawaii but
also in the East Polynesian or Tahitian region, by endemic genera of
tall shrubby and tree-like species.

(2) Tree-Lobelias occur in other parts of the world, as in South America
and tropical Africa; but it is especially on the higher slopes of the
mountains of Equatorial Africa that they attain a development comparable
with that of Hawaii.

(3) In Hawaii the Tree-Lobelias are most characteristic of the middle
forest-zone (3,000-6,000 feet), where the temperature is mild, the
rainfall heavy, and the atmosphere laden with humidity.

(4) The affinities of these endemic genera of the Lobeliaceæ are mainly
American; but their generic distinctions have been both exaggerated and
disguised by redundant growth.

(5) From the distribution of the genera and species within the Hawaiian
Group it is evident that, as with the early Compositæ, the original
Lobeliaceous immigrants were not all contemporaneous arrivals. Some of
the genera are on the point of extinction, whilst others are in their
prime.

(6) The absence of the Lobeliaceæ from the groups of the Fijian area
(Fiji, Tonga, Samoa) is probably to be connected, as in the case of the
absence of the early Compositæ, with the circumstance that the general
distribution of these two orders over the tropical Pacific occurred
during the Tertiary submergence of these archipelagoes.

(7) These endemic genera of the Lobeliaceæ possess the same facilities
for dispersal that are owned by other genera with minute seeds, such as
Cyrtandra, &c., that are dispersed over the Pacific; but in the case of
the Lobeliaceæ the agencies of dispersal have been for ages suspended.

(8) This suspension is to be associated with the diverting of the main
stream of migration from its source in America, during the early age of
the Lobeliaceæ and Compositæ, to a source on the Asiatic side of the
Pacific.

(9) The Hawaiian endemic genera other than those of the Compositæ and
Lobeliaceæ arrange themselves in two groups—an earlier group containing
highly differentiated Caryophyllaceæ and Labiatæ, and belonging to the
age of the Compositæ and Lobeliaceæ; and a later group, characterised by
Rubiaceæ and Araliaceæ, which marks the close of the first era, as well
as the change in the main source of the plants from America to the Old
World, the beginning of the Hawaiian forests, the appearance of the
Rubiaceous drupe, and the first active intervention of frugivorous
birds.

(10) Though there are no “difficult” or “impossible” fruits (fruits, the
dispersal of which is not easy to explain) amongst the forty and odd
endemic genera of Hawaii and Tahiti, it is noteworthy that in some cases
the fruits are seemingly little fitted for dispersal now, and that this
deterioration in capacity for dispersal is to be frequently associated
with more or less failure of the inter-island dispersal in the case of
Hawaii.

(11) The interest associated with the Hawaiian endemic genera fails to
attach itself to those of Fiji, where genera only seem to have become
peculiar because they have failed at their sources in the regions to the
west. The endemic genera of the Compositæ and Lobeliaceæ are here
lacking, and this is true also of the neighbouring Samoan and Tongan
Groups, it being held that the age of the general dispersion of these
orders over the Pacific corresponded with the Tertiary submergence of
the archipelagoes of the Western Pacific. Those of Fiji, which do not
amount to ten in number, belong to nearly as many orders and present a
motley collection such as one might look for in a group much less
isolated than Hawaii and exposed to wave after wave of migration from
the west.

CHAPTER XXIII

THE ERA OF THE NON-ENDEMIC GENERA OF FLOWERING PLANTS

THE MOUNTAIN-FLORAS OF THE PACIFIC ISLANDS AS ILLUSTRATED BY THE
NON-ENDEMIC GENERA

THE AGE OF THE ENDEMIC GENERA OF FLOWERING PLANTS.

WE are now entering an era distinguished from the preceding age of the
endemic genera, the age chiefly of the Compositæ and Lobeliaceæ, by the
fact that the extreme isolation that followed that era no longer
prevails. In a sense these island-floras are in touch again with the
world around, though the main stream of plant-migration now comes from
the south and from the west. Yet in a large number of cases, the amount
varying greatly in the different groups, it is evident that this stream
has not flowed continuously to the present day. The agencies of
dispersal are often no longer active; but the period of inactivity has
not been sufficiently prolonged to produce generic distinction, and the
differentiating energy has been restricted to the development of new
species.

Yet within these limits the development of new forms, as indicated in
Table B on p. 233, has often been very great. Thus, nearly half the
Hawaiian genera that are non-endemic are composed entirely of species
not found outside the group; and in this sense they may be regarded as
cut off from the regions around. In Fiji and Tahiti only about a fourth
are in this manner isolated, the agencies of dispersal being still
effective with the majority of the genera. It is apparent, therefore,
that the same question concerning the cause of the failure of the means
of dispersal presents itself in this era as in the last, and most
markedly in the instance of Hawaii.

The simplest and quickest plan for bringing into relief the prominent
features of this age is first to regard the genera from the standpoint
of the elevation of their stations. We have before remarked that in the
occurrence of extensive regions of great altitude the Hawaiian Islands
differ conspicuously from the groups of Tahiti and Fiji (and I may add
Samoa); and that they present conditions for the development of a
temperate mountain-flora that are not found at all in Fiji and are
barely represented in Tahiti. That the Hawaiian flora responds to this
contrast between the elevations of the three groups is well established;
and I will now proceed to refer more in detail to the subject.

THE MOUNTAIN-FLORAS OF THE PACIFIC ISLANDS.

In the Hawaiian Islands there are at least 37 or 38 genera, making up
about 19 or 20 per cent. of those belonging to this era, that may be
designated mountain genera, nearly all of them being characterised as
appertaining exclusively or in the main to temperate regions, or as
frequenting mountain-tops in tropical latitudes. In Tahiti there are
only 7 or 8 of such genera, about 4 per cent. of the total for the era.
In Fiji, excluding the Conifers, there are only 4 or 5, or not 2 per
cent. of the whole. In Samoa, which may be included in the Fijian area,
there are 3, or about 2 per cent. of the total. These are results which
we might have expected from the varying altitudes of these groups, as
described in Chapter XIX.

Few things give more pleasure to the botanist than his recognition in
some remote locality of plants long familiar to him in other regions.
This will often be his lot on the mountain summits of Hawaii. If he has
been a mountain-climber in many countries, he will there notice again
the genera Artemisia, Geranium, Plantago, Ranunculus, Rubus, Sanicula,
Vaccinium, and others that he has met perhaps either in the Rocky
Mountains or in the Andes or in Equatorial Africa or in the Himalayas.
If fresh from Chile he will find on these heights the familiar Gunnera
and the Chilian Strawberry (Fragaria chilensis). If he has been in New
Zealand and in the islands of the Southern Ocean he will find old
friends in the genera Acæna and Coprosma. He may handle once again
plants like Nertera depressa, that he gathered on Tristan da Cunha; and
on the boggy summits of some of the mountains he will find the
ubiquitous Sun-dew (Drosera longifolia).

Within the limited area occupied by the peaks of Tahiti he will find
genera like Astelia and Coprosma that are at home in New Zealand or in
Antarctic America, and may even find, as in the cases of Coriaria
ruscifolia and Nertera depressa, the identical species that are at home
in those distant regions. Even on the summit of Rarotonga he will gather
a species of Vaccinium. In Fiji, here and there on some isolated
mountain-top he may come upon a remnant of this Antarctic flora, such as
a solitary species of Coprosma or Lagenophora, that will carry him back
for a moment to high southern latitudes; and in the highlands of Savaii,
in the neighbouring Samoan Group, he will find again Nertera depressa
and a species of Vaccinium. But that which will interest him most in
Fiji will be the tall conifers of the genera Dammara, Podocarpus, and
Dacrydium, which will bring to him memories perhaps of New Zealand and
southern Chile, of South Africa, and of the mountain-woods of Java and
of Southern Japan.

Yet the influence of isolation has been at work amongst the
mountain-plants of all these groups. The agencies that have dispersed
over the tropical Pacific plants from the cold latitudes of the southern
hemisphere, and those that have borne the seeds of Plantago, Sanicula,
and Vaccinium from mountain-top to mountain-top, even though it be to a
peak in mid-ocean, are to a great extent inactive now.

THE MOUNTAIN-FLORA OF HAWAII AS ILLUSTRATED BY THE NON-ENDEMIC GENERA.

Let us look in the first place at Hawaii, where the breaking off of
communication with the outside world is especially pronounced. Here, all
the species of two-thirds or more of the mountain-genera are confined to
that group. Only in a relatively small number of cases are the species
in touch with the regions outside. The mystery of disconnection that is
so evident in the instance of the peculiar or endemic mountain-genera of
the Compositæ and Lobeliaceæ and other orders is here again presented to
us, and once more in the upland regions 4,000 to 10,000 feet above the
sea. We will now endeavour to discover from an examination of the
present distribution of the isolated mountain-genera (those non-endemic
genera possessing only peculiar species) along what tracks they arrived
at the Hawaiian uplands, tracks, as indicated by the local distribution
of the species, that have been more or less abandoned since.

_The Mountain Genera with only Endemic Species._—By referring to the
Table on the following page it will be observed that nearly a third of
these mountain genera have now their principal homes in the high
latitudes of the southern hemisphere. They are components of what
Forster and Hooker have termed the “Antarctic” flora, a collection of
plants that range round the globe in high southern latitudes, namely,
over Fuegia, New Zealand, southern Australia, South Africa, and the
islands of the Southern Ocean, the “Antarctic” islands, as they have
been termed. These genera are Acæna, Gunnera, Coprosma, Lagenophora,
Astelia, Oreobolus, and Uncinia. (It is necessary to observe that I am
entirely indebted to the Introduction to the _Botany of the “Challenger”
Expedition_ for my information on the “Antarctic” flora.)

We are thus led to expect that some of the other mountain genera may
have been similarly derived from cool southern latitudes, even though
they may be scarcely included in the “Antarctic” flora. This is very
probably true of Myoporum and Exocarpus, two genera that are chiefly
centred in Australia. A species of Sophora (S. tetraptera) is now one of
the most widely dispersed of the plants of high southern latitudes, a
circumstance which at all events explains the capacity for transport
that the ancestor of the Hawaiian “Mamani” (S. chrysophylla) must have
originally possessed (see Chapter XV.). Kinship between the Hawaiian
species and southern forms has been found in the case of a few of the
widely ranging genera here represented. Thus Decaisne placed Plantago
princeps next to P. fernandeziana of Juan Fernandez; whilst according to
Hillebrand, Plantago pachyphylla resembles P. aucklandica from the
Auckland Islands. These resemblances are consistently associated with
the respective range in altitude of the Hawaiian plants, since Plantago
princeps occurs usually between 2,000 and 4,000 feet, and P. pachyphylla
between 6,000 and 8,000 feet, the species of greatest elevation being
related with the species of highest latitude. It is thus seen that these
endemic mountain genera with peculiar species have very evident
affinities with the plants of extra-tropical southern latitudes, and
especially with the “Antarctic” flora. This affinity will also be found,
as will subsequently be noticed, in the case of genera like Cyathodes
and Nertera, where there is still a specific connection with the outside
world.

THE MOUNTAIN-FLORA OF HAWAII, AS REPRESENTED BY THE NON-ENDEMIC
GENERA (Compiled from Hillebrand’s Flora).

+----------------+------------+---------------+--------------+------------+
| | | Distribution |Distribution | |
| | | outside | in Hawaii, | |
| | | Polynesia. | Fiji, and | |
| | | | Tahiti. | |
| | Usual |---------------|--------------+ |
| Genus |altitude of |Both Worlds. |Hawaii only. | |
| |station in | |Old World. | |Hawaii, Fiji| |
| | feet. | | |New World. | | |Hawaii, | Fruit. |
| | | | | |Antarctic| | | Tahiti | |
| | | | | | flora. | | | |All | |
| | | | | | |Austra-| | | | three | |
| | | | | | |lia and| | | | groups | |
| | | | | | |New | | | | | |
| | | | | | |Zealand| | | | | |
+----------------+------------+-+-+-+-+-------+-+-+-+--------+------------+
| | | | | | | | | | | | |
| WITH ALL SPECIES ENDEMIC. |
| | | | | | | | | | | | |
|Ranunculus (2) |6,000- 7,000|+| | | | |+| | | |Achene. |
|Viola (5) |2,000- 6,000|+| | | | |+| | | |Capsule. |
|Silene (4) |2,000- 9,000|+| | | | |+| | | |Capsule. |
|Geranium (6) |5,000-10,000|+| | | | |+| | | |Carpels. |
|Vicia (1) |7,000- 8,000|+| | | | |+| | | |Pod. |
|Sophora (1) |5,000-10,000|+| | | | | | | | + |Pod. |
|Rubus (3) |4,000- 7,000|+| | | | | |+| | |Berry. |
|Acæna (1) |5,000- 6,000| | | |+| |+| | | |Spinose |
| | | | | | | | | | | | achene. |
|Gunnera (1) |3,000- 6,000| | | |+| |+| | | |Drupe. |
|Sanicula (1) |6,000- 8,000| | |+| | | | | | + |Prickly |
| | | | | | | | | | | | carpel. |
|Coprosma (9) |3,000- 9,000| | | | | + | | | | + |Drupe. |
|Lagenophora (1) | 6,000- | | | |+| |+| | | |Viscid |
| | | | | | | | | | | | achene. |
|Artemisia (2) |4,000- 8,000|+| | | | |+| | | |Achene. |
|Lobelia (5) |2,000- 6,000|+| | | | |+| | | |Capsule. |
|Vaccinium (2) |3,000- 8,000|+| | | | | | | | + |Berry. |
| | | | | | | | | | | Samoa | |
|Myoporum (1) |Coast-10,000| | | | | + |+| | | |Drupe. |
|Plantago (2) |2,000- 8,000|+| | | | |+| | | |Capsule. |
|Exocarpus (2) |3,000- 6,000| | | | | + |+| | | |Fleshy nut. |
|Sisyrinchium (1)|4,000- 7,000| | |+| | |+| | | |Capsule. |
|Astelia (2) |2,000- 6,000| | |+| | | | | | + |Berry. |
|Oreobolus (1) | 6,000 | | |+| | |+| | | |Toothed |
| | | | | | | | | | | | nutlet. |
|Uncinia (1) |3,000- 5,000| | | |+| |+| | | |Awned |
| | | | | | | | | | | | nutlet. |
|Agrostis (3) |4,000- 6,000|+| | | | |+| | | |Awned grain.|
|Deschampsia (3) |3,000- 6,000|+| | | | |+| | | |Awned grain.|
|Trisetum (1) |3,000- 5,000|+| | | | |+| | | |Awned grain.|
|Poa (2) | |+| | | | |+| | | |Grain. |
| | | | | | | | | | | | |
| WITH ENDEMIC AND NON-ENDEMIC SPECIES. |
| | | | | | | | | | | | |
|Cyathodes (2) |2,000-10,000| | | | | + | | |+| |Drupe. |
|Lysimachia (6) |Coast- 6,000|+| | | | |+| | | |Capsule. |
|Chenopodium (2) |Up to 7,000|+| | | | |+| | | |Seed-like. |
|Santalum (3) |Coast-10,000| |+| | | | | | | + |Drupe. |
|Carex (5) |2,000- 7,000|+| | | | | |+| | |Nutlet. |
|Rhynchospora (4)|Up to 10,000|+| | | | | | | | + |Nutlet. |
|Panicum (14) |Coast- 6,000|+| | | | | | | | + |Grain. |
|Deyeuxia (3) |Up to 10,000|+| | | | |+| | | |Awned grain.|
| | | | | | | | | | | | |
| WITH NO ENDEMIC SPECIES. |
| | | | | | | | | | | | |
|Fragaria | | | | | | | | | | | |
| chilensis |4,000- 6,000| | |+| | |+| | | |Fleshy. |
|Drosera | | | | | | | | | | | |
| longifolia | 4,000 |+| | | | |+| | | |Capsule. |
|Nertera depressa|2,500- 5,000| | | |+| | | | | + |Drupe. |
| | | | | | | | | | | Samoa | |
|Luzula | | | | | | | | | | | |
| campestris |3,000-10,000|+| | | | | | |+| |Capsule. |
+----------------+------------+-+-+-+-+-------+-+-+-+--------+------------+

It is evident that in one or two cases the connection between the
representatives of the “Antarctic” genera on the Hawaiian uplands and
those of high southern latitudes has only been recently broken off. Thus
with reference to the Hawaiian species of the Cyperaceous genus,
Uncinia, it may be observed that although Hillebrand regards it as a
distinct species, Hemsley (_Intr. Bot. Chall. Exped._, p. 31) remarks
that it is very near if not the same as a New Zealand species, an
affinity very significant of the source of the mountain plants of this
group that are derived from these southern latitudes.

The next component to be recognised in these Hawaiian mountain genera
with peculiar species is a small special American element; and in this
connection Sanicula and Sisyrinchium may be especially mentioned. The
first is mainly North American, and particularly Californian; but there
are two solitary species found on the continents and in oceanic islands
such as the Azores. The continental species, Sanicula europæa, occurs
not only in Europe and Central Asia, but in South Africa, and at high
elevations on the mountains of Equatorial Africa and of Madagascar. It
is not, however, with this widely ranging species that Sanicula
sandwicensis is related, but with S. menziesii, a species from
California and Oregon (Hillebrand). Sisyrinchium is confined to
temperate and tropical America; but a singular and suggestive outlier of
the genus (S. bermudiana) is found in Bermuda.

The mountain genera that are distributed on both sides of the Pacific
constitute about three-fifths of the total. So far as my scanty data
show, they seem to have reached Hawaii from the four quarters of the
compass. The probable southerly origin of Plantago has been already
indicated. Hillebrand notes the great resemblance between Lobelia
gaudichaudii and an undescribed species from the Liukiu Islands, lying
on the west side of the Pacific. It is likely, also, that the genus
Ranunculus reached Hawaii from the west, since one of the species, R.
mauiensis, resembles R. repens of the Old World (Hillebrand); whilst the
other, R. hawaiiensis, comes near R. sericeus of Mauritius (Drake del
Castillo). On the other hand, the genus Rubus may hail from an American
source, since, in the opinion of Gray, Rubus hawaiiensis, one of the
mountain raspberries, finds its nearest relative in R. spectabilis from
the north-west coast of America; and there are reasons for believing, as
will subsequently be shown, that the genus Artemisia has an American
source. It is also probable that some of these genera have reached
Hawaii from the north, since it is likely, as pointed out in a later
page, that the Carices of the Hawaiian uplands came originally from
north-eastern Asia.

In the previous paragraphs the mountain genera have been considered with
especial reference to their distribution and source beyond the confines
of the Pacific. If we now briefly discuss them from the standpoint of
their distribution within the Pacific, or rather as concerning their
presence or absence in the Fijian and Tahitian groups, we shall see that
to a large extent Hawaii has received its mountain genera of this era
independently of the other Pacific groups.

Mountain genera possessing only peculiar species, in Hawaii only 20
Mountain genera possessing only peculiar species, in Hawaii and Fiji 2
Mountain genera possessing only peculiar species, in Hawaii and Tahiti 0
Mountain genera possessing only peculiar species, in all three groups 4
—
26

It is here shown that three-fourths of the genera of the Hawaiian
mountains in this era are not found either in Fiji or Tahiti. This, as
before pointed out, is mainly to be attributed to the greater elevation
of the Hawaiian Islands. Had there been an island 13,000 to 14,000 feet
in height in Fiji, we cannot think that any such contrast in the floras
would have existed. The temperate genera of the Hawaiian uplands would
have been largely represented in the Fijian flora. Yet although we do
not find such genera as Ranunculus, Geranium, Sanicula, Uncinia, &c., in
Fiji and Tahiti, a small number of the Hawaiian mountain genera have
obtained a scanty footing. This is what we might have expected. Thus,
Lagenophora has been found on the mountains of Vanua Levu, and Vaccinium
in Tahiti and Rarotonga; whilst Coprosma and Astelia occur on the tops
of some of the mountains in both regions. In Fiji their distribution
seems sporadic, as shown not in Lagenophora alone, but also by Astelia,
which has been found only on the summit of Kandavu.

_The Capacities for Dispersal of the Hawaiian Non-endemic Mountain
Genera possessing only Peculiar Species._—As shown in the Table, seven,
or 27 per cent., of these genera have fleshy fruits that would attract
frugivorous birds. In three cases (Gunnera, Coprosma, Myoporum) they are
drupes, in three others (Rubus, Vaccinium, Astelia) they are berries,
and in one (Exocarpus) there is a nut with a fleshy perigone. It is
particularly interesting to notice that frugivorous birds, and I include
here granivorous birds that are known to be frugivorous at times, could
have transported seeds of the “Antarctic” flora to this group. We can
observe the process in operation in our own time within the limits of
the group. It has been long known, and we find it referred to in the
pages of Hillebrand’s work, that the wild mountain-goose (Bernicla
sandwicensis) feeds upon the fruits of Coprosma ernodeoides, and of
Vaccinium reticulatum, the famous “ohelo.” The fruits of the first are
known to the natives as “kukai neenee” (droppings of geese), and the
hard stones or pyrenes are very well suited for withstanding the risks
of the digestive process. I found a number of these pyrenes in the
stomach of a mountain-goose shot by my companion, Dr. Krämer, high up
the slopes of Mauna Loa.

According to Mr. Perkins, Chloridops kona, a big Hawaiian finch, feeds
on the fruits of the bastard sandal-tree (Myoporum sandwicense). There
are no “impossible fruits” among the mountain genera of Hawaii, that is
to say, fruits so large that bird agency must be excluded. All of them
are practicable in point of size. Thus amongst the largest, the “stones”
of Gunnera would not exceed 1/5 of an inch (5 mm.), and those of
Myoporum scarcely 1/4 of an inch (6 mm.); whilst the nuts of Exocarpus
range in the Hawaiian species from 3/10 to 6/10 of an inch (7-15 mm.),
and the beans of Sophora chrysophylla do not at the most exceed 1/4 of
an inch (6 mm.).

The principal feature, however, which these mountain genera exhibit from
the point of view of their dispersal is the number of plants possessing
seeds or fruits capable of adhering to plumage. Half of these genera are
thus characterised. Of these Sanicula and Acæna represent the ordinary
hooked fruits; whilst the fruits of the Grasses and Sedges, Agrostis,
Deschampsia, Trisetum, Poa, Oreobolus, and Uncinia, are enabled by means
of their awns or of their serrated beaks to attach themselves to
plumage, and the same may be said of the carpels of Geranium. The fruits
of Lagenophora and the seeds of Plantago display the capacity of
adhesiveness by means of a gummy secretion.

One or two of these genera need further mention. I will first take
Acæna, which is spread all over the south temperate zone both on the
continents and on the islands. The Hawaiian species (A. exigua) forms
tussocky growths on the swampy summits of Mount Eeka in Maui, and in
Kauai, at an elevation of 6,000 feet above the sea. Numerous observers
refer to the probable mode of dispersal of the genus in the “Antarctic”
and neighbouring islands. Captain Carmichael, in the instance of Acæna
sanguisorbæ on Tristan da Cunha, observes that it overruns the low
ground. Its burr-like fruit, as he describes, “fixes itself on the
slightest touch into one’s clothes, and falling into a hundred pieces
covers one all over with an unseemly crust of prickly seeds not to be
got rid of without infinite labour” (_Trans. Linn. Soc._, xii. 483,
1818). Both Mr. Moseley (Wallace’s _Island Life_, p. 250) and Dr. Kidder
(_Bull. U.S. Nat. Mus._, 2) refer to the burrowing habits of the
Petrels, Puffins, and other sea-birds amongst the vegetation covering
the ground in Tristan da Cunha, Marion Island, Kerguelen, &c., in places
where Acæna, amongst other plants, thrives. Mr. Moseley remarks that the
fruits of this genus stick like burrs to feathers, and he looks to
sea-birds for the dispersal of this and similar plants over the ocean.
He especially notes that the Petrels and other seafowl burrow and breed
high up the mountain-slopes of tropical islands as in Tahiti, Viti Levu,
Hawaii, and Jamaica.... It should be noted in the case of the Hawaiian
endemic species that it has been found only on two mountain tops; and
that however active may be the dispersal of the genus in south temperate
latitudes now, the Hawaiian Islands lie outside the present area of
dispersal.

The next mountain genus I will specially refer to is Lagenophora, one of
the Compositæ. The solitary Hawaiian endemic species, L. mauiensis, is
restricted to the summit of Mount Eeka, in Maui. In the mountains of
Vanua Levu, Fiji, another peculiar species, L. pickeringii, has been
found; and there is a species, L. petiolata, in the Kermadec Islands
(Hooker, in _Journ. Linn. Soc._, i. 127); but the genus is chiefly
characteristic of Australia, New Zealand, and temperate South America,
one species occurring both in Fuegia and Tristan da Cunha. The genus has
no pappus; but Hooker in the case of the Kermadec species considered
that the “viscid fruit” favoured its dispersal; and this may probably be
true of the genus.

With regard to the capacity for dispersal of the seeds of Plantago, it
may be pointed out that the seeds of Plantago major, P. lanceolata, &c.,
become coated with a mucilaginous material when wetted. In 1892, when
experimenting on these plants, I found that the wetted seeds adhered
firmly to a feather, so that it could be blown about without their
becoming detached. Species of Plantago are so characteristic of the
“alpine” floras of the summits of lofty mountains in the tropics, as in
Java and many other regions, that the mode of dispersal has always been
a subject of curiosity. I cannot myself doubt that this is the
explanation of the occurrence of the representatives of the genus that
now thrive as endemic species on the higher slopes of the Hawaiian
mountains. This method of dispersal for Plantago is recognised by recent
writers on the subject of seed-dispersal. (In a paper in _Science
Gossip_ for September, 1894, I dealt with the “mucous adhesiveness” of
such seeds as a factor in dispersal. The subject had previously been
discussed by Kerner in one of the earlier volumes of his
_Pflanzenleben_; and I have summed up some of the results in Note 43 of
the present volume.) My readers can readily ascertain by a simple
experiment that a bird pecking the fruit-spikes in wet weather would
often carry away some of the sticky seeds in its plumage. Several years
ago, when I was endeavouring to examine the condition of these seeds in
the droppings of a canary, my efforts were defeated by the bird itself,
since, in spite of all my care, some seeds and capsules were always
carried by the bird on its feathers into the clean cage reserved for the
experiment.

The plants of these mountain genera possessing dry seeds or fruits
neither very large nor very minute and suitable for bird-food are
Ranunculus, Viola, Vicia, Sophora, Artemisia, Sisyrinchium, six in all,
or 24 per cent. of the total. On the probable method of transport of the
ancestors of these endemic species the following remarks may be made.
With regard to Ranunculus, some authors like C. M. Weed
(_Seed-Travellers_, p. 48, Boston, 1899) perceive in the curved or
hooked beaks of the achenes a means of attaching the fruit to plumage.
This no doubt applies to some species, and it is advocated by Ekstam for
some of the plants of the Nova Zembla flora. There are others to which
this explanation would not be applicable, and the achenes of the
Hawaiian species do not appear to be specially fitted for this mode of
transport. I have found the achenes of Ranunculus frequently in the
stomachs of birds in England, in partridges frequently, and in wild
ducks at times. Those of certain species that possess buoyancy are
common in the floating seed-drift of rivers, as of the Thames (_Journ.
Linn. Soc. Bot._, xxix. 333), and they would probably in this way be
often swallowed by waterfowl.

I have but few data directly relating to the dispersal of seeds of Viola
by birds. From the frequent occurrence of species in alpine floras, as
in the Caucasus, the Great Atlas, in the mountains of Equatorial Africa,
in Madagascar, &c., it may be inferred that birds transport the seeds
between the higher levels of many continental ranges in tropical regions
and to the mountain-slopes of neighbouring large islands. Viola
abyssinica, for instance, which occurs in Madagascar, is spread over the
elevated mountain ranges of tropical Africa. With regard to the five
Hawaiian species, it may be remarked that three of them are bog species
and two occur in dry situations. The first are most characteristic of
the mountains, one species occurring on the summit of Mount Eeka, 6,000
feet above the sea. Judging from the stations alone, at least two
species were originally introduced into the Hawaiian Group.

Viola seeds, as indicated by my experiments on the different British
species, including Viola palustris, are not buoyant, and there is no
possibility of the seeds being picked up by birds in floating drift.
There is, however, a possible means of dispersal in birds’ plumage by
means of the mucosity of the seeds of some species. Thus, although this
is not exhibited, as shown by my experiments, by Viola canina and V.
palustris, it is well displayed by the Field-Pansy (V. tricolor). I
found that the seeds of this species, after lying a little time in
water, were thickly covered with mucus, and that they adhered to a
feather, on drying, as firmly as if gummed. This did not, however, come
under my notice in the case of the seeds of one of the Hawaiian species,
V. chamissoniana, examined by me. One sometimes observes Viola canina in
England growing in places, as in the crevices and on the tops of old
walls, where its seeds could have only been carried by birds. In some
cases the propellent force of the seed ejected by the contracting valves
of the capsule would explain queer stations. In its power of
seed-expulsion, Viola chamissoniana, the common Hawaiian species, is
just as active as our British species.

With regard to the Leguminous genus Vicia we have the observation of
Focke on the dispersal of its seeds by pigeons, as described before on
page 150.

Sophora chrysophylla, the “Mamani” of the natives and one of the most
familiar of the trees of the Hawaiian mountains, is discussed at length
in Chapter XV., where the difficulty of supposing that its seeds could
be transported unharmed in a bird’s stomach half-way across the Pacific
is pointed out; and it is suggested that it was more probably derived
from a littoral species brought by the currents. However, the point is a
debatable one, and the seeds of the “Mamani” can scarcely be regarded as
“impossible” from the standpoint of dispersal.

With reference to the possibilities of dispersal of the achenes of
Artemisia, some very suggestive indications are to be obtained from a
paper by Mr. D. Douglas on the North American Tetraonidæ published in
the _Transactions of the Linnæan Society_ for 1833. The “Cock of the
Plains” (Tetrao urophasianus), as we here learn, makes its nest on the
ground under the shade of Artemisia bushes, and lives on the foliage and
fruits of these and other plants. This bird is plentiful in Columbia and
North California, and another allied species is mentioned which lives on
the same sort of food. Later authors refer to these and other birds of
the same family as living chiefly on the Sage-brush (Artemisia
tridentata), a plant prevailing over great regions of the plains as well
as on the slopes of the Sierra Nevada and of the Rocky Mountains.
According to Dr. Sernander (page 228), birds when feeding on the fruits
of Artemisia vulgaris in the district of Upsala scatter them about and
thus aid in its dispersal. Artemisia achenes, since they have neither
pappus nor other appendages, nor any special adhesiveness when wetted,
depend largely on their small size and light weight to aid them in
dispersal. (Those of A. absinthium measure a millimetre in length, or
1/25 of an inch, whilst those of A. vulgaris measure 1·8 mm., or 1/14 of
an inch.) Driven as we are to look to bird-dispersal for the means of
transport of Artemisia achenes, it is interesting to find a possible
source of the Hawaiian endemic species on the nearest American mainland,
even though it is some 2,000 miles away. It is assumed that they would
be ordinarily carried in adherent soil or entangled in the feathers, and
on rare occasions in the bird’s stomach.

The small seeds of Sisyrinchium possess no means of adherence to
plumage. They are crustaceous, and in cases where the stomach and
intestines of a bird are well filled with other food they are quite
capable of resisting injury. The solitary Hawaiian species has,
according to Hillebrand, a range in altitude from 3,500 to 7,000 feet. I
found this pretty herb most abundant on the “cattle-plains” of Hawaii
between 5,000 and 6,000 feet, where it is evidently in part dispersed by
the cattle and other animals. The seeds are very small, being about a
millimetre in size, and when dried nearly 100 go to a grain (0·65
decigramme). They might thus also be transported in mud on birds’ feet.

For the mode of dispersal of the minute seeds of Lobelia, the last of
the mountain genera to be specially noticed, I must refer the reader to
the remarks on this subject in Chapter XXII. They would probably be
carried in soil adhering to the legs or feet of a bird.

There are one or two interesting points relating to the temperate genus
Silene, which is represented on these mountains. The four Hawaiian
species show a great range in altitude. Thus, whilst S. struthioloides
finds its home in Hawaii and Maui at elevations of 5,000 to 9,000 feet,
another species (S. lanceolata) thrives equally at elevations of 5,000
or 6,000 feet on the central plateau of Hawaii and at heights only of
300 to 500 feet above the sea. Although I have not yet come upon any
direct reference to the mode of dispersal of the small seeds of this
genus, there is little doubt that their rough tuberculated surfaces
would favour their attachment to plumage. A very significant
observation, however, is made by Jens Holmboe in a paper on littoral
plants in the interior of Norway. He refers to the occurrence in no
small quantity of Silene maritima on the top of “Linnekleppen,” 331
metres high, one of the highest peaks of Smaalenene, and distant about
29 kilometres from the nearest coast (_Strandplanter i det indre af
Norge_, “Naturen,” Bergen, 1899). Sernander (p. 405), commenting on this
observation, remarks that since bare hill-tops are frequented by birds,
such an agency in this instance is not impossible.

I will conclude these remarks on the non-endemic Hawaiian mountain
genera possessing only peculiar species, with a few observations on the
genus Vaccinium in the Pacific. This genus is known to be distributed
over the northern hemisphere and to occur on the uplands of tropical
mountains, as, for instance, on the summits of the Java mountains and on
the high levels of the Equatorial Andes at altitudes even of 15,000 to
16,000 feet. There are apparently only some four or five species known
from the Pacific islands, from Hawaii, the Marquesas, Tahiti, Rarotonga,
Samoa, and the New Hebrides, and it would almost seem that these can be
reduced to one or two species. Although not yet recorded from Fiji, the
probability of the genus being represented on some of the mountains is
pointed out by Seemann. Of these Pacific forms a single species, V.
cereum, is spread over the East Polynesian region including the
Marquesas, Tahiti, and Rarotonga; and, according to Hillebrand, V.
reticulatum, one of the two endemic Hawaiian species, is nearly related
to it. Even the New Hebrides species (V. macgillivrayi) resembles it,
according to Seemann, in general appearance. That there has been a
single Pacific polymorphous species is, as shown below, not impossible;
but Reinecke, in describing in 1898 the Samoan species, V. antipodum,
was under the impression that it was the only species known from the
southern hemisphere, and says nothing of its affinity to other Pacific
plants.

A few words on the station and habit of Vaccinium in the Pacific islands
may be here of interest. In Hawaii there are, according to Hillebrand,
two species, a high-level form, V. reticulatum, occurring at elevations
of 4,000 to 8,000 feet, and a low-level form, V. penduliflorum, ranging
between 1,000 and 4,000 feet. I may, however, remark that the last
species occasionally came under my notice at elevations of 6,000 to
7,000 feet. This species exhibits much variation, and Gray, Wawra, and
other botanists have evidently not been always able to distinguish
between the two species in their varying forms. It is not only
distinguished from the high-level species by its lower station, but also
by its epiphytic habit, a circumstance that, as pointed out below, may
explain some of the differences, since such a habit is bound up with the
difference in station. It seems, therefore, safer to regard them as
station forms of one species which is closely allied to V. cereum, the
species of the South Pacific, an inference which, if well founded, would
make highly probable the view that there has been a single polymorphous
Pacific species.... In Tahiti, as we learn from Nadeaud, V. cereum
occurs on the mountain-tops at altitudes exceeding 800 metres (2,600
feet). In Rarotonga, according to Cheeseman, it is found on the summits
of most of the higher hills extending almost to the summit of the
island, 2,250 feet above the sea. The Samoan species, V. antipodum of
Reinecke, which that botanist considers as probably one with V. whitmei,
a Polynesian (Samoa?) species originally described by Baron F. von
Müller, grows in the central mountains of Savaii at an elevation of
1,500 metres (4,920 feet).

These Pacific species of Vaccinium, as on tropical mountains of the
continents, occasionally assume an epiphytic habit, and it is here, as
above observed, that lies one of the distinctions between the Hawaiian
species. V. penduliflorum, the low-level form, occurs typically in the
forests, where, according to Hillebrand, it grows on the trunks of old
trees. The trees, however, may be quite in their prime, and I have
observed it growing in the fork of the trunk of an Olapa tree
(Cheirodendron gaudichaudii). It is in this connection of significance
to notice that a variety found in open glades and on grassy slopes is
described by Hillebrand as terrestrial in habit. The other high-level
form, V. reticulatum, grows gregariously on open ground, and is
typically terrestrial in its habit. The Samoan species, as we learn from
Reinecke, grows on trees, as on the branches of Gardenia. The epiphytic
habit of species of Vaccinium is especially discussed by Schimper in the
case of plants growing on the Java mountains. He there shows
(_Plant-Geography_, i. 14) that species which are epiphytes in the
virgin forest become terrestrial plants in the treeless alpine region.
This interchange of station, which is exhibited by several other plants,
including orchids and ferns, is connected with their xerophilous
characteristics, and is given by Schimper as an example of the
interchange of physiologically dry habitats.

Of the mode of dispersal of Vaccinium by frugivorous birds, much has
been written and much will be familiar to my readers. The berries of V.
reticulatum are known to be the principal food of the Hawaiian
mountain-goose. But probably birds of the grouse family have been the
chief agents in distributing the genus over the continents. I have
frequently found the fruits in the stomachs of the Black Cock (Tetrao
tetrix), the Scotch Grouse (Lagopus scoticus), and the Capercailzie
(Tetrao urogallus); but the same story comes from all over the northern
hemisphere. The Willow Grouse (Lagopus albus), which travels round the
globe, is known to feed on them. Hesselman in Sweden and Ekstam in Nova
Zembla have especially investigated the dispersal of Vaccinium by Tetrao
tetrix and Lagopus (see Sernander, pp. 6, 226); and according to Mr.
Douglas and others the different species of Tetrao that frequent the
subalpine regions of the Rocky Mountains and the uplands of Columbia and
North California subsist on Vaccinium fruits. This family is not now
represented in the Hawaiian avifauna; but it is noteworthy, as indicated
by the differentiation of the Pacific species of Vaccinium, that
dispersal of the genus is there almost suspended except within the
region of Eastern Polynesia. It is probable that numerous other birds,
except the Hawaiian goose, aided the original dispersal.

_The Mountain Genera with both Endemic and Non-endemic Species._—I pass
on now to consider those Hawaiian mountain genera that possess species
some of which are confined to the group, whilst others occur in regions
outside the islands. They are not many, as may be seen from the table
before given, and but few of them are entirely restricted to the high
levels, a range in altitude that may be frequently associated with great
lateral extension of the genus over different latitudes. Here the agents
of dispersal have through some species in each genus preserved a
connection with the outer world, though it may be restricted to the
limits of the Pacific islands.

Cyathodes tameiameiæ, an Epacridaceous species found also in the uplands
of Tahiti, occurs, according to Hillebrand, on all the Hawaiian Islands,
from 1,800 feet up to the limit of vegetation 10,000 feet and over above
the sea. I found it, however, at even lower levels. On the Puna coast of
Hawaii, associated with Metrosideros polymorpha, Osteomeles
anthyllidifolia, and other inland plants, it descends on the surface of
ancient lava-flows to the coast wherever the bolder spurs reach the
sea-border. The other species, C. imbricata, is more exclusively
confined to the greater altitudes. It is endemic, and may possibly be a
station form of the other species.

The six species of Lysimachia are found at different elevations, one
near the sea-shore, others at altitudes of 2,000 to 3,000 feet, and
others again at elevations of 6,000 feet. Chenopodium sandwicheum occurs
at all elevations from near the coast to the high inland plains of
Hawaii and to the upper slopes of Mauna Kea, that is to say, up to
altitudes of 6,000 or 7,000 feet. Hillebrand observes that it is a low
decumbent plant at the coast, and may become arborescent with a height
of 12 to 15 feet in the upper forests of Mauna Kea.

The species of Santalum (sandal-wood trees) also display great vertical
range in these islands. Though S. freycinetianum, which is also a
Tahitian species, is most at home in the forests 2,000 to 4,000 feet
above the sea, it has, as Hillebrand informs us, a dwarfed form that
extends far up the mountain slopes of Mauna Loa and Hualalai to
elevations of 7,000 or 8,000 feet, and another dwarfed shrubby variety
that grows only near the sea-shore. Another species, S. haleakalæ,
occurs as a tall shrub on Haleakala at elevations of 8,000 to 10,000
feet. Among the sedges, most of those of the genera Carex and
Rhynchospora are found at altitudes of between 3,000 and 7,000 feet, and
two grasses of the genus Deyeuxia occur at elevations of 6,000 to 8,000
feet.

Amongst these Hawaiian mountain genera with both endemic and non-endemic
species there are no plants possessing fruits which from their size
could be with difficulty regarded as dispersed by birds. The mode of
dispersal of these plants is in some cases indeed not far to seek. Thus
in the stomach of an Hawaiian goose (Bernicla sandwicensis), shot by my
companion Dr. Krämer on the slopes of Mauna Loa, I found a number of the
“stones” of Cyathodes tameiameiæ, the plant being abundant in fruit in
the immediate vicinity. It is highly probable that the seeds of Santalum
have been carried over the Pacific by frugivorous birds. We learn from
Dr. Brandis that Santalum album in India is mainly spread through the
agency of birds (_Bot. Chall. Exped._, iii. 13). The drupes of the
Pacific species, S. freycinetianum, that occurs alike in Hawaii, the
Marquesas, and Tahiti (Drake del Castillo), measure about half an inch.
There can be little doubt that with this tree, as with the species of
Cyathodes above mentioned, which also links together Tahiti and Hawaii,
there has been up to recent times an interchange by means of frugivorous
birds between these two regions, some 2,000 miles apart.

The small seeds of the capsular fruits of Lysimachia could be
transported in birds’ plumage or in dried soil attached to their feet or
feathers. The seed-like fruits of Chenopodium were probably dispersed by
some granivorous bird, much as nowadays our partridges carry about in
their stomachs the similar fruits of Atriplex. The long-awned fruits of
Deyeuxia were, it is likely, transported in birds’ plumage, and
doubtless also those of Panicum; whilst the nutlets of Carex and
Rhynchospora might have been carried about in a similar fashion.

The distribution of the non-endemic species of these Hawaiian mountain
genera may perhaps aid us in determining the original source of the
genus as well as in confirming the conclusions formed concerning the
other mountain genera that only possess species restricted to the group.
Lysimachia, Chenopodium, Carex, Rhynchospora, Deyeuxia, and Panicum are
found in both the Old and New Worlds. Since Hillebrand remarks that one
of the six species of Lysimachia (L. spathulata) occurs in Japan and in
the Liukiu, Bonin, and Marianne groups, we have here a valuable
indication of the route followed by a genus that has not been recorded
from the oceanic groups of the South Pacific.

The capricious distribution of the genus Carex in the Pacific is
remarkable, and it is noticed by Hemsley in the Introduction to the
_Botany of the “Challenger” Expedition_. No species have been recorded
from Tahiti, the Marquesas, and Rarotonga, but three Fijian species are
mentioned by Hemsley, and there is another in Samoa. Of the five
Hawaiian species given by Hillebrand, two are endemic. Of the rest, C.
wahuensis (oahuensis), Meyer, occurs also in Korea and Japan, whilst C.
brunnea, Thunb., is found in Japan and Australia, and the third, C.
propinqua, Nees., occurs all round the border of the Pacific Ocean, from
Kamschatka through Alaska south to the Straits of Magellan. These three
species all possess a home in common in north-east Asia, and probably
there lies the source of the Hawaiian species of Carex—a conclusion
which would help to explain the irregular distribution of the genus
amongst the South Pacific groups.

The genus Rhynchospora occurs alike in the Hawaiian, Tahitian, and
Fijian islands; but the groups in the North and South Pacific seem to
have been independently supplied with the original species, since R.
aurea, a widely spread tropical species, ranging the South Pacific from
New Caledonia to Tahiti, has not been recorded from Hawaii. A connection
between Hawaii and the Australian region seems to be indicated by a
species of Deyeuxia (D. forsteri) that is found also in Easter Island,
Australia, and New Zealand, and by the presence of the Australian and
New Zealand genus Cyathodes in Hawaii, though the existence of a species
common to both Tahiti and Hawaii goes to show that the route followed by
the genus lay through Eastern Polynesia. It is also not unlikely that
the genus Santalum reached Hawaii through Eastern Polynesia, since two
forms found in Hawaii and Tahiti are closely allied, and are, in fact,
regarded by Drake del Castillo as the same species. The genus occurs in
tropical Asia, Australia, and New Zealand.

Looking at the indications above given, I should be inclined to think
that the genera Lysimachia and Carex reached the Hawaiian mountains from
temperate Asia or the islands off its Pacific coast, and that Cyathodes,
Santalum, and Deyeuxia hail from the Australian or New Zealand region by
way of Eastern Polynesia.

_The Mountain Genera possessing no Endemic Species._—The few remaining
mountain plants of Hawaii to be considered are solitary, widely ranging
species of genera that here possess no peculiar species. Such may be
regarded as belonging to the latest age of the indigenous plants. They
still keep up, or kept up until recently, the connection with the world
outside Hawaii, and among them one may name here Fragaria chilensis,
Drosera longifolia, Nertera depressa, and Luzula campestris.

Fragaria chilensis, the Chilian strawberry, flourishes at elevations of
between 4,000 and 6,000 feet on the Hawaiian mountains. Its fruits,
according to Hillebrand and other authors, are much appreciated by the
wild goose of the islands. This plant ranges in America from Chile north
to Alaska; and Drake del Castillo is doubtless on safe ground when he
assumes that a congener of this bird originally brought the species from
the nearest part of the American continent, namely from California
(_Remarques_, &c., p. 8). In this connection it should be remembered
that one of the endemic mountain-raspberries of Hawaii (Rubus
hawaiiensis) finds its nearest relative, according to Gray, in Rubus
spectabilis, a species from the north-west coast of America.

The species of Sun-dew, Drosera longifolia, hitherto found only on the
marshy tableland of Kauai at an elevation of 4,000 feet above the sea,
occurs both in Asia and North America. Its minute fusiform seeds are
very light in weight, and might readily become entangled in a bird’s
plumage, or they could be carried in adherent dried mud.

Luzula campestris, which grows on the high mountains of the Hawaiian
group from 3,000 feet upward, is also found in Tahiti. It is widely
distributed in cool latitudes, and there is no special indication of its
source. Its seeds are especially well suited for adhering to birds’
feathers. When experimenting on these seeds in 1893 I ascertained that
whether freshly gathered or kept for more than a year they became on
wetting coated with mucus, and adhered firmly to a feather on drying.
There are many ways in which the “sticky” seeds in wet weather might
fasten themselves to a bird’s plumage. The plant-materials might be
used, for instance, for making nests. The Sea Eagle (Aquila albicilla),
as we learn from Mr. Napier (_Lakes and Rivers_), uses materials derived
from Luzula sylvatica in the construction of its nest.

Nertera depressa, a creeping Rubiaceous plant, with red, fleshy drupes
containing two coriaceous pyrenes, is found in all the Hawaiian Islands
at elevations of 2,500 to 5,000 feet, and it grows on the mountains of
Tahiti at altitudes over 3,000 feet. The genus is widely diffused over
the southern hemisphere. This particular species is characteristic of
the Antarctic flora, being found all round the south temperate zone
(excepting South Africa) in New Zealand, Fuegia, the Falkland Islands,
and Tristan da Cunha, and extending up the Andes to Mexico, occurring
also on the summits of Malayan mountains at elevations of 9,000 to
10,500 feet above the sea, as on Pangerango in West Java (Schimper), and
on Kinabalu in North Borneo (Stapf). Captain Carmichael, who resided on
Tristan da Cunha in the early part of last century, states (_Trans.
Linn. Soc._, xii. 483) that its drupes are eaten by a species of thrush
and by a bunting. Professor Moseley, who visited the island in the
_Challenger_ many years after, remarks that its fruits are “the
favourite food of the remarkable endemic thrush, Nesocichla eremita,”
the bunting being Emberiza brasiliensis (_Bot. Chall. Exped._, ii. 141).
It would seem most likely that the Hawaiian Islands received this
representative of the Antarctic flora through the Tahitian Islands, as
in the case of the species of Cyathodes common to both these groups.

Looking at the indications of these four widely ranging plants, the
Chilian strawberry (Fragaria chilensis), the Sun-dew (Drosera
longifolia), Nertera depressa, and Luzula campestris, it may be inferred
that with the exception of Nertera they all reached Hawaii from either
the Asiatic or American sides of the North Pacific, the last route being
evident in the case of the strawberry. Nertera depressa was probably
derived from southern latitudes.

_Summary._

(1) The second era of the flowering plants of the Pacific islands is
indicated by the non-endemic genera. Here also the isolating influences
have been generally active, and the work of dispersal is in some regions
largely suspended. Thus in Hawaii nearly half the non-endemic genera
possess only species that are restricted to the group, whilst in Fiji
and Tahiti about a fourth are thus isolated.

(2) The contrast in the elevations of the islands of the Hawaiian,
Tahitian, and Fijian regions is reflected in the development of an
extensive mountain-flora in Hawaii, in its scanty development in Tahiti,
and, excluding the Fijian conifers, in a mere remnant in Fiji and Samoa.

(3) The influence of isolation has been very active in the Hawaiian
mountains, since about two-thirds of the genera contain only species
confined to the group, and are thus disconnected from the world outside.

(4) Amongst these disconnected Hawaiian mountain genera, Antarctic or
New Zealand genera, like Acæna, Gunnera, Coprosma, and Lagenophora,
constitute nearly a third. The American element, represented, for
instance, by Sanicula and Sisyrinchium, is small; whilst the genera
found on both sides of the Pacific form more than one-half of the total,
and include genera like Ranunculus, Viola, Rubus, Artemisia, Vaccinium,
and Plantago, that often represent the flora of the temperate zone on
the summits of tropical mountains. Three-fourths of these genera are not
found either in Fiji or in Tahiti.

(5) The proportion of the disconnected Hawaiian mountain genera
possessing seeds or seedvessels suited for dispersal in a bird’s plumage
is very large, quite half belonging to this category; whilst only about
a fourth have fruits that would be dispersed by frugivorous birds.

(6) The Hawaiian mountain genera that still remain in touch with the
external world through species found outside the islands whilst other
species are confined to the group, present a later stage in the
plant-stocking. Their widely ranging species, which would be dispersed
either by frugivorous birds, as with Santalum and Cyathodes, or in
birds’ plumage, as with Lysimachia, Carex, and Deyeuxia, seem to
indicate that the main lines of migration for these genera have been
from temperate Asia and from the Australian and New Zealand region, the
last by way of Eastern Polynesia.

(7) The latest stage of the Hawaiian mountain-flora is exemplified by
those genera that are only represented in the group by a solitary
widely-ranging species, such as Fragaria chilensis, Nertera depressa,
Drosera longifolia, and Luzula campestris. It is our own age; and birds
are shown to be actual agents in the dispersal of the two first-named
species and to be probable agents with the two other species. The two
last-named species probably reached Hawaii from one or other side of the
North Pacific; whilst Fragaria chilensis doubtless hails from the
adjacent part of the American continent, and Nertera depressa from high
southern latitudes by way of Tahiti.

CHAPTER XXIV

THE MOUNTAIN-FLORAS OF THE TAHITIAN AND FIJIAN REGIONS

THE MOUNTAIN-FLORA OF THE TAHITIAN REGION AS ILLUSTRATED BY THE
NON-ENDEMIC GENERA

THIS floral region of the Pacific corresponds with the limits of Eastern
Polynesia, and includes not only the Tahitian group proper, but also the
Cook, Austral, Paumotuan, and Marquesan groups. It is only, however, in
Tahiti, the peaks of which rise to over 7,000 feet above the sea, that
we should expect to find such a mountain-flora, since the islands of the
other groups are much lower, the highest of them in the Marquesan group
barely exceeding 4,000 feet. Yet even in Tahiti it is not possible to
speak of a mountain-flora in the sense that we attach to it in Hawaii.
The elevated area of its interior is, as described in Chapter XIX.,
relatively very small; whilst, as Drake del Castillo points out, the
conditions presented by the steep mountain-slopes rarely afford a hold
for trees of any size, ferns often predominating in the higher levels.
Still, we can observe the traces of such a flora, and it is in this
sense only that the term “mountain-genera” is used in relation with this
group.

_Mountain-Genera of the Tahitian or East Polynesian Region._

Weinmannia, Saxifragaceæ, from New Zealand. }
Coprosma, Rubiaceæ, from New Zealand. }all species
Vaccinium, Vacciniaceæ, from the northern hemisphere. }endemic.
Astelia, Liliaceæ, from New Zealand. }

Coriaria, Coriariaceæ, from New Zealand }some
Cyathodes, Epacridaceæ, from New Zealand }species
}endemic

Nertera depressa, Rubiaceæ, a species of the Antarctic flora.
Luzula campestris, Juncaceæ, from the northern hemisphere.

The Tahitian non-endemic mountain-genera, though scanty in number, are
of considerable interest to the student of plant-dispersal. Among those
possessing only species that are confined to Eastern Polynesia, genera
that would be regarded as belonging to a past era of dispersal,
Weinmannia, Coprosma, Vaccinium, and Astelia may be mentioned.

Weinmannia, a Saxifragaceous genus of trees and shrubs, not represented
in Hawaii, but recorded from almost all the elevated oceanic groups of
the tropical South Pacific, as well as from the New Hebrides and New
Caledonia, has its home in South America, more particularly in the
Andes, and also occurs in New Zealand, Tasmania, and the Mascarene
Islands. One can scarcely doubt that, as in the case of Coprosma, the
Pacific Islands derived their species originally from high southern
latitudes, as from New Zealand, the absence of the genus from Hawaii
negativing an American origin. Of the two Tahitian species, one is
peculiar to Tahiti, whilst the other, W. parviflora, which is
conspicuous on the mountain-crests at elevations of 3,000 feet and over,
occurs also in the Marquesas. Another species grows in abundance in the
interior of Rarotonga. Samoa possesses two species, one of which, W.
affinis, occurs in Fiji, and the other, W. samoensis, which frequents
the mountains at elevations of 1,500 to 3,300 feet, is seemingly
endemic. Fiji possesses four or five species of Weinmannia occurring at
all altitudes up to 2,000 feet, of which some are evidently peculiar.
The capsular fruits of this genus contain hairy seeds that would
probably become entangled in a bird’s plumage. Dispersal by birds is
distinctedly indicated in the curious observation of Dr. Reinecke in the
case of the Samoan peculiar species. The seeds, he says, appear to
germinate by preference on the bark of other trees, young plants growing
epiphytically being of frequent occurrence.

There is some evidence that the species of Weinmannia, about ten in all,
found in the tropical islands of the open Pacific are derived from one
or two polymorphous species. As we learn from Mr. Cheeseman, the
Rarotongan species, W. rarotongensis, has considerable affinity to
several closely allied Polynesian species, and its nearest allies are a
Fijian and Samoan species, W. vitiensis and W. samoensis. Possibly, he
remarks, fuller materials may lead to the union of several of these
forms under one species.

The interesting New Zealand genus Coprosma, which we have noticed in
Hawaii, occurs also in the Tahitian region and Fiji; and it will be
further discussed under the last-named locality. The genus Vaccinium has
been previously dealt with in Chapter XXIII.

The Liliaceous genus Astelia may be considered as representing, like
Coprosma, the Antarctic or New Zealand flora in the higher levels
(usually) of the islands of the tropical Pacific, where it grows both on
trees and on the ground. The genus, according to Hemsley, is chiefly at
home in New Zealand, but is also found in Fuegia and in South-east
Australia. It is represented in Hawaii, Tahiti, Samoa, and Fiji. In
Hawaii there are two peculiar species ranging between 2,000 to 6,000
feet in elevation. The solitary Tahitian species, A. nadeaudi, is found
in the central mountains of Tahiti, reaching to the crests of Mount
Aorai, which attains a height of 6,700 feet. Fiji and Samoa possess a
species in common, A. montana, which is only recorded by Seemann, from
the summit of Kandavu, 2,750 feet above the sea; whilst in Samoa it
frequents, according to Reinecke, moist coast districts. The fruits of
Astelia are berries with crustaceous seeds that would be dispersed by
frugivorous birds.

Amongst the Tahitian mountain-genera that possess species ranging far
beyond this region as well as species confined to the group may be
mentioned Coriaria and Cyathodes. It is to their non-endemic species
that we look for further clues as to the general lines of migration by
which the mountain-genera that only possess peculiar species reached
this group. The evidence afforded by Coriaria is of some importance. The
genus has not been recorded from Hawaii, and, so far as the collections
of Seemann and Home show, not from Fiji. It is found in the
Mediterranean region, the Himalayas, Japan, New Zealand, and Antarctic
America, including Chile; and there are two particular species, C.
ruscifolia and C. thymifolia, that occur in both cases in New Zealand
and the adjacent islands and in South America (Introd. _Chall. Bot._ p.
53). The first of these, which is very common in Chile, exists also in
Tahiti on the crest of Aorai, 6,700 feet above the sea. Drake del
Castillo also describes a peculiar Tahitian species, C. vescoi, of which
the altitude is not given. Here one is in doubt whether Tahiti derived
its wide-ranging species from New Zealand or from Chile; but in the New
Zealand home of Coprosma, another Tahitian mountain-genus, we are
afforded the clue. The fruits of Coriaria possess fleshy cocci that
attract birds, though it would seem that the seeds of plants of this
genus are poisonous for man. Among the numerous fruits that form the
diet of the New Zealand fruit-pigeon (Carpophaga novæ zealandiæ) are
included, as we learn from Sir W. Buller in his _Birds of New Zealand_,
those of the “tupakihi” or “tutu” shrub, which Kirk identifies with C.
ruscifolia, the species that also occurs on the summit of Tahiti.

The Australian and New Zealand genus Cyathodes (Epacridaceæ) has been
already noticed in the case of Hawaii (page 282). The two Tahitian
species occur on the elevated mountain-ridges forming the summits of
Tahiti, one of them, C. tameiameiæ, occurring also in Hawaii, and the
other, C. pomaræ, being restricted to the group. I have shown that the
fruits are dispersed by frugivorous birds, and I can only include the
genus as another example of the representation of the New Zealand flora
in Tahiti.... There remain of these so-called Tahitian mountain-genera
the Antarctic Nertera and the north-temperate Luzula, each represented
by a solitary widely ranging species, N. depressa and L. campestris,
which I have fully discussed under Hawaii (Chapter XXIII), in which
group they also occur.

When we look at the evidence of origin supplied by the four Tahitian
mountain-genera possessing species that are found outside the group,
namely Coriaria, Cyathodes, Nertera, and Luzula, we find that the first
three hail from high southern latitudes, and more especially from New
Zealand; and when with this clue in our hands we take up the four genera
Weinmannia, Coprosma, Vaccinium, and Astelia, possessing only species
restricted to the Tahitian region, we find that all but the third-named
genus hail also from the south. It would thus appear that the element of
the Antarctic flora is much more evident in the Tahitian mountain-genera
than with those of Hawaii. In the Hawaiian mountain-flora, excluding, of
course, the endemic genera, it includes about a fourth of the
mountain-genera, which number about thirty-eight or forty in all; whilst
in the Tahitian mountain-flora it comprises six out of the eight genera.
It may, indeed, be said that the resemblance between the mountain-genera
of Hawaii and Tahiti is mainly restricted to genera that are found in
high southern latitudes, namely, Nertera, Coprosma, Cyathodes, and
Astelia, the only other genera linking the mountain-floras of both
groups together being Vaccinium and Luzula, which probably hail from
high northern latitudes. The agency of the frugivorous bird is plainly
marked in the case of five out of the six genera that connect the
cloud-capped peaks of Tahiti and Hawaii. In two of these genera,
Cyathodes and Nertera, the same species occurs in both archipelagoes.

_The Mountain-flora of Rarotonga._—A word may here be said on the
representation of these mountain-genera in Rarotonga, a small island
2,250 feet in height and about eight miles in length, which is, however,
the most important island of the Cook group. The recent important
explorations of Mr. Cheeseman show that its flora is essentially
Tahitian in character. As in Tahiti, the early age of the Compositæ and
Lobeliaceæ is well represented in the high levels by peculiar species of
Fitchia and Sclerotheca which are discussed in Chapters XXI and XXII. On
account, however, of its relatively low altitude and its small size, we
could not expect any extensive representation of the eight non-endemic
mountain-genera of Tahiti. Yet three of these occur, a Tahitian species
of Vaccinium (page 281) growing on its summits, whilst peculiar species
of Weinmannia (page 290) and Coprosma (page 295) are found in its
interior. The prevailing condition of many of the genera growing in the
higher levels is one of isolation, since other genera, like Pittosporum
and Elæocarpus, only possess peculiar species; but seeing that in
several cases the species are closely allied to others found in the
Western Pacific, as in Samoa, Fiji, and the Kermadec group, it is
apparent that the period of isolation has not long commenced.

THE MOUNTAIN-FLORA OF THE FIJIAN REGION.

Weinmannia, Saxifragaceæ, Fiji and Samoa. } Derived from
Lagenophora, Compositæ, Fiji. } New Zealand or
Coprosma, Rubiaceæ, Fiji. } from the
Astelia, Liliaceæ, Fiji and Samoa. } Antarctic flora.

Vaccinium, Vacciniaceæ, Samoa, from the northern hemisphere.
Nertera depressa, Rubiaceæ, Samoa, from the Antarctic flora.

Dammara, Coniferæ, Fiji. }Not as a rule belonging
Podocarpus, Coniferæ, Fiji and Tonga. }to the present age
Dacrydium, Coniferæ, Fiji. }of dispersal

But little can be said of the mountain-flora of Fiji, since on account
of the relatively low elevation of the islands there are but few special
mountain-genera; and as a rule we find only here and there a solitary
species on some isolated peak that recalls the upland flora of the
Hawaiian mountains. “None of the mountains of Fiji,” remarks Horne (page
60), “are high enough for an alpine flora to exist. Many of the plants
found on the tops of the mountains are also found near the level of the
sea. On the other hand sea-level plants may also be found on the tops of
the hills.”

Fiji lacks the endemic genera of Compositæ and of Lobeliaceæ that often
give a character to the mountain-floras of the Hawaiian and Tahitian
regions, though, as remarked in Chapters XXI and XXII., their absence
involves something more than a question of station. We find, however,
four genera of the Antarctic or New Zealand flora, Weinmannia,
Lagenophora, Coprosma, and Astelia. The first-named genus possesses four
or five species ranging up to 2,000 feet, some of which are endemic, and
it has been already discussed in this chapter. The United States
Exploring Expedition found a single species of Lagenophora (L.
pickeringii) on the mountains of the Mathuata coast of Vanua Levu, and
no other species seems to have since been found. The subject is dealt
with in Chapter XXIII in the case of Hawaii, but it may be here observed
that there is an Hawaiian mountain species, and that the route followed
by the ancestor of the Fijian species from the New Zealand home of the
genus is indicated by a species in the intermediate Kermadec group. The
genus Astelia has been discussed on page 291. It is represented in
Hawaii and in most of the oceanic groups of elevated islands. The
solitary species, A. montana, discovered by Seemann on the summit of
Kandavu in Fiji, has since been found in Samoa, and probably Mr. Horne’s
collections contain another species.

The Rubiaceous genus Coprosma needs a few special remarks, since a
particular genus of birds seems to have been concerned in dispersing it
in the South Pacific. About fifty species are enumerated in the _Index
Kewensis_, and if we include a few other species from the collections of
Hillebrand, Horne, Cheeseman, &c., the total would be about sixty. Of
these, about half are restricted to New Zealand, which may be justly
regarded as the home of the genus, the rest being confined to Australia
and the islands of the Pacific, excepting a Chilian and three or four
Malayan species. Hawaii with its nine species, Tahiti with two,
Rarotonga with one, and Fiji with two or three species represent
approximately the distribution of the genus in the oceanic archipelagoes
of the tropical Pacific. (It most probably exists on the high peaks of
Samoa, though it has not yet been recorded from the group.) In all, or
in almost all cases, the species are restricted to their particular
groups, so that we may regard the dispersal of the genus over the
Pacific as suspended, though, as will be observed below, the period of
suspension in the South Pacific has not been of sufficient duration to
obliterate the affinities of species in distant groups and to prevent us
from tracing out the route followed by the genus.

This genus of temperate latitudes, which in its New Zealand home ranges
from near the sea-level to the region of the alpine floras, finds its
usual station in the tropics on the summits of mountains. Thus, on Mount
Kinabalu, in Borneo, it is found at altitudes of 10,500 to 13,000 feet
(Stapf), and on the mountains of East Java at elevations exceeding 9,000
feet (Schimper). In Hawaii its species grow at elevations ranging from
3,000 to 9,000 feet, and in Tahiti at altitudes of 2,600 to 3,300 feet;
whilst in Rarotonga it grows in the hilly parts of the island, its
elevation in Fiji not being recorded.

When we come to consider the route by which the genus (Coprosma) entered
the tropical Pacific, we must remember that unless we establish some
special connection with its New Zealand home it will always be open for
any one to suggest that the genus might have been derived, like
Vaccinium, from other regions than the south, as from the summits of the
Malayan mountains. However, a curious connection has been discovered by
Mr. Cheeseman in his examination of the Kermadec and Rarotongan floras,
and it would indeed appear that he has traced the Rarotongan peculiar
species to its New Zealand home. Thus, he says that Coprosma lævigata,
his new Rarotongan species, is very closely allied to the Kermadec
endemic plant, C. acutifolia, Hook., which itself comes near C. lucida,
Forst., a New Zealand species. The connection between Rarotonga and New
Zealand by way of the Kermadec group is rendered yet more probable by
the occurrence of two New Zealand species of Coprosma in the Kermadec
flora (_Journ. Linn. Soc._ i. 1857; _Trans. Linn. Soc. Bot._ vi. 1903;
_Trans. N.Z. Instit._ xx. 1887).

When speaking of the genus in Hawaii (page 275), mention was made of the
inter-island dispersal of the fruits of one of the species by the native
mountain-goose, Bernicla sandwicensis. We learn from Sir W. Buller’s
_History of the Birds of New Zealand_ that when the Coprosma is in fruit
the Swamp-Hens (Porphyrio melanotus) come out to feed on it. These
birds, he says, are capable of prolonged flight; and I chance to have
beside me a cutting from the _Field_ of July 9, 1904, in which “Hy. S.”
refers to a Black-backed Porphyrio that was captured in 1876 four
hundred miles off the coast of New Zealand. This genus, which is widely
dispersed in the tropics, the birds being commonly known as Sultanas,
Blue Gallinules, Purple Water-Hens, &c., has probably been a very
important factor in the dispersal of plants, especially in connection
with insular floras. The birds live on a variety of food. The Messrs.
Layard observed that Porphyrio vitiensis, which abounds in the swamps of
New Caledonia, fed on maize, yams, &c. (_Ibis_, 1882); whilst in the
stomach of a bird of the same genus shot in the Rewa swamps in Fiji I
found a number of the stony fruits of Scleria, a genus of the Cyperaceæ.
According to Mr. Wiglesworth, each region in the South Pacific has its
own species of Porphyrio. There is one in the Tahitian Islands, and
another common to Fiji, Tonga, and Samoa; whilst New Caledonia and the
New Hebrides have their species (“Aves Polynesiæ”). However, it is
evident that the power of dispersing seeds from group to group is not
quite suspended, since, as we learn from Sir W. Buller, the New Zealand
species, above named as partial to Coprosma drupes, is distributed over
Tasmania and Australia, and reaches also Niue and New Caledonia; whilst
the Messrs. Layard evidently regarded one species as common to Fiji and
New Caledonia.

It is doubtless to birds of this description that we owe some of the
specific connections of Coprosma between groups of the Western Pacific.
That the dispersal of the species over distant regions was recently in
active operation is shown by the close affinity, according to Dr. Stapf,
of two species growing on the summit of Kinabalu, the Bornean mountain,
with certain species from New Zealand and South-east Australia. Other
Rubiaceous species, like Nertera depressa, possessing Coprosma-like
fruits and fitted for the same mode of dispersal, link the heights of
Kinabalu with the flora of high southern latitudes.

Being included in the Fijian area, the scanty mountain-flora of Samoa
may be here referred to. As in Fiji, the endemic genera of Compositæ and
Lobeliaceæ are not to be found, but we find in the central elevated
district of Savaii, which rises to over 5,000 feet above the sea, a
peculiar species of Vaccinium (4,900 feet), the Antarctic Nertera
depressa (4,000 feet), and two species of Weinmannia, a genus hailing
probably from high southern latitudes.

THE FIJIAN CONIFERÆ.

It has been found most convenient to discuss here these interesting
plants, which belong in a general sense to the mountain-flora of this
archipelago. That which the Fijian flora loses in interest in the eyes
of the student of plant-dispersal in not possessing the mysterious
Composite and Lobeliaceous genera of Hawaii and Tahiti, it regains in
the possession of its genera of Coniferæ. If he felt loth to apply his
empirical principles to the above-named Hawaiian and Tahitian endemic
genera, he feels more than uneasy when he comes to deal with the three
Coniferous genera of Fiji, Dammara (Agathis), Podocarpus, and Dacrydium.

These three genera represent an order that has not found a home either
in Tahiti or in East Polynesia generally, or in the more distant Hawaii;
and they present at first sight in their existence in Fiji a powerful
argument in favour of the previous continental condition of the islands
of the Western Pacific. But in advocating this view we should remember
that it involves the original continuity of the Fijian land-area, not
only with the neighbouring islands of the New Hebrides and of New
Caledonia where these genera alike occur, but also with New Zealand,
Tasmania, and Australia, where they sometimes attain a great
development.

In Fiji these trees often chiefly form the forests of the larger
islands, extending in the moister regions from near the sea to the
mountain-tops, and being often abundant on the great mountain-ridges of
the interior. It may be at once remarked that, viewed merely from the
standpoint of dispersal, there is no great difficulty in regarding it as
probable that the seeds of Podocarpus and Dacrydium have been dispersed
by frugivorous birds over tracts of ocean 500 or 600 miles across.
Dammara, however, so far as my Fijian observations show, possesses none
of the means of dispersal across oceans that we are at present
acquainted with. The two first-named genera occur in South America as
well as in the Australo-Polynesian region, some of the species in these
two regions, though the Pacific Ocean divides them, being closely
related. Dammara is, on the other hand, confined to a much more limited
area, extending from New Zealand to Borneo. It is from the distribution
of this genus that the continental theory derives its chief support.

Yet it may be remarked that something more than questions relating to
the capacity for dispersal are involved here. This is at once indicated
by the circumstance that although Podocarpus is known to be dispersed by
frugivorous birds, it is not found in Polynesia east of Tonga, and the
same may be said of Dacrydium, which does not occur east of Fiji. In
this connection it is necessary to notice the intrusion of Araucaria
into the tropical Pacific from Eastern Australia to New Caledonia and
the New Hebrides. The fact of this genus not having been recorded from
Fiji or any of the groups east of the New Hebrides is very remarkable,
and scarcely in accordance with the continental hypothesis. There is a
persistence in type of these genera of the Coniferæ during geological
time that prevents us from dealing with them on the lines that are
required by the mass of the flowering-plants. Other factors intervene,
and we apply with hesitation the same canons of dispersal that we employ
for the general bulk of the plants of the Pacific islands. If, as often
happens, a specific distinction alone separates the Conifers of the same
genus on either side of the Pacific Ocean, it must possess in point of
time a very different value from that which we would usually attach to
specific distinctions in the floras of the Pacific islands.

DAMMARA (AGATHIS).—The Dammara region includes Eastern Australia, New
Zealand, New Caledonia, with the New Hebrides, Fijian, and Santa Cruz
groups, and extends north-west to Java and Borneo. Only ten species are
named in the _Index Kewensis_, and of these four are assigned to New
Caledonia and two to Fiji, the focus of geographical distribution being,
therefore, as Seemann long since pointed out, in the islands of the
Western Pacific. The absence of the genus from the neighbouring Samoan
and Tongan groups is very significant; and it is evident that the
ordinary agencies of dispersal, whether birds, winds, or currents, have
here failed to extend the genus over a few hundred miles of sea.

When by means of observation and experiment we turn to the fruits and
look for a reply, we find in the first place that they are never to be
noticed either whole or in part in the floating drift of sea or river,
or amongst the stranded materials of the beaches. This is at once
explained when we ascertain that the fresh cones sink in the
river-water, and thus could never reach the coast in their entire
condition. Nor could they do so in fragments, since the detached cone
falls to pieces on the ground and the separate scales and seeds sink at
once or float only for a few hours. In order to test the buoyancy of a
cone after drying, it is necessary to bind it round with string to keep
it from breaking down. One such fruit, after being kept for ten days,
was placed in sea-water, where it floated heavily for eleven days and
then sank. This is, of course, a most unnatural experiment, but it was
well to have carried it out. That the entire fruit could never be
transported by water is indirectly implied by Kirk respecting the fruit
of Dammara australis, the Kauri Pine of New Zealand. In this case, when
the fruit reaches maturity the scales, he remarks, fall away from the
woody axis of the cone and the seeds are freed.

The fleshy, unprotected seeds, which, as above noted, possess little or
no floating power, could scarcely withstand the injurious effect of
sea-water; and they are absolutely unfitted for any known mode of
dispersal by birds. It is observed by Kirk that the seeds of the New
Zealand tree are widely spread by winds. But this could only avail them
for local dispersion, and they appear ill-suited for being transported
for more than a few paces. The seeds are winged, and are in form a
little like the samara of the Maple (Acer); but they have not the same
protective coverings, the wing being, however, only a little more than
half the length of the entire seed. Those of both Dammara australis and
D. vitiensis are about two-thirds of an inch in length, and are
heavy-looking; and the agency of the wind could never be invoked except
for local dispersion.

Looking at these results, the cones of Dammara may be regarded as most
unsuited for any of the ordinary means of dispersal over an ocean except
through the agency of man. There is, however, no necessity to introduce
man’s aid here, unless the gum or resin which the Fijian burns in his
torches and employs as a glaze for his pottery gave his ancestors an
object in carrying the cones with them in their migrations. But in that
case the same argument would have to be applied to all partially useful
plants, and much of the Fijian flora would lose its indigenous
reputation. The endemic character of the Fijian species also militates
against such a view, and we should have to apply the same explanation to
the New Zealand species, concerning which no one, so far as I know, has
ever ventured to suggest that it was introduced by the Maoris.

The native names of the trees seem to have been sometimes connected with
general words for gums or resins; whilst at other times the tree and the
resin have separate designations. Thus the Fijians call the tree
“ndakua” and the resin “makandre,” which last Hazlewood in his
dictionary seemingly connects with “ndrenga,” the word for “gum.” In my
work on the Solomon Islands, page 190, I have endeavoured to show that
the Maori name of “kauri” may be connected with “gatah,” the general
Malayan word for gums and resins, transitional stages being presented in
the names of resin-yielding trees in the intermediate regions, as, for
instance, by “gutur,” a species of Canarium, on the Maclay coast of New
Guinea, and by “katari,” a species of Calophyllum, in Bougainville
Straits, Solomon group. It may be pointed out that these facts of
plant-nomenclature do not promise us any aid in determining the mode of
dispersion of Dammara in the Western Pacific. There is a suspicious
resemblance between the Fijian name of “ndakua” and “dundathu,” the
Queensland aboriginal name for Dammara robusta; but even if the
comparison is legitimate, its explanation may lie far back in the ages
in some root-word as ancient as the Malayan “gatah.”

If there is a real difficulty in applying our canons of
plant-dispersal to the distribution of Dammara, it is merely the same
difficulty that has so often perplexed the botanist with other
Coniferous genera in continental regions, such as, for instance, the
occurrence of Pinus excelsa on the far-removed mountains of Europe and
of the Himalayas, and the existence of the cedar in its isolated homes
on the Atlas, the Lebanon mountains, and the Himalayas. Such
difficulties largely disappear if we regard the present distribution
of the Coniferæ as the remnant of what it was in an ancient geological
period. In the case of Dammara it seems almost as idle to puzzle over
its means of dispersal as to consider the mode of dispersal of the
Marsupials. The questions, indeed, that affect the Dammaras of Fiji
and the Western Pacific far ante-date any questions concerning a
previous continental condition of those regions. The attitude of the
palæobotanist to such questions would probably be one of indifference;
yet to the student of plant-distribution they are of prime importance;
and _nolens volens_ we must admit that Dammara may well be cited in
support of any continental hypothesis affecting the Western Pacific.

PODOCARPUS.—In this connection I will mainly depend on Pilger’s recent
monograph on the Taxaceæ (heft 18, Engler’s _Das Pflanzenreich_, 1903).
More than sixty species are here enumerated, which are distributed in
Africa, Asia, Australasia, and South America. With a range that extends
north to Japan and south to Southern Chile in latitude 48°, this genus
attains its greatest development in respect of species in Malaya, in the
region comprised by Australia, New Zealand, and New Caledonia, in South
America, and in Africa. Eastward of New Caledonia it is found in Fiji
and in Tonga, but not in Samoa, and it is altogether absent from the
Tahitian region as well as from Hawaii. Of the four species accredited
by Seemann to Fiji, two are enumerated by Pilger, namely, P. affinis and
P. vitiensis. The first-named, according to Stapf, is allied to P.
bracteata, which occurs on the upper slopes of Kinabalu, in Borneo, and
is distributed not only over Malaya, but occurs in Japan and in the
Himalayas. The Tongan species, P. elatus, is, according to Hemsley,
found in East Australia.

This Tongan tree is suggestive of bird-agency in the dispersal of the
genus, and the same may be said of the occurrence of another species, P.
ferrugineus, found in both New Caledonia and New Zealand. Since the
seeds of the genus possess an outer fleshy and an inner bony covering,
they would appear to be well fitted both to attract and to be dispersed
by birds. In fact, we learn from Sir W. Buller that the New Zealand
fruit-pigeon feeds on the seeds of the “matai” tree (Podocarpus spicata)
and of the “kahikatea” (P. dacrydioides), and no doubt to the agency of
frugivorous birds we can attribute the presence of the genus in Fiji and
Tonga. Yet it is strange that bird-agency should have failed both with
Tahiti and Hawaii. In point of size the seeds, which range from
one-quarter to an inch across, present no great difficulty, and one
would have thought that the birds that carried the “stones” of
Elæocarpus to Hawaii could have also carried the seeds of Podocarpus.

It is, however, necessary to remember, in dealing with a genus that has
a wide distribution both in time and space, that specific affinities may
have a very different significance with the Gymnosperms than with most
other flowering plants. When Hemsley remarks (_Introd. Chall. Bot._ p.
56) that the New Zealand Podocarpus spicata is closely allied to the
South American P. andina, he does not imply that the two regions are in
touch with each other though some 5,000 to 6,000 miles of ocean
intervene. One is prepared to credit these seeds with a capacity of
dispersal by birds over tracts of sea such as the extent of ocean
separating New Caledonia and New Zealand, which are some 900 miles
apart; but one hesitates to admit that frugivorous birds could carry
them across the Southern Ocean. If we assign a home in the high
latitudes of the northern hemisphere to a genus that was well
represented in Europe in the Tertiary period, a movement of migration
southward would explain most of the difficulties in its present
distribution. The great vertical range of some of the species leads us
to attribute a corresponding power of adaptation to the genus in respect
of widely different climates. Thus, according to Stapf, the vertical
range of P. bracteata in the Malay Archipelago extends, including
varieties, from the coast to an altitude of 12,000 feet. With such a
capacity for adaptation, migrations of the genus would be rendered easy
over the globe.

DACRYDIUM.—It may happen that some additional light on the mystery of
the Fijian Coniferæ may be afforded by Dacrydium elatum, a tree that
occurs not only in Fiji, but in Further India and in Malaya. Pilger
confirms Seemann’s view in his identification of the Fijian tree, and
this opinion is, in the main, shared by Stapf. This species, so to
speak, affords us a _point d’appui_ in the history of the distribution
of the genus in the Western Pacific. This distribution somewhat
resembles that of Dammara in extending from New Zealand (its principal
centre) to Malaya and Further India; but, unlike Dammara, Dacrydium is
represented in America by a solitary species in South Chile. Of the
sixteen species enumerated by Pilger, seven belong to New Zealand, four
to New Caledonia, three to Malaya, one to Tasmania, and one to Chile.
The seeds are, as a rule, smaller than those of Podocarpus, and on
account of their somewhat similar structure would serve as bird-food,
and might be distributed in this fashion. Yet the genus has been only
recorded from Fiji, and is not only unrepresented in Hawaii and Tahiti,
but is also not known from the Tongan and Samoan groups that belong to
the Fijian floral region of the Pacific. Capacities for dispersal appear
meaningless here, especially when we have regard to the solitary
American species, Dacrydium fonkii, that as a shrub finds a refuge in
the bleak region of Southern Chile.

The three Fijian genera of the Coniferæ, Dammara, Podocarpus, and
Dacrydium, appear at first sight to be beyond the reach of our canons of
plant-dispersal, by which we connect specific affinity with a continuity
of range, and by which we co-ordinate means of dispersal and area of
distribution. We begin to realise that there may have been an age of
Coniferæ in the Pacific islands that is even less amenable to our
methods than the later era of the Compositæ and Lobeliaceæ in Hawaii and
Tahiti. Such an age would be concerned only with that region in the
Western Pacific which is now held by the genera Dammara, Podocarpus, and
Dacrydium, a region that did not participate in the era of the Compositæ
and Lobeliaceæ. We thus have evidence of an ancient era of the Coniferæ
that was confined to the Western Pacific, and of a later era indicated
by the peculiar genera of Compositæ and Lobeliaceæ that was restricted
to Hawaii and to Eastern Polynesia (Tahiti, Rarotonga, &c.). The key to
the situation here presented seems to lie in the following
considerations.

It is assumed that there was an age of Coniferæ in the Pacific, or
rather that this region shared in an era of dispersion of existing
genera of the order. In this age only the islands of the Western Pacific
participated, neither the Hawaiian nor the Tahitian islands taking a
part in it. Such a result is to be attributed either to the inability of
these genera of Conifers to reach Hawaii and the islands of East
Polynesia, or to the non-existence of the Hawaiian and Tahitian
archipelagoes at that epoch. The first explanation seems scarcely
acceptable, since, although the powers of dispersal of the genus Dammara
are very limited, there seems no reason why the genera Podocarpus and
Dacrydium could not have reached those distant regions of the Pacific.
The second explanation is most probable, and it is the one suggested by
Hillebrand (p. xxx) in the case of Hawaii, namely, that “the absence of
Gymnosperms militates for the view that the islands were formed
subsequent to the age in which these were universally distributed.”

If this conclusion is legitimate we have here a datum-mark in the
history of the islands of this ocean. Before the appearance of the
Hawaiian and Tahitian islands (using the term Tahitian to cover the East
Polynesian region) there existed a land-area in the Western Pacific held
by the Coniferæ, probably in the late Secondary period. After the
formation of the Hawaiian and Tahitian islands, perhaps in the early
Tertiary epoch, came the age characterised by the ancestors of the
present endemic genera of the Compositæ and Lobeliaceæ, and of a few
other orders in Hawaii and Tahiti. In this age the islands of the
Western Pacific do not seem to have participated, and it is to be
inferred that this was an age of extensive but probably not of complete
submergence in that part of the ocean, since at least the genus Dammara
was able in places to hold its ground. Then ensued the great Tertiary
emergence of the land-areas of the Western Pacific, when small islands
that dotted the sea-surface in this region became the nuclei for the
formation of the large islands of the present Fijian, New Hebrides, and
Solomon groups. This prepared the way for the migration of Malayan
plants which now predominate over the islands of the tropical Pacific;
and in a later age man, following the same track from Indo-Malaya,
occupied these islands.

In my volume on the geology of Vanua Levu it was shown that the Tertiary
period was an age of submergence in the Western Pacific, and a disbelief
in any previous continental condition was expressed. My later view is
more in accordance with that of Wichmann, who, on geological grounds,
contended that the islands of the Western Pacific were in a continental
condition during the Palæozoic and Mesozoic periods, and that their
submergence and subsequent emergence took place in Tertiary times. The
distribution of the genus Dammara has thus led me to modify the views
expressed in the final chapter of my first volume on the geology of
Vanua Levu. Though still holding that there is no geological evidence
that the various islands of the Fijian group were ever amalgamated, or
that they were joined as such to the westward groups, it is quite
possible that their position was indicated by a few small islands a few
miles across and a few hundred feet in height in early Tertiary times.
On these small islands, which probably represented the remains of a
submerged Mesozoic land-area, such as is in part implied in Dr. Forbes’
_Antipodea_, or in Mr. Hedley’s _Melanesian Plateau_, the genus Dammara
survived. Such islands merely indicated the situation of some of the
present groups of the Western Pacific, which have been since largely
built up by submarine eruptions, and the greater number of the islands
were no doubt completely submerged. Between the groups as we know them
now there never was any land connection, since they are the product of
later eruptions, mainly submarine; and they have acquired their present
composite character during the emergence that followed the period of
volcanic activity. Except, perhaps, in New Caledonia, which does not
seem to have shared in the Tertiary submergence, the islands of the
Western Pacific have a configuration acquired in comparatively recent
times, and one that gives no idea of the character of the Mesozoic
continent.

Such, as I understand them, are the indications of the Fijian Coniferæ
and particularly of Dammara. In the distribution of this genus we have
outlined an ancient, more or less continuous land area which, with the
exception of a few isolated points, disappeared beneath the sea in
Tertiary times to re-appear near the close of that period in the form of
a number of archipelagoes that were largely built up by submarine
eruptions, and probably altogether mask the form of the original
land-area. It may be remarked that New Zealand, which largely shared in
the Tertiary submergence, especially in the Miocene age, is included in
the range of the genus Dammara, as well as in those of the genera
Podocarpus and Dacrydium.

_Summary._

(1) The evidences of a mountain-flora in Tahiti, as indicated by the
non-endemic genera, though, as we would expect, of a scanty nature when
contrasted with Hawaii, are nevertheless of considerable interest. There
is much kinship with the Hawaiian mountain-flora, but it is mainly
confined to genera from high southern latitudes, such as Nertera,
Coprosma, Cyathodes, and Astelia, which are all dispersed by frugivorous
birds. Amongst other plants linking the Tahitian mountains with the
region of the Antarctic flora, and with New Zealand in particular, may
be mentioned Coriaria ruscifolia and the genus Weinmannia.

(2) On account of their relatively low altitude the Fijian islands do
not present the conditions for an alpine flora. Traces, however, of the
Antarctic flora, or of the New Zealand flora, occur on occasional
mountain-tops, as is indicated by the occurrence of species of
Lagenophora, Coprosma, and Astelia. In Samoa the mountain-flora is also
scantily developed, as we might have expected; but here occurs the genus
Vaccinium as well as a widely-ranging species of the Antarctic flora,
Nertera depressa.

(3) The route by which some of the representatives of the flora of high
southern latitudes reached the mountains of the islands of the tropical
Pacific is directly indicated by the genus Coprosma to have been from
New Zealand by way of the Kermadec Islands.

(4) In the distribution of plants possessing drupes or berries that
connect the tropical islands of the South Pacific with New Zealand, it
is highly probable that birds of the genus Porphyrio (Swamp-Hens or
Purple Water-Hens) have taken a prominent part.

(5) In the possession of species of the three genera of Coniferæ,
Dammara, Podocarpus, and Dacrydium, which often largely form the forests
of the mountain-slopes, Fiji is distinguished from all the other groups
of the open Pacific with the exception of Tonga, which owns a species of
Podocarpus probably introduced by birds. From the circumstance that
Dammara has no known means of crossing a tract of ocean, whilst
Podocarpus and Dacrydium could be dispersed by frugivorous birds, all
three genera having, however, much the same limited distribution in the
Western Pacific, it is apparent that something more than a question of
means of dispersal is here involved. It is assumed that they mark the
site of a Mesozoic continental area in this region, and that at this
period the Tahitian and Hawaiian groups which possess no Conifers did
not exist. This area was submerged during the Tertiary period with the
exception of a few peaks that formed small islands on which the Conifers
held their ground. During the Tertiary submergence of the Western
Pacific region, the Hawaiian and Tahitian islands were built up by
subaërial volcanoes and received the ancestors of the Compositæ and
Lobeliaceæ that now exist as endemic genera in those groups. Then
followed the emergence of the islands of the Western Pacific and their
occupation mainly by Indo-Malayan plants that extended eastward over the
Pacific. Thus in the Pacific there has been first an age of Conifers in
which the islands of the Hawaiian and Tahitian regions could not
participate, since they did not exist. Then ensued an era of American
forms of Compositæ and Lobeliaceæ in which only Hawaii and Tahiti
participated, since the Western Pacific region was submerged. Lastly
came the invasion of Indo-Malayan plants, which have largely occupied
every group in the tropical Pacific.

CHAPTER XXV

THE ERA OF THE NON-ENDEMIC GENERA OF FLOWERING PLANTS (_continued_)

THE AGE OF THE MALAYAN PLANTS AS REPRESENTED IN THE LOW-LEVEL FLORA OF
HAWAII AND IN THE BULK OF THE FLORAS OF THE FIJIAN AND TAHITIAN REGIONS

_The Age of Wide Dispersal over the Tropical Pacific._

The widely dispersed genera which possess only peculiar species in
Hawaii.—Pittosporum.—Reynoldsia.—Gardenia.—Psychotria.—Cyrtandra—Freycinetia.—Sapindus.—Phyllanthus.—Pritchardia.—Summary.

WE pass now from the consideration of the mountain-flora of Hawaii and
its scanty representation in the Fijian and Tahitian regions to a
discussion of the low-level Hawaiian flora, belonging to stations under
4,000 or 5,000 feet, and of the corresponding floras of the other two
regions. It has been previously pointed out that in mass the plants of
Fiji and Tahiti correspond to the low-level flora of Hawaii.

There are numerous ways of comparing this era of the non-endemic genera
of these three regions of the Pacific. The necessities of space,
however, compel me to treat the subject only in an illustrative fashion,
and in adopting the plan which seems easiest and simplest I have also
been obliged to keep my limitations mainly in view.

THE WIDELY-DISPERSED GENERA WHICH POSSESS ONLY PECULIAR SPECIES IN
HAWAII.

Amongst the oldest denizens of the Pacific islands in this era of the
non-endemic genera may be taken those genera of flowering plants which
are found in all three regions, Hawaii, Fiji, and Tahiti, but possess in
the first group only endemic species, whilst in the other two regions
they may include species both confined to and occurring outside the
respective groups. They represent an age of wide dispersal over the
Pacific, an age which for Hawaii has long since passed away, since all
the genera have been disconnected from the outer world, whilst in the
groups of the South Pacific they as a rule in each case remain in touch
through some of the species with the groups around.

The problem of plant-distribution in the Pacific thus assumes a
different aspect in an age which we term Malayan or Indo-Malayan, since
the bulk of the plants are thence derived. The earliest age of the
Coniferæ was, as we have seen in the previous chapter, restricted to the
region of the Western Pacific. The following age of the Compositæ and
the Lobeliaceæ was concerned with the regions of Tahiti and Hawaii. Now,
however, in the Malayan era, the whole of the tropical Pacific is
concerned. Yet, although we shall still regard, for purposes of
convenience, the groups of Fiji, Tahiti, and Hawaii as the three foci of
plant-distribution, it will soon become apparent that in future there
will be in reality only two regions to deal with, the Hawaiian in the
North Pacific, and the whole region of the South Pacific extending from
Fiji to Tahiti and as far east as the islands stretch. It will be also
seen that in making our comparison we shall sometimes have to regard
each of the principal Hawaiian islands as the equivalent as a
plant-centre of an entire archipelago of the South Pacific.

The genera that are here selected to represent this epoch of wide
dispersion are very characteristic of the floras of the Pacific islands.
Genera like Pittosporum, Gardenia, Psychotria, Cyrtandra, Freycinetia,
and others one meets with everywhere in the larger islands, and it
should be observed that they are predominantly Old World, and more
especially Malayan, in their origin, not a single purely American genus,
unless we except the decadent genus of fan-palms, Pritchardia, occurring
among them. Here we notice [what we shall see is especially typical of
the era of the non-endemic genera, excepting those of the lofty uplands
of Hawaii] that the frugivorous bird has been the principal agent in
dispersing the plants, quite two-thirds of the total genera possessing
drupes or berries that would attract such birds. The transport of seeds
or seedvessels in birds’ plumage, which was a conspicuous feature in the
case of the mountain-flora of Hawaii, is not a feature of this age of
wide dispersal of tropical plants over the Pacific.

The genera selected to represent this age are given in the following
table. Those on which my observations directly bear, or in which I was
particularly interested when in the Pacific, will be discussed in detail
from the standpoint of dispersal; whilst only a brief reference will be
made to a few of the others, not, however, from lack of materials at my
disposal, but merely to keep this volume within moderate bounds.

_Genera selected to represent the Age of Wide-dispersal of Indo-Malayan
or Malayan Plants over the Pacific, and possessing in Hawaii only
Endemic Species._—Most of the genera of this age are exclusively from
the tropics of the Old World, whilst those found on both sides of the
Pacific can be shown in most cases to have been derived from the same
source, and only very few, like Pritchardia, can be traced to America.

Pittosporum (Pittosporeæ).
Sapindus (Sapindaceæ).
Reynoldsia or Trevesia (Araliaceæ)
Gardenia (Rubiaceæ).
Psychotria (Rubiaceæ).
Cyrtandra (Gesneraceæ).
Phyllanthus (Euphorbiaceæ).
Pritchardia (Palmaceæ).
Freycinetia (Pandanaceæ).

PITTOSPORUM (Pittosporeæ).

This genus, which contains nearly a hundred species, usually of small
trees, is widely spread in the warmer regions of Africa, Asia,
Australia, and New Zealand. It is also especially a genus of oceanic
islands, occurring not only in those of the Pacific but also in Madeira
and Teneriffe in the Atlantic.

Though found in most of the larger Pacific groups, it has apparently
never been recorded from Samoa. From Hawaii ten species are known, all
peculiar to that group. About half a dozen have been described from
Fiji, of which three at least have been observed outside the group in
the neighbouring Tongan Islands. Rarotonga possesses a peculiar species
which, however, is so near to two other Fijian and Tongan species that,
according to Cheeseman’s memoir, they may have to be subsequently
united. Tahiti is credited by Drake del Castillo with a solitary species
widely distributed in the Old World, whilst in the _Index Kewensis_ a
peculiar species is assigned to it. They form small trees of the wooded
mountain-slopes of Fiji; whilst in Hawaii, beside occurring in the lower
forests, they may extend to altitudes of between 5,000 and 7,000 feet.
In the connection that more or less exists between the species of the
South Pacific archipelagoes, and in the endemic character of all the
Hawaiian species, we see the principle exemplified that there are two
regions of distribution in the islands of the tropical Pacific—the
Hawaiian region and the South Pacific region.

Before their dehiscence, the wrinkled, woody capsules would seem very
unlikely to attract birds; but the observer on handling an opening
fruit, with its orange or brightly coloured lining and displaying black
or dark-purple seeds immersed in a semi-liquid pulp, would form a
different idea of the plant’s capacity for this mode of dispersal. The
mature dehiscing fruits are very conspicuous on the tree; and the seeds
covered with the “sticky” material of the pulp might possibly adhere to
birds pecking at the fruit. But this would only aid in local dispersion,
since the weight and size of the seeds, 5 to 8 millimetres (2/10 to 3/10
inch), would unfit them for this mode of transport across an ocean. They
are, however, sufficiently protected by their hard tests to be able to
pass unharmed through a bird’s intestinal canal.

Yet the distribution of the species of Pittosporum in the Pacific would
show that their dispersal is more a matter of the past than of the
present. Out of the ten peculiar Hawaiian species, Hillebrand designates
none as generally distributed over the group. But it is evident that,
though it is on the point of breaking off, some sort of connection still
exists in the South Pacific between the Tongan and Fijian species, and
until recently between the species of those two groups and of Rarotonga.

REYNOLDSIA (Araliaceæ).

The Polynesian genus of Reynoldsia, originally established by Gray, is
merged by Hooker and Bentham into the Malayan genus Trevesia, a step
that brings the Pacific plants into line with many other of the plants
hailing originally from the Old World. The significant fact in the
distribution of this genus of small trees in the Pacific is that its
dispersal over the ocean has ceased long ago, since the three species
here occurring are restricted each to a particular group, namely, to
Hawaii, Tahiti, and Samoa. Yet the inter-island dispersal still
continues in the Hawaiian Group, the species characteristic of that
archipelago being found in all the islands.

Reynoldsia sandwicensis came frequently under my notice in Hawaii, and
the fairly fleshy drupes, about one-third of an inch, or 8 millimetres,
in size, with their crustaceous pyrenes appeared to me well fitted for
assisting the dispersal of the plant by frugivorous birds. Yet here the
same question arises that presents itself with so many other Hawaiian
plants, and that is, How has it happened that the birds have continued
to disperse the species over the scattered islands of this group long
after they ceased to transport fresh seeds from the outside world? The
answer is an obvious one. The birds that originally brought the seeds of
the parent species from some distant region came at last to remain
permanently in the Hawaiian Group, and not only the plant but probably
also the bird has since undergone specific differentiation. This link
between bird and plant in the floral history of a group of Pacific
islands is the common theme of the story of most of the endemic species
of plants in this region of the globe.

GARDENIA (Rubiaceæ).

This genus, comprising about a hundred known species, is spread over
tropical Africa, Asia, and America, and over all the groups of the
tropical Pacific. On account of their handsome, white, scented flowers
these shrubs are much appreciated by the Pacific islanders, who employ
the flowers for personal decoration. Some ten species have been
described from the groups of the open Pacific, all of which, with the
exception of Gardenia tahitensis, which ranges the South Pacific from
Fiji to the Marquesas and Tahiti, are seemingly peculiar to the
different archipelagoes. Thus there are some six species endemic to
Fiji, one to Samoa, and two to Hawaii.

The Hawaiian Islands are, however, quite isolated in this respect, since
the group possesses only peculiar species; whilst a solitary species
keeps up the connection between the groups on the south side of the
equator. The Gardenias thus tell the same story of complete isolation in
Hawaii, and of partial isolation in the archipelagoes of the South
Pacific that is repeated by many other Pacific genera. Yet in Hawaii
there has subsequently been some inter-island dispersal, since the
species are not restricted each to a single island, but are found on two
or three islands. The significance of the relation of the Hawaiian
Gardenias to those of the combined Fijian and Tahitian areas consists in
regarding the two regions, the Hawaiian and the South Pacific, as of
equivalent value, and each large Hawaiian island as equivalent to one of
the southern archipelagoes.

_The Station of the Pacific Gardenias._—Although they may occur in the
forests, the Gardenias of the Pacific are most characteristic of dry,
thinly vegetated localities, and they have an inclination for the
vicinity of the coast. In the Tahitian Group, as we learn from the
writings of Nadeaud and Drake del Castillo, Gardenia tahitensis thrives
much better on coral islands than on volcanic soils, and, in fact,
rarely quits the “_région madréporique_.” It is sometimes planted in
Polynesia near the houses, and both Nadeaud in Tahiti and Cheeseman in
Rarotonga consider that it was probably introduced into those islands
before the arrival of Europeans. The aborigines may have assisted in the
dispersal of the genus to a small extent, but from the presence of
peculiar species in Hawaii, Samoa, and Fiji it is apparent that the
genus is truly indigenous in the Pacific islands, and long antedated
their occupation by man. This is also evident from the station of the
species in Hawaii, Samoa, and Fiji. In Hawaii they may be found on the
dry forehills in the vicinity of the sea-border. In Samoa, as Reinecke
informs us, Gardenia tahitensis is very widely spread in the
mountain-forests, whilst the endemic species is found thriving in
inundated coast districts. In Fiji I found the Gardenias to be
especially characteristic (as is also pointed out by Horne) of the dry
districts on the leeward side of the larger islands. On the rolling
“talasinga” or “sun-burnt” plains of the north side of Vanua Levu they
thrive in numbers; and here their leaf-buds and the extremities of the
young shoots are often tipped or covered over with an amber-like
gum-resin which the natives chew.

_The Mode of Dispersal of the Pacific Gardenias._—The fruits of this
genus are usually described as indehiscent. If this were true of Pacific
plants it would be very difficult to explain the dispersal of hard, dry
fruits an inch in size over this region. In the case of two or three
Fijian species, I paid especial attention to this point by examining the
plants in fruit. As exhibited in Fiji the fruits are globose, hard, and
almost stony, with persistent adherent calyx, the seeds lying
horizontally in a pulp at first firm and subsequently softening as the
fruit matures. The fruits are not as a rule to be observed opening on
the plant; but they are to be seen dehiscing septicidally on the ground
beneath, the detached woody valves being scattered around. If one of the
fruits gathered from the plant is kept soaking in water for some time it
will begin to dehisce; and this is probably what occurs with fallen
fruits in wet weather. Dr. Hillebrand regards the fruits of the Hawaiian
species as indehiscent. I did not myself examine them, but it is not
improbable that, like those in Fiji, they dehisce whilst lying soaking
on the ground.

Judged merely from the dispersal standpoint, the fruits of the Fijian
Gardenias come near to those of Pittosporum, and both can be in a sense
described as baccate capsules. The flat, crustaceous seeds of Gardenia,
which are usually two or three millimetres in size, are also well fitted
for passing without injury through the digestive canal of a bird. It is
likely that the two genera have been dispersed in the Pacific by the
same kind of birds; and it should be remarked that their distribution is
somewhat similar, both belonging to the warm regions of the Old World.

It might at first appear from some experiments of mine made in Fiji that
the dried fruits of Gardenia could be dispersed over oceans by the
currents. This receives some support by the preference for a littoral
station sometimes shown by G. tahitensis in Tahiti, and by the
occurrence of G. zanguebarica in the East African strand-flora
(Schimper’s _Ind. Mal. Strand-flora_, p. 131). It will, however, be
pointed out that currents could only have aided the dispersal of the
genus to a limited extent. The fresh fruits of Fijian species, with or
without the adherent calyx, have little or no buoyancy, and the seeds
sink even after drying for months. But it was ascertained that fruits
which had been kept for three months floated after four or five weeks’
immersion in sea-water. On examination, however, it was found that the
valves gaped a little, being only held in apposition by the adherent
calyx, and that water had penetrated into the interior, the pulp being
in a state of decay. The fruits were, in fact, kept afloat in the latter
part of the experiment partly by the investing calyx and partly by gas
generated in the decomposing pulp. Ultimately they broke down altogether
and the seeds sank. In the “rough-and-tumble” of ocean-transport this
could scarcely be deemed an effective means of dispersal; and in the
open sea a fortnight would probably represent the limit of the floating
power. It is to the agency that has distributed the genus Pittosporum
over the Pacific that we must look for the explanation of the dispersal
of Gardenia over the same ocean, namely, to birds.

PSYCHOTRIA (Rubiaceæ).

We find in this large genus of the Old and New Worlds a typical example
of the plants with fleshy drupes containing hard pyrenes that represent,
from the standpoint of dispersal, a common Rubiaceous type of plant in
the tropical Pacific. Such plants, of which those of Coprosma and
Nertera may be cited as other instances, are in a generic sense always
widely distributed in these islands. They are eminently suited for
dispersal by frugivorous birds; and it is a matter for surprise,
therefore, that in a genus like Nertera the solitary Pacific species has
such a wide range, whilst with Psychotria and Coprosma the numerous
species are usually restricted to particular groups. Genera doubtless
have their periods of development and decadence in the Pacific, and
probably Nertera is to be regarded as a decadent genus. These Rubiaceous
genera, however, appear to be well fitted for the investigation of the
centres of dispersal of particular genera and of their relative age.

The Psychotrias in these islands are typically shrubs of the shady
woods, and they may be seen thriving best where the forest-growth is
rank and the humidity greatest. Their bright red ovoid drupes, which
range from eight to twenty-five millimetres in length (1/3 to 1 inch),
would readily attract birds, and their crustaceous pyrenes, that vary
between five and eight millimetres (1/5 to 1/3 inch) in length, would
pass unharmed through a bird’s digestive canal. That fruit pigeons can
distribute their seeds over the Pacific has been long established, and
Mr. Hemsley includes Psychotria amongst those genera which, from the
collections of fruits and seeds found in the crops of fruit-pigeons,
made by Professor Moseley, myself, and others, in the groups of the
Western Pacific, are “known to be dispersed by birds in Polynesia”
(_Introd. Bot. Chall. Exped._, p. 45). It is thus hardly necessary to
point out that neither the entire fruits nor the separate pyrenes could
be transported by the currents, my observations showing that in both
cases they sink at once or in a day or two.

Psychotria, however, is an enormous genus including, according to the
_Index Kewensis_, some 600 or 700 described species, distributed in the
tropics all over the world, and also extending into subtropical regions,
the greatest concentration being in America. It is described in the
_Genera Plantarum_ as a polymorphous genus distinguished by no certain
characters from some other genera of the tribe of the Rubiaceæ to which
it has given its name. We have here a genus that has overrun the
tropical regions of the world, probably originating in America; and we
may contrast it with the relatively small Rubiaceous genus of Coprosma
(with its three score of species, and quite comparable with it from the
standpoint of capacity for dispersal), that, having its birthplace in
New Zealand, is only beginning to reach the mainlands of the New and the
Old World.

One is a genus of the tropics and the other is a genus of south
temperate latitudes; and both have occupied the Pacific islands; but
Coprosma naturally finds its most appropriate station on the cool
uplands of Hawaii and Tahiti. We may ask, indeed, whether the great
contrast in the fecundity of the two genera, dispersed as they are in
the same fashion by the agency of frugivorous birds, is to be connected
with questions of relative antiquity or with geographical position. It
would certainly have been a more difficult task in the past, other
things being similar, for a New Zealand genus to stock the temperate
regions with its species than for a tropical American genus to overrun
the warmer regions of the globe. However that may be, the age of
dispersal of both genera is largely over now.

A vast genus like Psychotria, that is not sharply defined from other
genera, presents difficulties to the systematic botanist which are
reflected in a complex synonymy; but there are certain broad facts which
the student of dispersal can gather for himself without much difficulty.
When we look at its distribution in the islands of the open Pacific, we
find that the genus attains its greatest development in the Western
Pacific, there being from thirty to forty species known from Fiji and
quite a dozen from Samoa, and that it shades away as we proceed eastward
and northward, some six species being recorded from Tahiti and the
Marquesas, two from Hawaii, and one from Juan Fernandez near the South
American mainland. The arrangement of the species shows fairly
conclusively that the genus Psychotria, as it is found in the Pacific,
has, like most of the other plants of this era of non-endemic genera,
been derived from the Asiatic side of the ocean. (The absence of species
of this genus from Mr. Cheeseman’s Rarotongan collections seems strange.
It is represented by some species in Tonga, and it is extremely probable
that it will be subsequently found also in the Rarotongan group.)

That the age of dispersal of the genus Psychotria over the Pacific
islands has almost passed away is evident from the circumstance that of
the half-hundred species known from these groups, all but some four or
five are confined to particular groups. There is one species, P.
insularum, that ranges over the South Pacific from Fiji to the Tahitian
region; and there are two or three others that keep up a connection
between the adjacent groups of Fiji, Samoa, and Tonga, the last having
no peculiar species; but, apart from these indications, isolating
influences generally prevail. The two Hawaiian species are both endemic
and are only recorded from the island of Kauai, so that in that
archipelago there has not even been inter-island dispersal of the genus.
For Fiji it would seem from the _Index Kewensis_ and other authorities
that at least two-thirds of the species are confined to the group. Of
the dozen Samoan species only two or three are known outside the
islands. Four out of the five Tahitian species are peculiar, and the
only Marquesan species named by Drake del Castello is endemic. Even the
solitary species of Juan Fernandez is endemic, there attaining the
dimensions of a fair-sized tree. It forms the subject of an illustration
in Schimper’s _Plant-Geography_, page 491.

Speaking generally, birds may be said to have almost ceased dispersing
this genus over the Pacific. This is not because birds have ceased to be
partial to the fruits, but because the frugivorous birds that used to
range over the Pacific archipelagoes now restrict their wanderings to
the limits of a single group. If we find occasionally in other parts of
the world, as in the occurrence of a Florida species of Psychotria in
the Bermudas, some evidence of a dispersal still in operation, this is
nothing more than we observe in the case of a few of the Polynesian
species now. The connection between birds and plants in the Pacific is
discussed in Chapter XXXIII. In this ocean the dispersal of the genus is
now practically dead, and Psychotria presents no exception to that
general tendency towards isolation and differentiation exhibited by most
genera of the tropical Pacific as the result of failure of the means of
dispersal.

CYRTANDRA (Gesneraceæ).

This remarkable genus of shrubs, which forms the subject of an important
memoir by Mr. C. B. Clarke (_De Cand. Mon. Phan._ v. 1883-87), offers,
as Mr. Hemsley remarks, an example of a Malayan genus extending to
Polynesia and there developing numerous species. Of some 180 known
species, about 80 or nearly half are confined to Polynesia, the rest
being mainly Malayan. Of the Polynesian species about thirty are
Hawaiian, twenty Fijian, fifteen Samoan, and twelve Tahitian; whilst
solitary species are restricted to Tonga and Rarotonga respectively.

The most significant feature in the distribution of this genus in
Polynesia is not only, as is pointed out by Mr. Clarke, that every group
has its peculiar species, but that very few species are found in more
than one group, and that even in the same archipelago each island has
its own species. Thus, of the thirty Hawaiian species, all of which are
peculiar to the group, only two or three, according to Hillebrand, are
at all generally distributed over the islands, whilst four-fifths have
not yet been found to be common to more than one island. So again, all
the species found in the Tahitian Group proper are peculiar, with the
exception of one extending to the neighbouring Paumotu Islands; and even
Rarotonga has its own species. In the region comprising Fiji, Tonga, and
Samoa the same rule prevails, only two or three species connecting the
three groups together. There thus seems to be not only a complete
suspension of the dispersal agencies between the various archipelagoes,
but also often between the several islands of a group. This is
particularly to be remarked with the relatively contiguous groups of
Fiji, Samoa, and Tonga, since with most other genera a number of species
are common to all three archipelagoes. “The polymorphism of the Hawaiian
Cyrtandras,” says Hillebrand, “is extraordinary: no single form extends
over the whole group, and not many are common to more than one island.
The variations affect nearly every part of the plant, and branch out and
intercross each other to such an extent that it is next to impossible to
define exact limits of species.” Genera, however, run riot in other
groups of the Pacific besides Hawaii, and Reinecke uses much the same
language with reference to Elatostema, an Urticaceous genus in Samoa,
attributing the wealth of forms to the sensitiveness of the plants to
the varying conditions of station (see Chapter XXVII).

The behaviour of Cyrtandra in the Pacific is rather startling to the
student of plant-dispersal when he reflects on the suitability of the
berries for dispersing the plant through the agency of birds. That the
vegetation of oceanic islands should be of an endemic character is a
fact, remarks Mr. Clarke, that is illustrated by many other orders
besides the Gesneraceæ. But the point we have to remember is that not
only does the genus Cyrtandra display the same prolific character in the
large continental islands of Malaya, such as Java, Sumatra, and Borneo,
each of which possesses at least a couple of dozen species, but that
this seems to be a feature of the tribe Cyrtandreæ and of the whole
order. The genera, as observed by Mr. Clarke, are very continuous in
their areas of distribution, and in the tribe Cyrtandreæ there are very
few species that extend to more than one region, whether on the mainland
or in an oceanic archipelago. In the Himalayas, he says, closely allied
species of Didymocarpus are confined to single districts, although there
appears no reason either in soil or climate why they should not spread
to the adjacent valleys.

There is therefore, we may infer, nothing peculiarly characteristic of
insular floras in this prolific display of the genus Cyrtandra in the
Pacific, except that it is rather more pronounced in an oceanic group
than in a continent. The same general cause is working alike in an
island in mid-ocean, in a large continental island bordering the
mainland, and on the mainland itself. With the Pacific Cyrtandras as
with the British species of Rubus the variability may be so great that
the ordinary agencies of dispersal fail to keep it in check; and when,
as in the Pacific islands, the suspension of the activity of these
agencies is complete, the formative energy of the species knows no
bounds other than the determining limits of station. Our lesson from the
Pacific Cyrtandras is therefore this. The isolation of the oceanic
archipelagoes may not explain the endemic character of the flora, but
only the extreme degree to which the endemism is carried. When a genus
is in its prime, it can defy all the limiting conditions imposed by
similarity of station and by free and unchecked means of dispersal, the
essential marks of a species or a genus having probably in their
development little or no connection with environment.

The Cyrtandras of the Pacific Islands are most frequent where vegetation
is rank, as in moist woods, in humid valleys, and in shady ravines and
gorges; but they may also occur in more exposed and drier stations. They
often grow gregariously, and Schimper says the same of them in the Java
forests (_Plant-Geography_, pp. 291, 297).

The fruit of the genus is described by Clarke as a fleshy or a
coriaceous berry. Almost everywhere in the Pacific groups the berry is
white and fleshy; but it is noteworthy that out of the nine Tahitian
species where the fruit is particularised by Drake del Castello, in two
cases it is designated a capsule and in seven a berry. It is in this
connection worth remarking that in Malaya other genera of the tribe
often have capsular or dry and coriaceous berries. The conspicuous white
berries of the Pacific species would readily attract birds, and their
minute roughened seeds scattered through the pulp might readily adhere
to their plumage or even be ejected unharmed in their droppings. As
respecting the capacity for dispersal, the Pacific Cyrtandras come near
the Hawaiian endemic genera of Lobeliaceæ with baccate fruits and minute
seeds. Speaking of Malayan genera of the tribe Cyrtandreæ, Mr. Ridley
says that their dry, dull-coloured, and inconspicuous corky fruits are
often devoured by animals. The seeds, on account of their roughened
surface, adhere to rocks and other surfaces and readily germinate.

FREYCINETIA (Pandanaceæ).

If there is any genus of tropical plants to which the student of
distribution can look for guidance in the region of the Pacific, it is
to Freycinetia as dealt with by Dr. Warburg in his monograph on the
order (Engler’s _Pflanzenreich_, iv. 9, 1900). Its characters and its
distribution are well defined; and here, if anywhere, we might be able
to work out the history of a genus. In the words of the German botanist,
it stands quite apart from Pandanus and Sararanga, the two other genera
of the order. When Hillebrand was preparing his work on the Hawaiian
flora, more than a quarter of a century ago, only about thirty species
were known. Warburg’s list, excluding doubtful forms, comprises sixty
species, and even this number the author surmises will be doubled in
future years. The later investigators, however, have not materially
extended the range of the genus; and the statement of the botanists of a
generation ago, that it extends from Ceylon through Malaya and Australia
to New Zealand, and is found on almost every elevated island of the
Pacific, can only be supplemented by extending its area to the Asiatic
mainland in Burma where a wide-ranging Malayan species exists.

It is, however, remarkable that no endemic species can be with certainty
accredited to the mainland of Asia either in Burma or in the Malay
peninsula where the genus also occurs. The Malayan region from Java to
the Philippines possesses quite three-fifths of the species, and it is
singular how few wide-ranging species there are. The Philippine Islands,
Borneo, Celebes, Sumatra, Java, New Guinea, &c., have all their own
species, the only wide-ranging plant being Freycinetia angustifolia,
which occupies the region from Burma to Java and Borneo. So also in the
Pacific, there is no widely distributed species, every group possessing
its own plant or plants, and there does not appear to be any Freycinetia
that is common to two groups. Thus, Hawaii and Tahiti each have their
own species. Rarotonga, according to Cheeseman, owns a peculiar but not
yet fully described form. Samoa has two and Tonga has one species.
Westward from Tonga and Samoa the numbers of species increase, Fiji
possessing five and New Caledonia four. Australia and New Zealand each
claim two species as their own.

Dr. Warburg, who has studied the genus in its home, remarks on page 43
that none of the species possess any means of dispersal enabling them to
cross an ocean; and he connects with this the fact that the genus is
only found (to use his own words) on islands like those of Samoa,
Tahiti, and Hawaii, that possess a “palæobiotic” nucleus
(_paläobiotischen Kern_) and not on islands like the Bonin Islands of
new formation (_auf Neubildungen_). This attitude towards the problem of
plant-distribution in the Pacific is backed by a great experience; but
it is one, of course, that is directly opposed to the line of argument
followed in these pages; and it is needless to say that it is not
encouraging to the student of plant-dispersal. Yet one could hardly look
upon the islands of the Tongan Group with their representative of the
endemic Freycinetias as of more ancient origin than the Bonin Islands
that have none; and plants that find their homes on the peaks and in the
forests of mountainous islands would rarely find a suitable station on
the low coral islands of the Pacific. It is, however, noteworthy that
Professor Schimper is inclined to include a species of Freycinetia as
amongst the strand-flora of the coral islands of the Java Sea (_Ind.
Mal. Strand-flora_, p. 134). With regard to the question of the means of
dispersal of Freycinetias, it will at once be shown that these plants
possess many opportunities for dispersal by birds.

Though in our own time dispersal by birds between the various Pacific
archipelagoes is often largely suspended, the inter-island dispersal in
each group is usually active through the agency of birds, now like the
plants they distribute confined to each group. Thus with Freycinetia we
find that, notwithstanding that each Pacific group is, as regards this
genus, isolated from the others, the separate islands, as in the case of
those of Hawaii, may possess a common species dispersed over the area.
The ripe fruit, which consists of a number of berries in a head or
spike, is juicy and pulpy, and contains in each berry a large number of
minute oblong or fusiform seeds, usually one or two millimetres long and
possessing thick toughish tests. Birds, indeed, are fond of pecking at
the ripe fruit-heads in Hawaii. Thus we learn from the _Aves
Hawaiienses_ of Wilson and Evans that a Grosbeak (Psittacirostra) and
the Hawaiian Crow (Corvus tropicus) feed principally on ripe Freycinetia
fruits, the seeds having been often found by Mr. Wilson in the stomach
of the former bird. No doubt these birds distribute the seeds over the
islands of the group. Mr. Perkins tells me that the Grosbeak is found
unmodified all over the group, and that it no doubt frequently gets
carried _nolens volens_ from one island to another. In his memoir on the
birds in the _Fauna Hawaiiensis_, he remarks that the essential food of
the “Ou,” the native name of this bird, is the fruiting inflorescence of
Freycinetias. The “Oo” (Acrulocercus) and the Hawaiian Crow above
mentioned, as he also observes, feed on these ripe red fruits. Like Mr.
Wilson, he sometimes found the Crow absolutely filled with this food to
the exclusion of all others (see Chapter XXXIII). Facts of a similar
kind came under my notice whilst in these islands. Thus on one occasion
I observed, on a leaf below a fruit-head that had been partly eaten by a
bird, a pellet half an inch long composed entirely of Freycinetia seeds
well soaked with the gastric juices and apparently only recently
disgorged. Sir W. Buller refers to different New Zealand birds, as the
Banded Rail (Rallus philippensis), the Kaka Parrot (Nestor
meridionalis), and the “Tui” (Prosthemadera), that live on the “sugary
flowering spadices” of Freycinetia Banksii. One can legitimately suppose
that they also attack the juicy berries. It is singular that as we learn
from Dr. Warburg (p. 17), Flying-Foxes (Pteropidæ) feed on the flowers
and top-leaves of many species of Freycinetia, and he considers that
they would aid in fertilisation by carrying about the pollen in the hair
of the head. Here again it would seem to us highly probable that whilst
brushing past a ripe fruit-head these bats might readily carry away in
their fur some of the minute seeds, which in the fresh berry are
“sticky” or adhesive.

Just as it was possible in the case of Coprosma in the South Pacific
(see page 296) to connect its distribution with the range of the Purple
Water-Hens (Porphyrio), so it may perhaps be legitimate to associate the
range of Freycinetia over Polynesia with the distribution of the
Honey-Eaters (Meliphagidæ) in the Pacific, a family sometimes possessing
peculiar genera as in New Zealand and Hawaii, and one in which the
species have usually a very confined range, being sometimes limited to a
single island (Newton in _Encycl. Brit._ xii. 139). To this family
belongs the New Zealand “Tui” above mentioned; and it may be remarked
that these birds as a rule feed on soft fruits, such as figs, and
bananas. It is to Acrulocercus, one of the Hawaiian genera of the
Meliphagidæ, that Mr. Perkins refers me, on my asking him to name some
of the fruit-eaters in that group.

These climbing shrubs, as Dr. Warburg observes, mostly frequent the
tropical forests up to 4,000 feet and over. Though their most familiar
habit is as tree-climbers in the forests, in localities where there are
no trees they adopt a trailing habit and cover mountain peaks and ridges
with a dense growth to the exclusion of almost all other plants. Many a
peak in the Pacific islands would be inaccessible if it were not for the
dense growth of these plants on their precipitous sides. It was owing to
the friendly aid of a tangled mass of Freycinetia stems that Lieutenant
Heming and myself were able to clamber to the summit of Fauro Island
(1,900 feet) in the Solomon Group, where I discovered a tree that under
the name of Sararanga forms the type of the third genus of the
Pandanaceæ.

Whilst describing their station, it will be of interest to also record
the altitudes at which these plants have been observed in the tropical
Pacific. Since they can be independent of trees and are as much at home
on treeless rocky peaks and mountain crests, the upper limit would
usually be determined by climatic conditions, abundance of rain and
great humidity being the chief requisites; but, as will be seen below,
this limit does not seem to be reached in the tropical islands of the
South Pacific except perhaps in Tahiti. In the Fijis the Freycinetias
ascend to the highest mountain peaks. Thus, three of the species
discovered here by Seemann were found at elevations of about 4,000 feet
on Voma Peak in Viti Levu and in the highlands of Taviuni. In Vanua
Levu, as I found, they cover the highest peaks 3,500 feet above the sea.
They are especially abundant on the lofty mountain ridges, and clothe
the higher slopes of the Mbatini Ridge which terminates in the highest
peak of the island. In no locality did I find them growing in such
densely tangled masses as on the long ridge-like crest that forms the
upper part of Mount Freeland, 2,740 feet above the sea. For more than an
hour in order to reach the summit I had to clamber along the crest of a
ridge covered with a dense growth several feet deep of these trailing
plants, without touching the ground beneath.

In Samoa, as we learn from Reinecke, Freycinetias are common on the
mountain ridges, climbing the trees and forming also a dense undergrowth
covering the ground and concealing the rocks. They occur at all levels
from 1,000 feet above the sea up to the highest region of Savaii, rather
over 5,000 feet in elevation. In Rarotonga, according to Mr. Cheeseman,
the Freycinetias are very abundant on the mountains, which reach a
height of 2,200 feet, the plants scrambling up the trunks of trees or
over rocks and frequently rendering the forest almost impenetrable. In
Tahiti, Nadeaud tells us, the Freycinetias often cover in an
inextricable network the sides of the valleys at elevations of 2,000 to
3,300 feet, extending in their vertical range from the lower levels of
the island to the highest inaccessible peaks which attain a maximum
height of about 7,300 feet.

These plants in the Hawaiian group are common in the lower woods as
Hillebrand informs us, that is to say, at elevations of 2,000 or 3,000
feet. During my descent from Mauna Kea through the Hamakua forests on
the north-east side I observed that the Freycinetias commenced at an
altitude of 3,900 feet, and that they attained their greatest
development between 3,200 and 2,000 feet. These plants ascended quite a
thousand feet higher on these mountain slopes than the Bird’s Nest Fern
(Asplenium Nidus), which reached an altitude of 2,800 feet. In the
forests on the west side of Mauna Loa they were abundant at altitudes of
3,500 to 4,000 feet and were not noticed above 4,500 feet. On the slopes
of Mount Eeka in West Maui they abounded between 3,500 and 4,400 feet.
In those localities where the forest descends to the sea, Freycinetias
occur at the coast, and on Oahu they are often found at elevations under
a thousand feet.

I have but few data showing the altitude obtained by Freycinetias in
other regions, as, for instance, in their most southerly habitat in New
Zealand, where they give a tropical luxuriance to the forests, or in
their chief home in Malaya. From Schimper’s observations
(_Plant-Geography_, p. 293) it would seem that they thrive in the Gedeh
forest of Java at elevations of about 5,000 feet. Except for the lower
levels, Warburg makes but few references to this subject in dealing with
the species. It appears to me that some very interesting results might
be obtained by comparing the vertical range of this genus in different
regions, as, for instance, in New Zealand and in Borneo or in Java. We
might get indications that since the age of Freycinetia began the
climate in tropical latitudes has been getting warmer, and that the
erstwhile plants of the lower levels are now as a result climbing the
mountain slopes. The student of distribution may find here a genus that
has been “cornered” not only in space and time, but as regards its
conditions of existence. Since it is obvious that during a gradual
increase of temperature it would ascend the mountains and during a
lowering of temperature it would descend to the plains, it follows that
in the mountains of an oceanic island it might be driven into the sea or
await extinction on a mountain-top. In the tropics also there would be
no escape during a gradual increase of temperature. Here again it would
make its last stand on the strand, and, forced to choose between Death
and Adaptation, the genus might select the latter alternative and
present us with a startling new form. In this sense Freycinetia seems to
offer itself as “fair game” for the speculative botanist, and at all
events he will be able to interrogate it as to the connection between
its existing range of altitude and the climatic conditions of the
earlier phases of its history.

The Freycinetias bear the same name over Polynesia, “ie-ie” in Hawaii,
“ie” and “ie-ie” in Tahiti and Samoa, which appear in their full form in
the Rarotongan and Maori “kie-kie.” The secret of the wide distribution
of the name lies in the circumstance that this is a mat-word over much
of Polynesia, as in Fiji, Tonga, Samoa, the Gilbert group, Tahiti, &c.,
Freycinetia leaves being often employed for making mats, as in Samoa and
New Zealand. The same word is applied in some groups to small species of
Pandanus that were also used in mat-making. Thus in Fiji “kie-kie” was
not only the name for a mat-dress, but also of Pandanus caricosus that
supplied the material. In the home of the Polynesians in Malaya and its
vicinity the same word for mat and Pandanus occur. Thus, “gerekere” in
the Motu dialect of New Guinea and “keker” or “kekel” in Amboyna are the
names of small species of Pandanus employed in mat-manufacture; whilst
“kihu” and “kiel” in Celebes are the words for the mats themselves.
Therefore in one form or another the word, originally applied to the
mats, but now often restricted to the plants from which the materials
were derived, ranges over the great region extending from Malaya to New
Zealand, Tahiti, and Hawaii, and, as I have shown in the table given in
my paper on Polynesian Plant-Names (_Journ. Victor. Inst._, London,
1896), it may be traced even to Further India, as in Annam, and to
North-East Australia. It thus covers the area to which the migrations of
the Polynesians of the Pacific have been confined, and it covers also
the area of the genus Freycinetia. There is something far more than mere
analogy between man and plants in their occupation of the Pacific
islands. The plants are Malayan and the Polynesians are from Malaya
also, whilst in both man and plants we experience the same difficulty in
explaining their dispersal over the ocean. Divesting his mind of all
previous conceptions, the ethnologist might profitably study _de novo_
the dispersion of man in the Pacific from the standpoint of
plant-dispersal (see Chapter XXVIII).

SAPINDUS AND PHYLLANTHUS.

Brief reference can alone be made to these two genera. Foremost comes
Sapindus, which is represented by two endemic species, one in Hawaii and
one in Fiji, and by another species, found in Tahiti, the Marquesas, and
Easter Island, which is identified by some botanists with the well-known
American “soap-tree,” S. saponaria. There are several difficulties
connected with the presence of this genus of the Old and New World in
the Pacific. Not the least of them is connected with the transport of
the large seeds of this genus, an inch in size, to the isolated Hawaiian
Group, where it is represented by a solitary endemic species in the
island of Oahu. The fleshy mesocarp of the fruits might attract birds;
but it is not easy to perceive how birds could carry such large seeds
over some 1,500 or 2,000 miles of ocean. Yet the same difficulty exists
with a few other genera, such as Osmanthus and Sideroxylon, that are
only represented in Hawaii by endemic species, genera which require the
agency of birds to explain their occurrence unless we wish to postulate
a continental connection for this group. (See under those genera in
Chapter XXVII.)

The large Euphorbiaceous genus Phyllanthus, spread universally over the
tropics and containing some 500 known species, clearly indicates by its
distribution in the Pacific islands that genera with dry fruits, such as
are typical of the order, are as widely distributed and just as much at
home in these islands as the genera with fleshy fruits, such as
Psychotria and Cyrtandra. The small trees and shrubs of Phyllanthus are
common in dry, open, partially wooded districts near the sea-border. The
genus attains its greatest development in this ocean in New Caledonia
and Fiji; and since the number of species diminishes the further we
penetrate the Pacific, it can be scarcely doubted that the genus has
entered this ocean from the west. In Fiji there are at least 20 species,
of which probably half are not recorded from elsewhere. In Samoa there
are seemingly but few peculiar species. In Hawaii there is only one
indigenous species, and that is endemic. The genus, however, has
developed a lesser centre of distribution in East Polynesia, there being
about a dozen species known from Tahiti and the Marquesas, of which half
are peculiar to one or other of those groups. From experiments made by
me in Fiji on the fruits and seeds of two species it was evident that
they possessed little or no capacity for dispersal by the currents. We
look, therefore, to the birds, and in this connection it is of interest
to note that this genus is included amongst those known to be dispersed
by birds in the Pacific, some of the fruits having been found in the
crops of fruit-pigeons shot by Prof. Moseley in the Admiralty Islands
(_Bot. Chall. Exped._, Introd. 46; iv. 308).

PRITCHARDIA (Palmaceæ).

This genus of Fan Palms supplies an instructive lesson for the student
of plant-distribution, more especially with reference to the loss of the
endemic reputation of a genus. Regarded by the earlier botanists who
visited the Pacific as identical with the familiar Asiatic Talipot Palm
(Corypha umbraculifera), the Fan Palms of this region, as represented in
Fiji and Hawaii, were subsequently placed by Seemann and Wendland in a
new genus restricted to Polynesia and named after a former British
Consul in Fiji. Since that time it has lost its reputation as a
peculiarly Pacific genus, since a species (Pritchardia filifera) has
been found lingering in a few valleys in Arizona, where it enjoys the
distinction of being the most northerly in station of all the world’s
palms (Linden in _Illustr. Hort._ vol. 24, 1876-77). It would thus
appear that the Pacific islands have derived this genus of palms from
the western part of North America, but the whole question is beset with
many difficulties, and not the least is that connected with the
confusion that seems to reign in several cases as regards the allocation
and identity of the species.

Six species are named in the _Index Kewensis_, viz.: Pritchardia
macrocarpa, restricted to Hawaii; P. martii and P. gaudichaudii, of the
Pacific islands; P. pacifica, assigned to Fiji; P. vuylstekeana, from
the Paumotus; and P. filifera, from the west side of North America.
Though it is sometimes difficult to reconcile this account of the
distribution of the genus in the Pacific with views held by other
botanists, it offers the safest basis for the future investigation of
the subject. It would be, however, necessary to remember that
Pritchardia gaudichaudii and P. martii are regarded by Hillebrand as
peculiar to the Hawaiian Islands, and that the exact locality of the
Paumotu species is not very definitely settled, if it depends on the
remarks made on this species in the _Gardeners’ Chronicle_ for 1883. No
mention is indeed made by Drake del Castillo of any Tahitian or
Paumotuan species.

Whilst in Hawaii and Fiji I was much interested in these palms, and the
following remarks are merely intended to be a contribution to the
subject. According to Seemann, Hemsley, Drake del Castillo, and Burkill,
Pritchardia pacifica, which often attains a height of thirty to
thirty-five feet, occurs in Fiji, Tonga, Samoa, and the Marquesas, but
it does not exist in Tahiti, and Cheeseman does not include it in the
Rarotongan flora. Except in the Tonga Group, where, according to Lister
as quoted by Hemsley, the palms form conspicuous objects along the
weather shore of the island of Eua, this species is rarely found in the
wild state in the South Pacific. This especially applies to Fiji, as Mr.
Horne also observes; and at most one is accustomed to see (to employ the
words of Dr. Seemann) one or two trees outside a village which are
reserved, as in many parts of Polynesia, for the use of the chiefs who
employ the leaves for fans and for other purposes. But even this reason
for preserving the palms scarcely now exists in Fiji, and at the time of
my sojourn in Vanua Levu (1897-99) the trees were rare enough to be
regarded as curiosities. In the Marquesas, according to Bennett (quoted
by Seemann), they grow in groves in the valleys of the interior. Dr.
Reinecke does not even include the species in the Samoan flora, but
mentions it with the Date-Palm (Phœnix dactylifera) as if it were
recently introduced. It was, however, found in that group by the United
States Exploring Expedition about 1840, and this is evidently the palm
referred to by Captain Cook as existing at his time in the Tongan Group.

The Hawaiian species of the palm appear to be three in number,
Pritchardia gaudichaudii and P. martii, both regarded by Hillebrand as
confined to the group, and P. macrocarpa of Linden, also endemic
(_Illustr. Hort._ vol. 26). The two first-named species are evidently on
the road to extinction in the wild state, and often find their last
refuge on rocky, almost inaccessible, inland cliffs. Pritchardia
gaudichaudii, about twenty feet in height, is found in the wild state,
as we learn from Hillebrand, on the islands of Molokai and Hawaii. It
was at one time frequently met with near native dwellings; but during my
sojourn in 1896-97 on the last-named island it was not at all frequent,
and as a rule only came under my notice occasionally in clumps of three
or four trees on the Kona and Puna coasts, as near Kiholo, Milolii, and
Kalapana. However, it was more frequent in the Waimanu district of
Kohala in the same island. Here I noticed it growing in clumps in
precipitous rocky situations at elevations ranging from 1,200 to 2,000
feet. The other palm mentioned by Hillebrand, P. martii, is only five or
six feet high, and is confined mostly to Oahu and Molokai.

The agency of man in introducing these interesting Fan-Palms into the
Hawaiian Islands seems out of the question, since they are home
productions in a specific sense and are doubtless ancient components of
the flora; and, of course, grave objections exist on ethnological
grounds, if this genus had originally its home in America. With
reference, however, to Pritchardia pacifica of the South Pacific, it is
not unlikely that man has aided in the distribution of a palm mainly
preserved by planting in and about the villages and set apart from time
immemorial for the use of the chiefs.

In this connection the aboriginal names are of some importance and may
be very briefly here referred to. The Fijian “Viu,” the “Piu” of Samoa,
Tonga, and Futuna, and the Tongan “Biu” are forms of the same name
applied to this palm all over West Polynesia; and I have shown in my
paper on Polynesian Plant-Names that in the form of “Firo” in the
Solomon Islands (Bougainville Straits) and of “Wiru” in Sundanese, one
of the Malayan languages, the same name is given to another genus of Fan
Palms, namely, Licuala. But since these West Polynesian names do not
always conform with the laws of consonantal interchange in this region,
they cannot all be considered as indigenous in the languages concerned.
If, for instance, “Viu” is an indigenous Fijian name, as no doubt it is,
since it follows the phonetic laws affecting the Malayan and Fijian
languages, “Piu” must be a foreign word in Samoa and Tonga, and “Biu”
must be another introduced Tongan name.... The Fijians have in “Sakiki”
(contracted into “Saii” in the Somosomo dialect) another name for this
palm. This is probably derived from “Kiekie,” a mat-word in different
forms in various Polynesian groups, and applied in many islands to the
plants that supply the materials for mat-making, such as Pandanus and
Freycinetia.

The Hawaiian generic name of “Loulu” for these palms appears to be quite
local; but it may possibly have a common origin with “Roro,” one of the
Fijian names of Cycas circinalis. It is pointed out by Hillebrand that
the Hawaiian name of the edible kernels of these palms, “Hawane” or
“Wahane,” occurs in the Marquesas as “Vahana” applied to the palm, a
comparison that is on linguistic grounds quite legitimate. “Vaake” is
another Marquesan name, which recalls “Vakoa,” the Malagasy word for
Pandanus.

When we compare the variety of the names of the Pritchardia fan-palms in
the Pacific Islands with the prevailing uniformity of the names of
cultivated plants transported by the aborigines in their migrations from
Malaya, such as the taro, the yam, the sugar-cane, the coco-nut, and the
Malay-apple, we perceive that the testimony of the names points to the
same conclusion as the botanical evidence, namely, that the ancestors of
the Hawaiians found these palms in the group at the time of its
occupation. In the South Pacific much uncertainty prevails. The
ancestors of the West Polynesian peoples evidently brought the word for
a fan-palm from their Malayan home; but it is doubtful if they found
Pritchardia already established in all the islands; and the apparent
home of the genus in America prevents us from attributing to a palm,
that is by some botanists regarded as confined to the Western Pacific, a
home in the neighbouring regions to the west. There is thus a lack of
agreement between the botanical and ethnological indications as regards
the original American origin of Pritchardia in the South Pacific.

There remain then the agencies of the currents and of birds. A singular
feature in the distribution of the Hawaiian species, Pritchardia
gaudichaudii, at once affords a clue as concerning the dispersal in the
North Pacific. Dr. Hillebrand remarks that this palm covers part of Bird
Island, a small volcanic rock forming an outlier of the Hawaiian group
about 400 miles north-east of Kauai. Here the agency of birds is
suggested, since it is scarcely likely, though, as shown below, not
impossible, that stranded fruits of the palm could have established
themselves in this fashion. Mr. Perkins has an interesting note on the
food of Ciridops anna, an Hawaiian bird, now nearly extinct, that feeds
principally on the blossoms and unripe fruits of the Loulu palms,
probably of this species. The drupes when fresh have a somewhat fleshy
mesocarp and are about 9/10 of an inch (22 mm.) across, and their
crustaceous inner shell would undoubtedly fit the seeds for dispersal by
frugivorous birds like pigeons. The fruits of the other two Hawaiian
species are considerably larger, that of P. macrocarpa being, according
to Linden, of the size of a nut of Juglans regia, that is, about 1-1/8
inch or 29 mm., whilst that of P. martii, as we learn from Hillebrand,
is from 1-1/2 to 2 inches or 37 to 50 mm. Allowing for the variation in
size of the fruits within the limits of the genus, there need be no more
difficulty in assuming that the original species had fruits that could
have been brought by birds, than in holding that the fruits of
Elæocarpus have been carried to Hawaii in the same fashion. The drupes
of Pritchardia pacifica are barely half an inch in diameter. They are
fitted by reason of their hard crustaceous endocarp for dispersal by
fruit-pigeons; and I may here add that these birds are known to
distribute the fruits of other palms, such as Kentia and Areca, in the
islands of the South Pacific (_Bot. Chall. Exped._ iv. 308, 312).

Both in Hawaii and in Fiji I experimented on the capacity of Pritchardia
drupes for dispersal by the currents. Those of the Hawaiian species, P.
gaudichaudii, have when well dried a light buoyant rather fibrous
mesocarp which enables them to float in the case of a good proportion of
the fruits for at least five weeks. I had no opportunity of testing the
buoyancy of the fruits of P. martii, another Hawaiian species; but,
judging from the existence in the coats of a fibrous layer as described
by Hillebrand, they ought to display some floating power. The fruits of
P. pacifica, the South Pacific species, lack the light buoyant covering
of the Hawaiian species above referred to, and display little or no
floating power even after drying for weeks. Looking at the results of
these experiments, it would seem that it is not impossible that Hawaii
received the genus through the agency of the currents; but it seems
scarcely probable, since it could only have been derived from America,
and the American species grows in the interior of the continent and not
near the sea-border. The possibility of course exists; but I am inclined
to attribute the presence of Pritchardia in Hawaii to bird-agency.

My position from the standpoint of dispersal with regard to Pritchardia
in the Pacific is this. The Hawaiian species I would consider as
American in origin. The Marquesan species, unless recently described,
still awaits detailed investigation. The West Polynesian species of Fiji
and Tonga, according to the principles of distribution prevailing in the
South Pacific, ought to hail from the west.

_Summary._

(1) Whilst the earliest age characterised by the Coniferæ was restricted
to the Western Pacific, and whilst the following age of the Compositæ
and Lobeliaceæ, mainly American in their affinities, was concerned with
the regions of Hawaii and Tahiti, we have now to discuss the Malayan era
during which the bulk of the plants were derived from the nearest
tropical regions of the Old World. Here we have to deal with the
low-level flora of Hawaii, that is to say, with the plants of the levels
below 4,000 or 5,000 feet, and with almost the entire floras of the
areas of Fiji-Samoa and of East Polynesia. The whole of the tropical
Pacific is here concerned, and not a portion of it, as in the two
preceding eras; and in our comparison we shall see that there are two,
and not as heretofore three, regions to be regarded—the Hawaiian in the
North Pacific, and the whole Polynesian area of the South Pacific
extending from Fiji to Tahiti.

(2) Here the frugivorous bird has been the principal agent in dispersing
the plants, quite two-thirds of the genera possessing drupes or berries
that would attract such birds.

(3) The genera representative of the first part of this era are those
which have only peculiar species in Hawaii, and are composed in the
South Pacific either entirely of peculiar species or sometimes of a
mixture of endemic and non-endemic species. It is an era of complete
isolation in Hawaii and often of a partial connection between the groups
of the southern region. Except to some extent in the South Pacific, the
dispersing agencies are now no longer active between the groups.

(4) Amongst the genera typical of this period are Pittosporum, Gardenia,
Psychotria, Cyrtandra, and Freycinetia.

(5) The two genera of the Rubiaceæ, Psychotria and Coprosma (the last
belonging to the mountain-flora), appear to be well suited for the
investigation of the effect on distribution of the geographical position
of the home of the genus, the first with 600 to 700 species distributed
over the tropics of the Old and New Worlds, the second with some sixty
species having its home in New Zealand.

(6) From the Pacific Cyrtandras we derive the lessons that the display
of great formative power in a genus may not be a peculiarity of an
insular flora; that the isolation of an oceanic archipelago does not
necessarily induce “endemism,” but merely intensifies it; and that the
production of new species within the limits of a genus like Cyrtandra
may be nearly as active on the mainland as in an island in mid-ocean.

(7) From the Freycinetias we learn that it may be possible to connect
the distribution of a genus of plants with that of a genus or a family
of birds. Just as in Chapter XXIV we endeavoured to connect Coprosma and
Porphyrio (the Purple Water-Hens), so we here suggest a connection, in
their range over the Pacific, between the Freycinetias and the
Meliphagidæ (the Honey-eaters), a connection that in the last case at
least belongs to the past.

(8) From the genus Phyllanthus we learn that genera with dry fruits may
be as widely distributed and may display the same formative power in the
Pacific as those with fleshy fruits that would seem much more likely to
be dispersed by birds. Here again we obtain an indirect indication that
species-making in these islands is not altogether dependent on
isolation.

(9) In the case of the genus Sapindus we are apparently compelled to
infer that its large seeds (in the present species an inch in size) have
been transported by birds to Hawaii. Yet in point of size the
difficulties here raised are no greater than those arising from the
existence of such genera as Sideroxylon and Elæocarpus in Hawaii, the
fruits of which are known to attract frugivorous birds.

CHAPTER XXVI

THE MALAYAN ERA OF THE NON-ENDEMIC GENERA OF FLOWERING PLANTS
(_continued_)

THE AGE OF WIDE DISPERSAL OVER THE TROPICAL PACIFIC (_continued_)

A LATER period in the era of the general dispersal of Malayan plants
over the Pacific is indicated by those genera that as a rule are never
entirely represented by endemic species in any archipelago. Hawaii now
comes into touch with the world outside, and all the groups possess some
connecting link. But the beginning of the effect of the isolating
influence is shown in the association in each principal archipelago of
peculiar species with those that occur in other groups.

We see here illustrated in all but the final stage that process by which
a solitary widely-ranging species, alone representing its genus, becomes
ultimately in each group the parent of a number of peculiar species. The
polymorphous, or extremely variable, species plays in this period the
all-important part. The earliest stage is exhibited by such genera as
Alphitonia, Dodonæa, Metrosideros, Pisonia, and Wikstrœmia, that possess
in the tropical Pacific a solitary widely-ranging species, varying
independently in every group and giving rise to forms that, in their
degree of differentiation, sometimes approach a specific value. Later
stages are shown when the polymorphous species, having done its work of
distributing the genus, settles down and “differentiates” in every
group; and this we see now illustrated in the genera Elæocarpus, Alyxia,
Peperomia, and others.

The bulk of the genera of this period, of which only a few can be
mentioned here, hail from the tropics of the Old World through Malaya.
Thus Alyxia, Elæocarpus, Morinda, and Wikstrœmia are Malayan; whilst
genera like Eugenia, Peperomia, and Pisonia, that occur in the Old and
New Worlds, can similarly be traced to the Asiatic side of the ocean by
the distribution of their species. Others again have their home in New
Zealand like Metrosideros, or in Australia, as with Dodonæa and Scævola.
None are exclusively American. Some of the genera, as Morinda and
Scævola, have littoral as well as inland species; but, as shown in
Chapter XIV, there is rarely anything to suggest a derivation of the
inland from the coast species, both being, from the standpoint of
dispersal, of independent origin.

About half of the plants have fleshy or sappy fruits (drupes and
berries) that would attract frugivorous birds, such as we find in
Xylosma, Elæocarpus, Eugenia, Scævola, Wikstrœmia, &c., whilst the
others have often dry capsular fruits, with minute seeds as in
Metrosideros, or with larger seeds as in Dodonæa. Some of them, like
Pisonia, have fruits that excrete a viscid material that causes them to
adhere firmly to plumage. Birds both granivorous and frugivorous have
been actively at work; and there are few difficulties relating to
dispersal connected with the genera, except with such as Gossypium and
Elæocarpus.

I will adopt the method employed in the preceding chapter of discussing
in detail from the standpoint of dispersal some of the genera that came
most frequently under my notice, or in which I am greatly interested,
and of dealing briefly with some of the rest. Those dealt with in other
connections will not be treated.

ELÆOCARPUS (Tiliaceæ).

This is a genus of trees containing, according to the _Index Kewensis_,
about 130 species, most of which are confined to tropical Asia,
including Malaya; but a fair number occur in the Pacific region, in
Australia, New Zealand, and the islands of the tropical Pacific, and the
genus is also found in Japan. It will thus be seen that Elæocarpus is
not only a continental but also a typical insular genus. It has reached
not only some of the most isolated island-groups of the Pacific, but it
is to be found also in the smaller islands of the Indian Ocean, there
being an endemic species in Mauritius. Amongst the Pacific Islands, a
region with which we are more immediately concerned, it has been
recorded from the Solomon Islands, New Caledonia, Fiji, Tonga, Samoa,
Rarotonga, and Hawaii. It is strange that the genus is not accredited to
Tahiti, but since it is represented in Rarotonga we may regard it as not
altogether absent from East Polynesia. Reinecke does not include it
amongst the Samoan plants, but Horne, in a short list of plants
collected in Upolu about 1878, mentions Elæocarpus græffei, a Fijian
species (_Year in Fiji_, p. 285).

New Caledonia represents the principal centre of the genus in the
tropical Pacific, thirteen species being accredited to it in the _Index
Kewensis_. Seemann found six species in Fiji, a number that does not
seem to have been added to by Horne. Of these one is found in Tonga and
Samoa, and of the rest perhaps most are peculiar; but one of them is
closely allied to a second peculiar Tongan species. Tonga possesses the
two species just alluded to, whilst Rarotonga and Hawaii have each a
peculiar species.

From an interesting comparison made by Mr. Burkill of some of the
Polynesian species, it would seem that Elæocarpus, if not actually
possessing a widely-spread polymorphous species in the tropical Pacific,
presents us with the next stage in the differentiation of the species.
Thus, he says in his paper on the flora of Vavau that an endemic Tongan
species, E. tonganus, is allied to three different species—E. græffei
from Fiji, E. floridanus from the Solomon Group, and E. glandulifer from
Ceylon—three species, he remarks, which are “so closely allied that it
is possible to regard them as insular subspecies.” It would thus appear
that some of the species of the Western Pacific are almost in touch with
Asiatic species. It would be of importance to determine whether some
affinity can be detected between the species of this part of the Pacific
and some of the widely-ranging species of Indo-Malaya, such as E.
ganitrus and E. oblongus. Mr. Burkill goes on to say that the solitary
Hawaiian and Rarotongan species are closely allied, an inference which
is of interest as indicating the route by which Hawaii received its
species. The genus, we may fairly infer, once possessed a widely-ranging
polymorphous or very variable Asiatic species in the tropical Pacific;
and we see it now in the next stage of specific differentiation in
various far-removed regions. In this connection Seemann significantly
remarks that all the Fijian species are evidently very local in the
group.

It will be appropriate here to refer briefly to the station and mode of
occurrence of the species. They occur most typically as forest-trees,
often of considerable height. In New Zealand, according to Hochstetter,
they form a feature in the temperate rain-forest; and, as we learn from
Kurz, they are similarly conspicuous in the tropical rain-forests of
Pegu. To this seeming indifference to the varying thermal conditions of
different latitudes we shall have subsequently to refer again. The tree
of the Hawaiian Group, as Hillebrand tells us, is common in the forests
of Oahu and Kauai, but is scarce in Maui and Hawaii, a singular
distribution that may be due to the inflorescence being “often
monstrously deformed by oviposition of some dipterous insect.” The
Rarotongan species, according to Cheeseman, is common throughout the
island from the sea-level to the tops of the hills. In Vanua Levu I
found that these trees preferred the crests of wooded mountain-ridges or
the partially vegetated mountain peaks. They came under my notice in the
forests of the island of Fauro, in the Solomon Group, associated with
other large trees of the genera Canarium and Calophyllum.

Much interest is attached to the mode of dispersal of this genus, since
in some species the size of the drupes and of the included “stone” is so
great that, judged by those species only, it might be deemed impossible
to attribute the existence of the genus in isolated oceanic groups to
the agency of frugivorous birds. We are, however, compelled to appeal to
the bird, since, as my experiments in Fiji indicate, the genus has
little or no capacity for dispersal by currents, the “stone” when
containing a seed always sinking, whilst the entire fruit either sinks
at once or floats heavily for a few days.

The degree of fleshiness of the drupes of Elæocarpus varies in different
species, being sometimes slight and at other times pronounced, but,
speaking generally, they would be expected to attract frugivorous birds.
The colour of the fruits of some species is dark and purplish, whilst in
others it is a bright blue. In the last case the fruits are very
conspicuous and sappy. A Solomon Island species collected by me and a
Malayan species observed by Ridley had bright blue fruits, and Cheeseman
refers to the Rarotongan species as possessing fruits of this hue. Their
colour, therefore, would often aid in attracting birds, and we are not
surprised to learn that they form a favourite food with fruit-pigeons,
parrots, and other frugivorous birds in different regions. Amongst the
fruits found by Professor Moseley in the crops of fruit-pigeons in the
Admiralty Islands were those of Elæocarpus; whilst in the Solomon
Islands I noticed that the blue fruits of the “Toa,” a species of the
genus, were a favourite food of the same birds (_Bot. Chall. Exped_.,
iv. 307, 308; Guppy’s _Solomon Islands_, 293, 295). We learn also from
Hochstetter and from Sir W. Buller that the drupes of the “Hinau”
(Elæocarpus) form a favourite food of the parrots and fruit-pigeons of
New Zealand (Hochstetter’s _New Zealand_; Buller’s _Birds of New
Zealand_).

The question of size acquires considerable importance when we come to
consider the transport of the seeds of the genus to a group of islands
lying, like Hawaii, in the middle of the Pacific Ocean. The protection
of the seed is also another important matter. There can, however, be no
doubt that the hard woody or often osseous “stone” sufficiently protects
the seed. With regard to size, if we were to judge from the dimensions
of the fruits of some of the Fijian species, where, as I found, the
“stone” measures from 3 to 5 centimetres (1-1/4 to 2 inches) in length,
we might be led to form a very erroneous opinion of the capacity of the
genus for conveyance through the agency of frugivorous birds to Hawaii.
But when we turn to the Hawaiian species we find the difficulty much
diminished, though still serious, the fruits being smaller and
possessing a “stone” 2-1/2 centimetres or about an inch long. In other
regions, however, the genus may possess fruits yet smaller in size. The
Tongan endemic species, as described by Burkill, has fruits 1·7 cm. or
7/10 of an inch in length; and closely similar dimensions are given by
Kirk for a New Zealand species. In both these cases the “stone” would
not be more than half an inch or 1·2 cm. in length, and this would also
apply to the Solomon Island species above mentioned. In another New
Zealand species, where the drupe is only half an inch, the “stone” would
be still smaller. It is thus evident that the fruits of different
species vary greatly in size in different regions, and that there is no
difficulty in assuming that a small-fruited species could be dispersed
over the Pacific by frugivorous birds, and carried either to Hawaii or
New Zealand.

It might be an interesting point to determine to what extent a species
in an oceanic island could effect its own isolation by developing a
“stone” too large and too heavy to be transported across an ocean by
birds, such as seems to have happened with some Fijian species. But a
similar curious question is raised by the deterioration of a drupe in
its capacity for dispersal by frugivorous birds, when, as in the case of
the Hawaiian species of Elæocarpus, the drupes become dry and almost
sapless. As remarked in Note 68, this same feature is to be noticed in
the fruits of some of the Hawaiian endemic genera. This, of course,
would be quite in accord with what we should expect from the standpoint
of dispersal.

I will conclude these remarks on Elæocarpus with a reference to the
similarity of its distribution with that of Freycinetia. Both genera are
at home in the temperate rain-forests of New Zealand and in the tropical
rain-forests of the Pacific islands and of Malaya. Their capacities for
dispersal are so different and so unequal, the dispersal of Freycinetia
being seemingly so much more readily effected, that we can only suppose
that time has long since discounted any special advantage one genus
possesses over the other as regards distribution.

DODONÆA (Sapindaceæ).

This genus of small trees and shrubs includes between fifty and sixty
known species, of which about forty are confined to Australia; but a few
species are found over the tropical and subtropical regions of the
world, extending sometimes into temperate latitudes. There are, it
seems, only three species known from the oceanic groups of the tropical
Pacific: one, the cosmopolitan Dodonæa viscosa, that occurs in every
island of volcanic formation; and two others associated with it in the
Hawaiian Group, to which they are restricted. We have thus repeated in
this genus what is true of several other genera in Hawaii, such as
Metrosideros and Wikstrœmia, namely, the occurrence in that group of a
widely-ranging species accompanied by other species peculiar to those
islands. In the case of Dodonæa in Hawaii we should not expect to find
it very difficult to connect the endemic species with the widely-ranging
D. viscosa, which is a very variable species. The extreme forms in
different parts of the world are so different in character that Bentham
viewed this species as probably including the whole of the
extra-Australian species, excepting perhaps the Hawaiian endemic species
and one or two South African and Mexican plants (_Bot. Chall. Exped._,
iii. 136).

Of the two Hawaiian peculiar species, one, Dodonæa eriocarpa, is a
mountain shrub found in most of the large islands and occurring
sometimes at elevations of 6,000 to 8,000 feet. The other species, D.
stenoptera, is, according to Hillebrand, a very distinct species found
only on Molokai. Bentham was only acquainted with the first-named, and
his hesitation to include it as one of the innumerable forms of the
widely-ranging D. viscosa is very suggestive. However, whether or not
one or both of these peculiar forms are connected in their origin with
this species, it is certain that the genus has been established for ages
in Hawaii; and from D. viscosa we can learn how a species of the genus
can cross an ocean, and also how from a widely-ranging species
exhibiting extreme variability species peculiar to a group of islands
could have been derived.

The great variability of Dodonæa viscosa is associated with great
adaptability to different stations. Thus, as Mr. Hemsley tersely puts
it, it is one of those plants that thrive on the sea-coast as well as
inland, and in almost any soil or situation—provided, it may be added,
that the station is well exposed to the sun. Although Mr. Ridley
characterises it as a regular sea-shore plant in the Malay peninsula,
and although Prof. Schimper places it in the Indo-Malayan strand-flora,
it is as an inland plant that it is most characteristic of the Pacific
islands; and the key to its powers of adaptation to different stations
is to be found in its xerophilous habit. It is essentially a plant of
sunny places, and is equally at home on the parched inland plain, in the
open wood, on the sandy beach, on an old lava-field, or on rocky
declivities. It is not a plant of the rain-forest, preferring dryness to
humidity and sunshine to shade.

The following remarks on the mode of dispersal of the wide-ranging
Dodonæa viscosa will serve to roughly indicate the capacity of the genus
for distribution. It is a subject, however, that requires further
detailed investigation. The light, inflated, winged capsules of this
species, about an inch across, could be blown for long distances along
the ground and carried for short distances in the air by strong winds,
but, as is also remarked by Prof. Schimper (_Ind. Mal. Strand-flora_, p.
157), they are much too large to be transported by winds across a broad
tract of sea. The currents, however, may have aided in the dispersal of
the species in the case of island-groups 500 or 600 miles apart.
Although the membranous capsules before dehiscing would be unable to
withstand the “rough-and-tumble” of ocean-transport for more than a few
days, the seeds possess some floating powers of a purely accidental
nature due to the imperfect filling up of the seed-cavity in some of the
seeds. In an experiment made in Hawaii I found that only half the seeds
floated in sea-water. Prof. Schimper, in an experiment conducted in
Germany with seeds that must have been well dried by keeping, found that
they floated for from ten to sixty days. This limited capacity for
flotation might possibly allow the species to reach Tahiti by easy
stages from Fiji; but it is not sufficient to explain its occurrence in
the more isolated Hawaiian Group. The fruits and seeds of this plant
never, however, came under my notice in the floating or stranded
seed-drift of Fiji; and I am not inclined, for this and the reasons
above mentioned, to consider that the currents have been very effective
agents in dispersing this plant over the Pacific islands.

Hillebrand endeavoured to account for the wide distribution of Dodonæa
viscosa by “the glutinous capsules which would easily adhere to the
plumage of birds.” It may be here remarked that in the dried state
specimens of the plant have a varnished appearance as respecting the
leaves, branchlets, and capsules. In the living condition this is
represented by a glutinous or viscid condition of the surface of these
portions of the plant, rendering them adhesive to the touch. I found,
however, that only the immature capsules are markedly “sticky,” and that
in any case the adhesive power was quite insufficient to allow of
adherence for any length of time of fruits of this size to a bird’s
feathers. Mr. Ridley, who allows much latitude to birds in matters of
dispersal, remarks that the stickiness only appears when the specimen is
dry (_Trans. Linn. Soc. Bot._, 1888-94, p. 289). It is, nevertheless,
likely that the crustaceous seeds, which do not exceed 1/5 of an inch (5
mm.) in size, when swallowed by a bird granivorous in its diet, might be
voided unharmed, and the dispersal of the species assured. It is in this
fashion, I imagine, that the plant reached distant groups like Tahiti
and Hawaii.

There is, of course, the possibility that man has in past times aided in
the distribution of Dodonæa viscosa over the warmer regions of the
globe. But such an agency seems largely discounted in the case of an
isolated archipelago like Hawaii by the occurrence of endemic species.
Nor does the usual station in the Pacific islands support the view that
it was introduced by the aborigines. According to Hillebrand, it
possesses a variety (var. spathulata) in Hawaii which seems also to
occur in Tahiti and New Zealand. Nadeaud observes that in Tahiti it
grows as a bush on dry crests, and as a small tree, ten feet in height,
in the mountains.

Nor do the aboriginal names of Dodonæa viscosa point in the direction of
man’s agency. It possesses a different name in every group, and is
evidently not a plant with which the ancestors of the Polynesians were
familiar in the home of the race. Thus it is named “aalii” in Hawaii,
“apiri” in Tahiti, “ake” in Rarotonga, “lala vao” in Samoa, and I may
add “usi” or, as Seemann writes it, “wase” in Fiji.

Looking at these various facts, I am not inclined to exclude altogether
any one of the three agencies above discussed; but I should imagine
that, placed in their order of effectiveness, we should have first
birds, then the currents, and lastly man.

METROSIDEROS (Myrtaceæ)

Whilst this genus of trees and shrubs has its home in New Zealand and
Australia, there is an extremely variable Polynesian species,
Metrosideros polymorpha, ranging over all the volcanic groups of the
tropical Pacific, from Fiji to Pitcairn Island and from Hawaii to the
Kermadec group, but seemingly only in the Hawaiian group associated with
endemic species. According to the _Index Kewensis_ the genus comprises
about forty known species, of which two-thirds are confined to New
Zealand and Australia in equal proportions; whilst, among the rest, six
species belong to New Caledonia, two to Hawaii, and three to Malaya, and
there are solitary species in Chile, Madagascar, and South Africa.

I will attack the problem connected with the distribution of the genus
through the widely-ranging Polynesian species, Metrosideros polymorpha.
“This genus,” wrote Dr. Seemann, “is in a fair way of becoming in
Polynesia what Rubus is in Europe. It is very much given to variation,
and it is very difficult to find out the limits of the different
species.” In making these remarks he had this species in view, and his
adoption of Gaudichaud’s specific name of “polymorpha” to cover almost
all the Polynesian forms has been generally followed. Although so widely
distributed over the Pacific, it is in the Hawaiian Islands that this
tree attains its greatest development, growing gregariously and often
forming almost exclusively entire forests; and it is here that it
displays the greatest variation. But it was remarked by Seemann, and
this was confirmed by Hillebrand, that almost all the Hawaiian forms
occur in the Society or Tahitian Islands.

In connection with the great variability of Metrosideros polymorpha must
be considered its variety of stations and its great range in altitude.
Hillebrand describes seven Hawaiian forms of this species, and their
various stations and characters are well illustrated in his
descriptions. Thus, whilst the trees may attain a height of forty feet
in the forests, in elevated exposed situations they may be small and
gnarled or low and shrubby; whilst in the bogs and swamps of the high
levels of Maui and Kauai the plant grows as a prostrate shrub. It is not
at all unlikely that the two peculiar Hawaiian species of the genus had
a common origin from a widely-ranging species, which, if not the present
M. polymorpha, was its immediate ancestor. One of them was, indeed,
included by Dr. Seemann within the wide limits of this species, and the
other was accepted with a doubt.

To illustrate the great vertical range in the Hawaiian Group of
Metrosideros polymorpha, I will take it as I found it in the island of
Hawaii. Here it ranges from the coast up to about 8,000 feet above the
sea. But it is in the middle forest-zone at elevations of 2,000 to 4,000
feet, where it is often associated with the Koa and Olapa Trees (Acacia
koa and Cheirodendron Gaudichaudii), that it is most at home and attains
its greatest size. Higher up at heights of 5,000 to 7,000 feet in the
more open forests it is still in the company of the trees just named
together with Sophora chrysophylla and Myoporum sandwicense. At 8,000
feet it becomes very stunted and is accompanied usually by bushes of
Cyathodes and other plants of similar bushy growth. In the lower parts
of its range, from 2,000 down to 1,000 feet, it forms forests with the
Kukui Tree (Aleurites moluccana), mingled also with smaller trees such
as the Hawaiian Olive (Osmanthus), and the Kopiko (Straussia). Below
1,000 feet, and wherever bold promontories reach the coast and the
inland forest descends to the sea, we find it associated with such trees
and shrubs as the Lama (Maba sandwicensis) and different Akeas
(Wikstrœmia). On the partially vegetated surfaces of old lava-flows near
the coast it grows beside bushes of the Ulei (Osteomeles
anthyllidifolia) and of Cyathodes.

Compared with its behaviour in Hawaii, Metrosideros polymorpha takes a
relatively unimportant part in the vegetation of Fiji. As Horne
observes, the trees are most common in the dry parts of the two largest
islands and grow in the poorest soil. I found them in Vanua Levu usually
in open exposed situations, generally in the dry “talasinga” plains on
the north side of the island, where they were associated with Acacia
Richii, Dodonæa viscosa, and Casuarinas; and sometimes they occurred in
a shrubby form on the rocky peaks of the highest mountains. In Rarotonga
also, as we learn from Cheeseman, it is on the tops of the rocky peaks
and along the crests of the ridges that this species, which is abundant
in the island, is frequently found.

I may here allude to the curious fact observed by me on the upper open
wooded slopes of Mauna Kea at elevations of 6,000 to 7,000 feet, and
therefore on the outskirts of the true forest-zone. Here the Ohia Tree,
as the Hawaiians name Metrosideros polymorpha, often grows in close
association with the Olapa Tree (Cheirodendron Gaudichaudii). In one
locality, for instance, a large Olapa was growing in the fork of an Ohia
at about eight feet from the ground, and sending down roots on either
side. Sometimes the trunks of the Olapa and the Ohia were to be seen
growing in such close contact as to look like one tree. In one such case
a young tree, four feet high, of Myoporum sandwicense was growing in a
fork of the Ohia, whilst in a fork of the Olapa a plant of Vaccinium
penduliflorum, three or four feet in height, had established itself.
This remarkable instance of epiphytic growth also proved to be quite a
revelation with regard to the dispersal of seeds in this island. Amongst
these four associated plants, which include three trees and one shrub,
all except the Ohia, which was probably the original tree, have fruits
that would attract frugivorous birds; and in succession these birds had
first dropped a pyrene of the Olapa in the fork of the Ohia, and
afterwards the seeds of Myoporum again on the Ohia, whilst finally the
Vaccinium seeds were dropped into the fork of the Olapa after it had
developed into a tree.

The mode of dispersal of the seeds of Metrosideros polymorpha now
invites our attention. Since the fruits are dry, dehiscent capsules
possessing minute fusiform seeds, we are not able to appeal directly to
the agency of frugivorous birds to explain the wide dispersal of this
species. The seeds are light in weight and remind one a little of those
of the succulent fruits of Freycinetia. For purposes of dispersal,
however, they must be placed in the same category with other plants with
dry, dehiscent fruits and small seeds, such as the Vota (Geissois
ternata) of Fiji, a tree that in those islands grows in similar
stations. On a later page I have suggested that the seeds of the Vota
are dispersed by large bats that visit the trees for the sake of the
honey in the red flowers. With Metrosideros polymorpha birds act
probably in the same way. We are, in fact, informed by Mr. Perkins that
the nectar-feeding birds of the Hawaiian Drepanids now obtain their main
supply of this food from the blossoms of this tree. If bats or birds
visit the large red flowers of Metrosideros polymorpha for the same
purpose, it is not difficult to imagine that they might carry away in
their fur or in their plumage some of the small seeds shaken out of old
dehiscent capsules. In this connection we may note that the Kaka Parrot
(Nestor meridionalis) of New Zealand is said to feed largely on the
scarlet blossoms and nectar of Metrosideros robusta (Evans’ _Birds_, p.
374).

The seeds of Metrosideros polymorpha might no doubt be carried by winds
from one mountain-top to another and across narrow straits, but only
whilst adherent to a bat or a bird could they be carried across a wide
tract of ocean. Speaking of the genera Metrosideros and Lobelia in
connection with their occurrence in the Kermadec Islands, Sir J. Hooker
long ago referred to their minute seeds as not adapted for transport
across oceans unless their minuteness and number fitted them for it
(_Journ. Linn. Soc._, i. 127). The point that is raised here for these
genera in the Kermadec Group can be raised for the same two genera in
Hawaii and for a multitude of other small-seeded genera in those
islands.

ALYXIA (Apocynaceæ).

This genus of climbing or straggling shrubs tells its own story of the
widely dispersed Indo-Malayan genera in the Pacific islands. Containing
about forty known species, it is distributed over the tropical regions
from Madagascar and the Mascarene Islands eastward to the Paumotu Group
and Pitcairn Island in mid-Pacific, and has its focus in the area
comprised by Malaya, Australia, and New Caledonia. In the _Index
Kewensis_ about eight species are assigned to New Caledonia, seven to
Australia, and seven to Malaya. One species, Alyxia stellata, ranges
over nearly the whole of the area of the genus from tropical Asia,
through Malaya, across the South Pacific to Tahiti. It will be for the
future investigator to determine how far the present distribution of the
genus can be connected with one or two widely-ranging polymorphous
species. The data at my disposal seem to show that in the open Pacific,
at all events, the history of the genus has gone a step beyond this
stage.

Of the seven or eight species recorded from the Pacific islands east of
New Caledonia, only two or three seem to be now recognised as restricted
to particular groups, namely, one in Hawaii (Schumann), one in Fiji, and
one in Rarotonga. The other species indirectly connect together all the
groups, although no single species occurs over the whole region. Thus
the Hawaiian species, Alyxia olivæformis (Gaud.) has in recent years
been found in Upolu, in the Samoan Group, by Dr. Reinecke, an
exceedingly interesting though unusual specific link between these two
archipelagoes. Two species, A. stellata and A. scandens, range over the
South Pacific from Fiji to Tahiti, the last-named also occurring in the
Paumotu or Low Archipelago; whilst Rarotonga possesses a form closely
allied to the first-named, and to it Cheeseman has given specific rank.
Another species, A. bracteolosa, links together the contiguous Fijian,
Tongan, and Samoan groups. This distribution is what we should have
expected if one or two polymorphous species had originally ranged over
the Pacific and were advancing towards that stage of differentiation
when each group possesses its own peculiar species. (It may be here
remarked that an undetermined species of Alyxia is accredited by Maiden
to Pitcairn Island, which indicates that the genus has extended east in
the Pacific almost as far as the extreme limit of the Polynesian
region.—_Australas. Assoc. Reports_, Melb., 1901, viii.)

All visitors to these islands that are interested in their floras will
be familiar with the Alyxias; and there are few of their plants that the
natives take more pleasure in pointing out to white men. They are
readily recognised on account of their black moniliform drupes and their
milky sap. All over Polynesia, whether in Hawaii, Tahiti, Samoa, or
Fiji, the aborigines value the plants on account of the delicate
fragrance of their foliage and bark. These materials they use for
personal decoration and in making wreaths, stripping off the bark of the
young branches with their teeth in the same fashion in Fiji and Hawaii
and probably in all the Pacific islands. Throughout Polynesia, excluding
Fiji, they bear the same name, which takes the form of “maile” in Hawaii
and Samoa, and of “maire” in Tahiti and Rarotonga—a name which the
Maoris, remembering the Alyxias of their tropical home in the South
Pacific, have applied to New Zealand species of Olea and Eugenia. The
Fijian generic name for Alyxia is “vono.”

A word may be said about the station of these plants in the Pacific
islands. In Hawaii they occur in the middle and lower forests, and
usually between 2,000 and 4,000 feet in elevation. In Tahiti they
frequent the crests and precipitous rocky slopes of the mountains at
elevations of from 3,000 to over 6,000 feet. The Rarotongan species
often forms extensive thickets in rocky localities on the hills. In
Samoa they are found usually in the mountain forests. In Fiji they grow
on the outskirts of the virgin forests and on rocky sparingly vegetated
mountain peaks. I found them often in Vanua Levu growing amongst the
open vegetation on the summits of isolated mountains at elevations of
2,000 to 2,500 feet, where they were associated with other plants like
Elæocarpus, Pleiosmilax, and Scævola, possessing similar fleshy fruits
likely to be dispersed by frugivorous birds.

The Alyxias indeed seem well suited for dispersal by birds. The black
fleshy drupes would readily attract them; and the solitary seed
protected by a very tough horny albumen might be ejected unharmed in
their droppings.

--------------

It would be possible to enter into similar detail with several other
genera of this period; but here I can only direct attention to their
principal indications, permitting myself a little more license when
discussing the means of dispersal.

ALPHITONIA (Rhamnaceæ).—Amongst other genera with polymorphous species
closely following the lines taken by Metrosideros in the Pacific is
Alphitonia, a small Malayan and Polynesian genus of tall trees,
containing at most three or four species, one of which (A. excelsa) has
almost the range of the genus and is found in most of the Pacific
archipelagoes. So variable is this widely-ranging tree that Bentham
suggested that there was only one species in the genus (_Bot. Chall.
Exped._, iii. 133), a suggestion especially interesting in connection
with the _rôle_ taken by polymorphous species in the Pacific. As bearing
on the mode of dispersal of this species, it may be observed that my
Fijian experiments show that the fruits are not fit for transport by
currents. With the mature drupe the outer coverings become pulverulent,
and the fruit breaks down, freeing the pyrenes which do not float; nor
have the seeds any buoyancy. Although the dry drupes would seem
unattractive to birds, it is to birds we must look for the dispersal of
the genus.

PISONIA (Nyctagineæ).—Like Dodonæa, Metrosideros, and Alphitonia, the
cosmopolitan genus Pisonia possesses a polymorphous species that
displays its variation in every Pacific group and occupies a
considerable number of stations. The earlier botanists in the Pacific
differed much as to the species of this region, and this led Mr. Hemsley
to observe in his paper on the Tongan flora that it is difficult to
understand the various Polynesian and Australian species except on the
assumption that there is one very variable species. Recognising this
difficulty, Drake del Castillo deals somewhat summarily with nearly all
these forms, uniting them under one comprehensive species, P.
umbellifera (Seem.), thus constituting “une espèce très-polymorphe” that
ranges (generally in maritime districts) over tropical Asia and the
islands of the Indian and Pacific Oceans, extending to North-East
Australia and to New Zealand. On account of the unusual capacity for
dispersal possessed by this species—a subject to be immediately
discussed—the tendency to specific differentiation has been kept in
check, though the process has gone farther in some groups than in
others, as in the case of Hawaii, where Hillebrand’s endemic species
has, however, been included by Drake del Castillo in his polymorphous
species, P. umbellifera.

The fruits of this genus possess no capacity for dispersal by currents.
They never came under my notice either in floating or stranded
seed-drift, and have little or no buoyancy. Prof. Schimper,
experimenting on the well-dried fruits of Pisonia aculeata, a seaside
shrub common in America and in the Old World, and destined probably to
be brought by the systematist into touch with the polymorphous P.
umbellifera, found that they sank in a day or two (_Ind. Mal.
Strand-flora_, p. 156). Dismissing the agency of the current, he looked
to that of the bird for the explanation of the dispersal. The
probability of the effectiveness of this last-named agency has long been
surmised. It attracted the notice of Darwin and especially invited the
attention of another student of plant-dispersal, Dr. H. O. Forbes. The
long, narrow, often fusiform fruits are invested by a somewhat
coriaceous perigone and range from less than an inch to three inches in
length (2-7·5 cm.). They excrete a very viscid fluid often in quantity,
and sometimes also possess glandular spines. The Hawaiians, according to
Hillebrand, used this material as bird-lime for catching birds, and the
fruits, he says, will stick fast to the paper in the herbarium for
years. In that group I often found the fruit adhering firmly to my
clothes. Writing of these trees on Keeling Atoll, Forbes observes that
their sticky fruits are often such a pest to birds roosting in their
branches that they have proved fatal to herons and boobies by collecting
in their plumage. “It is easy to perceive,” he remarks, “how widely this
tree might be disseminated by the birds that roost on it” (_The Eastern
Archipelago_, p. 30). In New Zealand, as we learn from Kirk, the viscid
fruits of Pisonia brunoniana attract small birds which become firmly
caught and die miserably. A cat has been known to wait under a tree
watching its opportunity of preying on the entangled birds. Sir W.
Buller states that the New Zealand fruit-pigeon feeds at times on the
green fruits of P. umbellifera; and we can infer that it occasionally
carries off some of the riper fruits in its feathers.

WIKSTRŒMIA (Thymelæaceæ).—This is a small genus of shrubs and small
trees, with red or yellowish drupes fitted for dispersal by frugivorous
birds, that is confined mainly to tropical Asia, Australia, and
Polynesia. Following Seemann and Drake del Castillo, we may say, that
like several other genera of this period, this genus possesses in the
tropical Pacific a widely-ranging species, W. indica, that occurs in
Hawaii, the Marquesas, Tahiti, Samoa, and Fiji, growing amongst the
vegetation immediately behind the beaches and in the plains and open
wooded districts inland. In Hawaii it is associated with half a dozen
peculiar species, and in Tonga there is also an endemic species. The
widely-ranging species has its home in the Indian Archipelago and in the
Asiatic mainland, and occurs also in Australia. According to Gray, the
American botanist, it is represented by a different variety in almost
every group in the tropical Pacific, and it presents us therefore with
another example of a polymorphous species which links Polynesia directly
with Malaya. As bearing on the dispersal of the genus by birds, it may
be added that Mr. Perkins in the _Fauna Hawaiiensis_ speaks of some of
the Drepanids and of a species of Phaeornis as feeding at times on the
fruits of these plants.

PEPEROMIA (Piperaceæ).—All observers of tropical plant-life will be
familiar with this genus of low herbs growing on tree-trunks, on the
soil, on rocks, and on stonewalls, and comprising about 500 known
species distributed over the warmer regions of the globe and sometimes
extending into cooler latitudes. In Polynesia it attains its greatest
development in Hawaii, where Hillebrand enumerates about twenty species,
of which, after excluding doubtful forms, at least a third must be
endemic. Tahiti, Samoa, and Fiji are each known to possess three or four
species, of which one is usually restricted to the group. Two species,
P. reflexa and P. leptostachya, link together nearly all the groups of
the tropical Pacific, including Hawaii, the first cosmopolitan, and the
second hailing from North-East Australia and indicating that the genus
has entered Polynesia from the west.... These plants possess spikes of
small berries containing a single seed, and are evidently, like other
Piperaceæ, dispersed by frugivorous birds. It is to be noted that the
presence of a West Indian and Mexican species in the Bermudian caves is
attributed by Mr. Hemsley to frugivorous birds (_Bot. Chall. Exped._,
Introd. 49, i. 62). In Vanua Levu they occur on the bare rocky peaks of
some of the mountains under such conditions that the seeds could only
have been brought by birds. Thus, on the bare surface of a large block
of tuff forming the highest peak of Koro-Mbasanga, 2,500 feet above the
sea, I found only two plants, Oxalis corniculata and a species of
Peperomia.

EUGENIA (Myrtaceæ).—This is a very extensive genus split up into
different subgenera, and comprising some 600 or 700 known species
scattered over the warm regions of the globe. Their fleshy, usually red,
berries contain as a rule one or two large seeds, and attract birds and
animals of all descriptions. The feature most interesting to us is the
dispersal of the genus over the Pacific islands eastward to the Low
Archipelago and northward to Hawaii. The track by which it has entered
the Pacific from the west is indicated in the distribution of the
species. The genus is only well represented in the Western Pacific,
whilst eastward and northward of Samoa and Tonga the distribution is
fitful and irregular, it being evident that the extension beyond these
two groups has been accomplished with difficulty.

There are at least twenty-five species in Fiji, of which perhaps half
would be peculiar; in Tonga eight species, of which two may be endemic;
in Samoa thirteen species, of which four are peculiar; in Rarotonga
none; in Tahiti a single non-endemic species; and in Hawaii two species,
of which one is peculiar. Only truly indigenous species are here
recorded, and Eugenia malaccensis, which has accompanied the aborigines
in their migrations, is not included. A solitary species, E. rariflora,
connects together all the principal archipelagoes from Fiji to Tahiti
and the Gambier Islands, and northward to Hawaii. Nine species are known
to be common to the region in which lie the three groups of Fiji, Tonga,
and Samoa; and since some of these species occur in the groups further
west they may be regarded as keeping up the connection with the original
home of their ancestors in the Malayan region.

Looking at these facts of distribution of the genus Eugenia in the open
Pacific, it is evident that whatever dispersal of the genus is now in
progress in this ocean is mainly confined to an interchange between the
groups of Fiji, Tonga, and Samoa in the Western Pacific, and doubtless
between the islands further west of these groups. The smaller islands
lying between and around these three groups participate in the
distribution of the species common to all. Thus Wallis Island, according
to Drake del Castillo, possesses two of these species. Over the rest of
the ocean the dispersal of the genus seems to be no longer effective,
since Eugenia rariflora, which links together Fiji, Tahiti, and Hawaii,
shows signs of differentiation in nearly every group. In Hawaii, where
it is very rare and is only recorded from two of the islands, it has
developed a small-leaved variety. In Tahiti it displays the same
variation; and Seemann observes that there are differences between the
Tahitian and Fijian species which may be almost specific in value. It
would also appear that both in Hawaii and Tahiti the fruits have become
less attractive to birds, being described as “dryish” and “dry,” which
is, as Dr. Seemann remarks, certainly not true of the Fijian plant.

In Fiji the Eugenias, as small trees and shrubs, find their home usually
on the banks of streams and rivers, on the outskirts of forests, and
occasionally at the coast. One of them, E. richii (Gray), is a
characteristic littoral tree in the group. A tree near it in character
was found by me of common occurrence in the interior of coral islets in
the Solomon Group (_Solomon Islands_, p. 297). E. rariflora occurs also
in the interior of coral islets in Fiji and amongst the vegetation at
the back of the mangrove-swamps.

Coming to the mode of dispersal of the genus in the Pacific, I may
remark that all the species, with the doubtful exception of the Fijian
and Samoan Eugenia neurocalyx (the Lemba of Fiji), are wild trees and
shrubs useless to man, but much appreciated by pigeons, pigs, &c., on
account of their fleshy fruits. Since exact observations on the
possibility of their dispersal by currents seemed to be wanting, I made
some experiments in Fiji. Out of six species, which included E.
corynocarpa, rariflora, richii, and rivularis, the mature fruits of most
species sank in sea-water in from seven to ten days. However, those of
the beach tree, E. richii, floated for a fortnight. The cause of sinking
in all cases lay in the decay of the outer fleshy covering. As I have
observed in river and sea drift, fish bite at the floating fruits, and
in this manner the seeds would soon be liberated and sink. The seeds of
all the plants sank at once in my experiments except with one species,
where the seed loosely filled its test and thus a floating-power of a
few days was acquired. Currents, it is apparent, could never account for
the dispersal of the genus over a broad extent of ocean, though in a few
cases, as in that of the littoral tree above noted, it is quite possible
that the fruits could be successfully transported across a tract of sea
200 or 300 miles in width.

It has long been known that fruit-pigeons are fond of the fruits of wild
species of Eugenia, and I found the Solomon Islanders and the Fijians
well acquainted with the fact. The fruits of a tall Eugenia tree, near
E. richii, common in the interior of the coral islets of Bougainville
Straits in the Solomon Group, were found by me in quantities in the
crops of fruit-pigeons shot by Lieut. Heming and Lieut. Leeper on the
islets (_Solomon Islands_, pp. 293, 297; _Bot. Chall. Exped._, Introd.
46, iv. 312). Dr. Seemann remarks that in Fiji the red fruits of E.
brackenridgei are eaten by pigeons. The somewhat thin coverings of the
seeds of this genus would seem to offer but a slight protection in a
bird’s stomach, though in one species the test was almost crustaceous.

Most species possessed only one or two large seeds in each fruit, though
this number may vary in the same individual. Thus, out of ten fruits of
Eugenia rariflora in Fiji, six had one seed, three had two seeds, and
one had three seeds. In the fruit of E. neurocalyx, however, the seeds
range from three to five.

It is the question of size that is of importance in considering the
possibility of birds transporting the seeds over a broad tract of ocean.
Eugenia rariflora, the species found all over the Pacific, has seeds
that measure in the Fijian plant one-fourth to one-third of an inch (6
to 8 mm.) across; and in Hawaii, according to Hillebrand, they would
perhaps be rather smaller. In point of size there is less difficulty
with regard to the transport by birds across the ocean to Hawaii of the
seeds of Eugenia rariflora than with the “stones” and seeds of some
other genera, like Elæocarpus, Osmanthus, and Sideroxylon, that must
have been conveyed there by the same agency. The fruits of several of
the Fijian species are of the size of a large cherry; but it is
noteworthy that in those species like E. corynocarpa and E. neurocalyx,
where the fruits are large and the seeds about an inch in size, the
plants are confined to the Western Pacific only, namely, to the
Fiji-Samoa region.

There is therefore no difficulty, from the standpoint of size, in
accounting for the distribution by birds of the widely-ranging Eugenia
rariflora over Polynesia; but at first sight there seems to be a real
difficulty with regard to the protective coverings of the seed. Yet
Nature speaks with no hesitating voice in the matter. The West Indian
and Florida species, E. monticola, regarded as indigenous in the
Bermudas, must have reached that group through the agency of birds that
carried its seeds over quite 800 or 900 miles of sea; and it may here be
noted that South Trinidad, lying some 600 miles off the coast of Brazil,
and Rodriguez, distant about 330 miles from Mauritius, each possess
species (_Bot. Chall. Exped._, Introd., 12, i. 32, ii. 128). If
fruit-pigeons can transport Eugenia seeds across 600 or 800 miles of
ocean, there would be no difficulty in accounting for the stocking of
the Fijian, Tongan, and Samoan Islands with the genus from regions to
the west. But the occurrence of the genus in Hawaii seems to compel us
to assume that the seeds have been carried in a bird’s stomach over
1,500 to 2,000 miles of ocean. This difficulty, however, does not really
exist. Eugenia rariflora, the Polynesian species found in Hawaii,
frequents, as before observed, coast districts and coral islets in Fiji,
and if we suppose that the low islands of the Fanning and Phœnix Groups,
lying between Hawaii and Samoa, have served as stepping-stones, a
capacity of crossing 1,000 miles of ocean would be alone required. This
is not much in excess of the distance that must have been traversed by
the bird that first brought the seeds of Eugenia monticola to the
Bermudas.

Other genera like Morinda and Scævola, possessing fleshy fruits
dispersed by frugivorous birds, have been mentioned in different
connections in other parts of this work, and will not be further dealt
with here. But before concluding this chapter I will refer briefly to
one of the disquieting mysteries in the flora of the Pacific which is
presented to us in the genus Gossypium. Three species are, or were,
truly indigenous in this region. One is Gossypium drynarioides, a small
endemic tree found by Nelson, the companion of Captain Cook, in Hawaii,
which was very rare in Hillebrand’s time, and is perhaps now extinct.
The second is G. tomentosum (Nuttall), which is also peculiar to Hawaii,
where it is found on the beaches. I am following here the _Index
Kewensis_; but it should be remarked that this species occurs also in
Fiji, though Seemann regards it as introduced. The third is G.
religiosum (L.), found by Captain Cook’s botanists growing wild in
Tahiti, and hailing from the tropics of the Old World. The seeds of the
first species are covered with a short brownish tomentum, and could
never have been of any value. The tawny wool of the seeds of the second
species has a staple too short for cultivation; whilst the Tahitians do
not seem to have made any use of the third species. It is difficult to
draw any conclusion concerning the presence of these plants in the
Pacific islands at the time of their discovery; nor can Dr. Seemann, who
was especially well informed in these matters, aid us much in our
endeavours to solve the mystery. From the aboriginal names we get no
clue. The Hawaiian name of “huluhulu” seemingly refers to the hairy
covering of the seed; whilst the Tahitian “vavai” and “ovari” simulate
the Fijian “vauvau,” which is merely the reduplicated form of “vau” (the
word in many shapes for Hibiscus tiliaceus in Malaya and Polynesia), and
is applied by the Fijians to Hibiscus esculentus and to the introduced
species of Gossypium.

When in Hawaii I ascertained that neither the seeds of the littoral
plant, Gossypium tomentosum, nor those of two cultivated species
possessed any fitness for dispersal by the currents, the scraped seeds
sinking at once, whilst when covered with the wool they floated only for
a few days. Further references to G. tomentosum in Hawaii are given in
the index of this volume.

_The Last Stage of the General Dispersal of Plants of the Malayan Era._

We arrive now at the close of the era of the general dispersal of
tropical plants, mainly Malayan, over the Pacific, and this brings us
down to our own age. The few genera that are still dispersed have no
peculiar species in particular groups. The species which often range
over all the groups, and retain as a rule their characters in most of
them, do not therefore display, except in a few cases, that extreme
variation which would give them a place in the ranks of the polymorphous
species. The dispersing agencies, in fact, are sufficiently active to
check marked variations, and the process of isolation has scarcely
begun.

We perceive the reason of this when we look at the nine genera which are
taken as samples of this period, viz., Rhus, Osteomeles, Viscum,
Plectronia, Boerhaavia, Polygonum, Pipturus, Boehmeria, and Dianella,
most of them being known to be dispersed by birds at the present day.
Six of the genera possess fruits likely to attract frugivorous birds;
whilst one of them, Boerhaavia, has sticky fruits that would be apt to
adhere to plumage. Actual observations in the cases of Rhus, Viscum, and
Plectronia establish the fact of their dispersal by fruit-eating birds;
and there is no difficulty in postulating the same agency for
Osteomeles, Pipturus, and Dianella. A method by which Boerhaavia fruits
would be transported in the plumage of birds has been observed by Mr.
Lister; whilst the nutlets of Polygonum are known to afford food to a
variety of birds and to be thus distributed.

In this period the plants all hail from the Asiatic side of the Pacific.
Three of the genera, Plectronia, Pipturus, and Dianella, belong almost
exclusively to the Old World. Five occur in both the Old and New Worlds,
but, as with Rhus, Viscum, Boerhaavia, and Boehmeria, are represented by
Old World species in the Pacific, or, as with Polygonum, possess a
cosmopolitan species (P. glabrum) ranging over the warm regions of the
globe. Even Osteomeles presents no exception to the rule, since the
Pacific plant is the only one of its species that is not American.

We have in Polygonum glabrum the only aquatic or semi-aquatic plant
widely distributed over the Pacific islands that can lay claim in all
groups to be indigenous. It is associated in Hawaii with species of
Potamogeton and Naias, aquatic genera that have, however, a limited
distribution in Polynesia.

I will now make a few remarks on each genus such as bear on their
distribution and on their mode of dispersal in the Pacific.

RHUS (Anacardiaceæ).—The representation of this genus by indigenous
species in oceanic islands not only in the Pacific but also in the
Atlantic, as in the Bermudas, is of especial interest in connection with
dispersal by frugivorous birds, since the drupes are typically dryish
and might appear to be not very attractive to birds. There are two Old
World species known from the Pacific islands: one being R. simarubæfolia
(Gray), distributed over the South Pacific groups from Fiji to Tahiti
and hailing from Malaya; the other, R. semialata (Murray), alone
recorded from the Hawaiian Group and derived probably from China or
Japan. This indication that the groups of the North and South Pacific
have derived their species, the first from Temperate Asia and the second
from Tropical Asia, is of some interest. In Samoa, according to
Reinecke, the fruits of R. simarubæfolia, which are of the size of a
pea, form the favourite food of the fruit-pigeons. That birds disperse
the seeds of the various Sumachs is familiarly known. In the United
States, as we learn from Barrows, Beal, and Weed, crows, woodpeckers,
and other birds feed extensively in winter on the fruits of different
species of Rhus, including the Poison Ivy (R. toxicodendron). The crows
discharge the seeds in pellets after retaining them for about thirty
minutes. Some seeds we must infer would pass into the intestines, where
they might be retained for ten to twelve hours (see Chapter XXXIII.),
which would be long enough, according to Gätke’s views of bird-velocity,
to enable them to be transported over a thousand miles of ocean.

OSTEOMELES (Rosaceæ).—One of the most interesting cases of dispersal in
recent times over the Pacific islands is that of O. anthyllidifolia. Of
the ten known species of the genus, nine are confined to South America;
whilst the Pacific species, which is not recorded from America, has been
found in Upper Burma, Japan, the Liukiu and Bonin Groups, Hawaii,
Pitcairn Island, Mangaia, and Rarotonga. The remarkable distribution of
the Pacific plant at once attracts attention. I was very familiar with
it in Hawaii, where it forms one of the commonest bushes in open-wooded
and thinly vegetated districts at elevations usually ranging from the
coast to 3,000 feet. Its small, white, somewhat fleshy fruits would
attract birds, and the hard pyrenes would be able to pass unharmed
through a bird’s digestive canal. It seems probable that, like Rhus
semialata, this plant entered the Pacific Ocean from the north-west,
taking the route by Japan and the Bonin Islands, and following the trend
of the archipelagoes over Polynesia (see _Bot. Chall. Exped._, Introd.
p. 18; _Journ. Linn. Soc. Bot._, vol. 28, 1891, &c.).

VISCUM (Loranthaceæ).—A single species, V. articulatum, which has its
home in Southern Asia, is found in most of the Pacific groups, such as
Hawaii, Marquesas, Tahiti, Rarotonga, Fiji, &c. The dispersal of the
genus by frugivorous birds is well known.

PLECTRONIA (Rubiaceæ).—I have found it more convenient to place this
genus here, although there are probably one or two species peculiar to
Fiji. This genus of shrubs, which is spread over the warm regions of the
Old World, is represented by two widely distributed species in
Polynesia, Plectronia odorata (B. and H.) and P. barbata (B. and H.),
the first alone extending to Hawaii. I was very familiar with P. odorata
in Hawaii and was much interested in its mode of dispersal, since the
species has also been found in Fiji, Tahiti, the Marquesas, and Pitcairn
Island (Maiden). In one locality, where an old lava-field was partially
covered by its bushes then in fruit, the doves were feeding greedily on
the drupes, the “stones” of which, as well as the partially digested
fruits, were to be seen in quantity in their excrement near a
water-hole. The stones are very hard and about a third of an inch (8
mm.) in length, and are exceedingly well suited for transport by
frugivorous birds. It was very probably to one of these species of
Plectronia that Peale alluded when he wrote of the berries of a species
of Canthium forming the principal food, on one of the Paumotu Islands,
of Numenius tahitensis, a curlew that has its home in Alaska, migrating
south in autumn to Hawaii, Tahiti, and the Paumotu Group (Wilson’s AVES
HAWAIIENSES).

BOERHAAVIA (Nyctagineæ).—Two or three Asiatic species of this genus, B.
diffusa, B. tetranda, &c., are spread all over the Pacific islands from
the Fijis to the Paumotus and northward to Hawaii. Similar or allied
species occur on the coral islands of the Indian Ocean, as on Diego
Garcia and on Keeling Atoll. Though these plants have often been
accidentally spread by man with his cultivated plants, it is probable
that sea-birds have regularly aided in their dispersal. The fruits, on
account of their small size and their glutinous sticky surfaces, are
well suited for transport in a bird’s feathers. Mr. Lister, as quoted by
Hedley (from _Proc. Zoolog. Soc._, 1891), made an interesting note in
this connection on one of the islands of the Phœnix Group, where he
found a fruit of Boerhaavia tetrandra entangled in some of the down that
had been preened by a booby (Sula piscatrix) out of its feathers whilst
roosting in a clump of Tournefortia trees.

POLYGONUM (Polygonaceæ).—This genus is represented by the cosmopolitan
Polygonum glabrum, the only aquatic or semi-aquatic plant that is
generally distributed in the Pacific islands. It occurs in fresh-water
swamps and beside streams and ponds in Tahiti, Tonga, Fiji, Hawaii, &c.,
and was gathered by Banks and Solander when Captain Cook first visited
Tahiti. That this plant has been distributed by geese, ducks, and
waterfowl over the tropics of the globe can scarcely be doubted. In
England I have found the nutlets of Polygonum convolvulus, P.
persicaria, and P. aviculare in the stomachs of a wild duck and a
curlew; and they came frequently under my notice in the crops and
intestines of different kinds of partridges and of wood-pigeons. Though
most of the fruits were generally injured, a few of them were not
uncommonly obtained in a sound condition.

PIPTURUS (Urticaceæ).—This is a genus of small trees and shrubs found in
the Mascarene Islands, Malaya, Australia, New Zealand, and throughout
Polynesia. Besides P. albidus, which is confined to Hawaii and Tahiti,
there are two Malayan species, P. argenteus and P. velutinus, which are
widely distributed over the islands of the South Pacific, extending to
Tahiti and the Marquesas. The fleshy receptacle and small achenes of the
compound fruit of Pipturus give it the appearance of a white immature
strawberry, and as such it would be likely to attract frugivorous birds.
Plants of this genus are included amongst the numerous plants from the
bast of which the natives used to prepare their native cloth or from
which they obtained the fibres for their fishing-lines.

BŒHMERIA (Urticaceæ).—There is an Asiatic species widely spread in the
South Pacific and another closely-allied species in Hawaii; but I
possess no data relating to the dispersal of the genus. The fruits are
dry and consist of an achene in a persistent perianth.

DIANELLA (Liliaceæ).—This is a genus of herbs, possessing often pretty
blue berries, that extends over tropical Africa, tropical Asia, the
Mascarene Islands, Malaya, Australia, and New Zealand, and is found in
all the larger Pacific archipelagoes. Of the twelve species named in the
_Index Kewensis_ only two belong to America, occurring respectively in
Cuba and Venezuela. There are two species in the islands of the tropical
Pacific: (_a_) Dianella ensifolia, found in Hawaii and ranging over the
Mascarene Islands, India, China, Malaya, and tropical Australia; and
(_b_) D. intermedia, recorded from most of the groups of the South
Pacific (Fiji, Tonga, Rarotonga, Tahiti), and occurring also in Norfolk
Island and New Zealand. These two plants occur in similar stations all
over Polynesia, sometimes growing in the grassy plains on the dry side
of an island, at other times extending up the thinly wooded mountain
slopes and reaching the hill-crests some 2,000 or 3,000 feet above the
sea. Their berries would readily attract birds; and their seeds, about
one-fifth of an inch (5 mm.) in size in the case of D. ensifolia, could
be carried uninjured in the stomach and intestines of a bird.

_Summary._

(1) A later period in the era of the general dispersal of Malayan plants
over the Pacific is indicated by the genera that contain species found
outside each group as well as species restricted to it.

(2) In this period the extremely variable or polymorphous species plays
a conspicuous part, as represented in such genera as Alphitonia,
Dodonæa, Metrosideros, Pisonia, and Wikstrœmia.

(3) The first stage is displayed by a solitary widely-ranging species
found over most of the Polynesian archipelagoes, and varying
independently in every group.

(4) The next stage is shown where the polymorphous species, having done
its work of distributing the genus, ceases to wander and settles down
and “differentiates” in all the groups; and the genus thus includes both
peculiar and widely-ranging species in each group. Most of the genera
possessing polymorphous species are in this stage.

(5) The following stage is displayed by those genera like Elæocarpus,
Eugenia, and Peperomia, where peculiar species are especially developed
in particular groups, and we get subcentres of distribution for the
genus, that is to say, small gatherings of peculiar species. A few
species, however, still keep up a connection with neighbouring
island-groups. Should this be severed we get the type of genus belonging
to the earlier period of the Malayan era as described in the preceding
chapter, a genus possessing only peculiar species and destined, after
ages of further isolation through the failure of the dispersing
agencies, to give rise to a new generic type or types.

(6) Frugivorous birds were chiefly active in dispersing these genera
over the Pacific. Some of the genera possess seeds or “stones” of such a
size that at first sight their transport by frugivorous birds to Hawaii
seems improbable; but, as in the case of Elæocarpus, it is shown that
this difficulty does not apply to all species of a genus, some of them
having much smaller seeds or stones.

(7) The close of the era of the general dispersal of Malayan plants over
the Polynesian Islands is indicated by those genera that are represented
more or less entirely by widely ranging species. Though such species may
vary among the different groups, they rarely take the rank of
polymorphous species, the agencies of dispersal being sufficiently
active to check marked variations.

(8) Several of the genera of this concluding stage, like Rhus, Viscum,
and Plectronia, are known to be dispersed by frugivorous birds, whilst
others, like Osteomeles and Dianella, are equally well suited for this
mode of dispersal.

(9) Distinct indications are afforded by the genera Rhus, Osteomeles,
and Dianella that the Hawaiian Group has been often supplied with its
plants directly from the Old World by the Asiatic mainland, whilst the
groups of the South Pacific have received different species of the same
genus by Malaya and tropical Australia.

CHAPTER XXVII

THE MALAYAN ERA OF THE NON-ENDEMIC GENERA OF FLOWERING PLANTS
(_continued_)

THE AGE OF LOCAL DISPERSAL

_Synopsis of the Chapter._

HAWAII.—(1) _The Hawaiian residual genera, being those not found in
either the Fijian or the Tahitian regions._ The genera especially
discussed are Osmanthus, Sicyos, Jacquemontia, Cuscuta, Rumex,
Dracæna, Naias, Potamogeton; and amongst others mentioned are
Perrottetia and Embelia.

(2) _The Hawaiian genera found in Tahiti and not in Fiji._ Very few,
and illustrated by Byronia, Reynoldsia or Trevesia, Phyllostegia, and
Pseudomorus, though it is likely that most of these will be
subsequently discovered in Fiji.

(3) _The Hawaiian genera found in Fiji and not in Tahiti._ Illustrated
by Eurya, Gouania, Maba, Sideroxylon, Antidesma, Pleiosmilax, Ruppia.

(4) _The absentees from Hawaii._ Illustrated amongst the orders by the
Sterculiaceæ (see text), the Meliaceæ, the Rhizophoreæ, the
Melastomaceæ, and the Coniferæ, and amongst the genera by
Trichospermum Loranthus, Stylocoryne, Ophiorrhiza, Alstonia, Hoya,
Ficus; and a great many others might be cited.

TAHITI.—(1) _The Tahitian residual genera._ Only six in number—Cratæva,
Buettneria, Berrya, Coriaria, Bidens, Lepinia.

(2) _The Tahitian genera found in Hawaii and not in Fiji._ See above
under (2).

(3) _The Tahitian genera found in Fiji and not in Hawaii._ (a) Those
possessing only species confined to the Tahitian region or to East
Polynesia, of which Meryta, Ophiorrhiza, Alstonia, and Loranthus are
examples.

(b) Those possessing widely-ranging species besides, often, species
confined to the Tahitian region, such as Grewia, Nelitris, Melastoma,
Randia Geniostoma, Tabernæmontana, Fagræa, Bischoffia, Macaranga, and
Ficus. The widely-ranging species is in many genera polymorphous.

(4) _The absentees from Tahiti._ Amongst the orders are the Meliaceæ,
the Rhizophoreæ, and the Coniferæ. Amongst the genera, usually those
with “stones” or large seeds an inch in size, such as Canarium,
Dracontomelon, Myristica, Sterculia, Veitchia, &c. Numerous other
absent genera might be named.

FIJI.—_The Fijian genera not found either in Tahiti or Hawaii._ These
genera compose about half the Fijian flora, being at least 160 in
number. Those especially discussed here are the following:—Hibbertia,
Cananga, Sterculia, Trichospermum, Micromelum, Canarium,
Dracontomelon, Begonia, Geissois, Dolicholobium, Lindenia, Myrmecodia,
Hydnophytum, Couthovia, Limnanthemum, Myristica, Elatostema,
Ceratophyllum, Gnetum, Veitchia, Rhaphidophora, Lemna, Wolffia,
Scirpodendron. The Coniferæ are dealt with in Chapter XXIV.

--------------

Note appended on Marsilea

HAVING completed our discussion of the general dispersal of tropical
genera, chiefly Indo-Malayan, over the Pacific islands, we pass on now
to consider the more restricted distribution of non-endemic genera over
this region. Here as before we take Hawaii, Tahiti, and Fiji as the
three centres of distribution; and here also we deal with the flowering
plants after excluding the orchids, the sedges, the grasses, the
mountain-plants, and all plants introduced either by the aborigines or
by white men.

HAWAII.

After excluding the endemic genera as well as those that are confined to
the mountains, we find that this group possesses very few genera that do
not occur in the Fijian and Tahitian regions, and fewer still that it
owns in common with Tahiti to the exclusion of Fiji. On the other hand,
we observe that Fiji possesses a great number of genera, mostly Asiatic
in origin, that have not reached Hawaii, and in several cases are not
known, from the Tahitian region. These contrasts might have been
expected, since the Pacific islands have in later ages been mainly
stocked from the Asiatic side of the Pacific, the principal route lying
through the Fijian region.

As far as the flora of the lower levels (below 4,000 feet) is concerned,
Hawaii only possesses a portion of that which Fiji has derived from the
Old World, chiefly through Malaya. Although, as will be shown below,
there is a noticeable contribution from America, it is very far from
counterbalancing the loss which the Hawaiian flora has sustained in
comparison with Fiji through the isolated position of the group. The
want of variety, however, in the flora of the Hawaiian lower levels,
which up to 4,000 or 5,000 feet represent the islands of the less
elevated Fijian region, is in a small degree compensated for by the
development of new genera and new species and by the great number of
individuals. Trees like Metrosideros polymorpha and Aleurites moluccana,
that in the southern groups form only one of many contributors to the
forests, rise suddenly into prominence in the northern archipelago and
form entire forests. Pandanus odoratissimus largely composes extensive
forests in the province of Puna in the large island of Hawaii, extending
several miles inland and nearly 2,000 feet up the mountain slopes.

The remarkable contrast between the Fijian flora, which is almost
entirely tropical, and the Hawaiian flora, which on account of the great
elevation of the islands is temperate as well as tropical, is brought
into yet greater prominence when we look at it more closely and treat it
numerically. The Hawaiian Group, it must be first observed, though
possessing the same area as Fiji and presenting a far greater variety of
climatic conditions, has only two-thirds the number of genera of
flowering plants (see Chapter XXI., Table B). Whilst at least 200 of the
Fijian genera of indigenous plants (excluding the orchids and the
grasses) are not found in Hawaii, only about 100 of the Hawaiian genera
are absent from Fiji, and the two groups possess about 100 genera in
common. When we look more closely at the hundred Hawaiian genera not
found in Fiji, we find that about sixty represent endemic genera
(thirty-seven) and non-endemic mountain-genera (twenty-two), which
naturally are not to be found in Fiji, so that there remain but a small
number of genera distinguishing the tropical flora of Hawaii from the
Fijian flora. When we take from them a few that occur in the Tahitian
region, there is left a very small residuum characteristic of Hawaii
alone to the exclusion of the Fijian and Tahitian regions of the South
Pacific.

THE HAWAIIAN RESIDUAL GENERA.

It is my purpose now to deal in an illustrative fashion with this
Hawaiian residual flora which is composed, as above explained, of the
non-endemic tropical genera that are not represented in the Fijian and
Tahitian regions. Up to the present we have been dealing with the
characters that the floras of Fiji, Tahiti, and Hawaii possess in common
as far as tropical genera are concerned. We will now proceed to discuss
their differences in this respect, and will begin with the residual
Hawaiian flora.

After eliminating two or three genera that will probably be found in
Fiji, but including one or two others that are best treated under the
endemic genera, about twenty-seven present themselves for our purpose.
Nearly all of them possess only endemic species, and belong therefore to
an age of dispersal that has passed away. These residual genera plainly
indicate that although Hawaii largely received its flora during the age
of general dispersal of Old World genera over the Pacific, it was at the
same time independently stocked with plants from other sources. They
include among others—Cocculus (4), Cleome (1), _Perrottetia_ (1),
Mezoneuron (1), _Lythrum_, _Sicyos_ (8), Peucedanum (2), Campylotheca
(12), Senecio (2), Lobelia (5), Embelia (1), _Chrysophyllum_ (1),
Rauwolfia (1), _Nama_ (1), Osmanthus (1), _Jacquemontia_ (1), Breweria
(1), Cuscuta (1), Lycium (1), _Sphacele_ (1), _Phytolacca_, Rumex (2),
_Urera_ (2), Pilea, Dracæna (1), Naias, Potamogeton. Those printed in
italics are regarded as derived from America; whilst the figures in
brackets indicate the number of endemic species, nearly all of the
genera except the five above indicated possessing only peculiar species,
and these five (Lythrum, Phytolacca, Pilea, Naias, Potamogeton) are only
represented by species found outside the group.

American genera form a more conspicuous element than they do amongst the
genera that have been generally dispersed over the Pacific, those
exclusively American being fairly represented, making a third of the
whole. We find, for instance, in the Hawaiian “Olomea,” Perrottetia
sandwicensis, a small tree that represents in the woods of all the
islands the Perrottetias of Mexico and the Andes; whilst with some of
those genera that, like Sicyos and Urera, are at home in both the Old
and New Worlds, we obtain indications of America being the source of the
Hawaiian plants. A few genera again, like Lythrum and Phytolacca, are
represented in Hawaii by American species.

Plants with drupes, berries, or other fleshy fruits likely to attract
frugivorous birds compose about a third of the total number of these
residual genera, whilst fruits or seeds, that were in all probability
originally brought entangled in a bird’s feathers, are represented by
Sicyos. Some of the genera with stone fruits, such as Osmanthus, to
which belongs the Hawaiian Olive, present special difficulties on
account of the size of the stone, in this case two-thirds of an inch in
length. There are also a number of genera with large dry fruits and
sometimes large seeds, of which the method of dispersal is not easy to
discover. Thus, Mezoneuron, a Leguminous genus with seeds an inch across
(2·5 cm.), and Peucedanum, of the Umbelliferæ, with mericarps half to
three-quarters of an inch (1·2 to 1·8 cm.) in length, offer serious
difficulties to the student of plant-dispersal. In discussing the
difficulty connected with Mezoneuron (see Chapter XV.) he will keep in
view the possibility that the original species may have been a littoral
plant possessing seeds dispersed by the currents, seeds that lost their
buoyancy when the plant established itself inland, just as is now taking
place with Afzelia bijuga, a Leguminous littoral tree of Fiji (see
Chapter XVII.).

He will also find much to puzzle him in the mode of dispersal of the
Hawaiian residual genera of the Convolvulaceæ (Breweria, Jacquemontia,
and Cuscuta) that possess only endemic species, and he will speculate as
to the manner in which seeds that would seem to possess but little
attraction for birds and have no capacity for transportation by the
currents could ever have reached these islands, and he will ask himself
why it is that the agencies of dispersal, whatever they are, have now
ceased to be active. He will perhaps see a way out of his difficulties
when he perceives that if isolation has led to the development of
peculiar species in Hawaii, it has strangely enough in the case of the
Myrsinaceous genus Embelia produced the same effect over the whole range
of the genus, and that Hawaii has in this respect derived no advantage
from being an oceanic group. According to Carl Mez, nearly all the
ninety species of this Old World genus are restricted in their areas,
whether continental or insular (“Myrsinaceæ,” _Das Pflanzenreich_,
1902); and indeed we do not seem justified in assuming that the
isolating influences in the case of this genus have been more effective
in Hawaii in the mid-Pacific, or in Mauritius in the Indian Ocean, than
they have been in continental regions like the Deccan and Nyassa Land,
in all of which localities endemic species occur.

The remarkable development of the Cucurbitaceous genus Sicyos, in Hawaii
alone of all the tropical Pacific groups, will attract his attention,
and he will find here another instance of that predominant principle in
the distribution of Pacific plants, where in a widely-ranging genus we
find one of its species covering most of its area, whilst the other
species are more or less localised. He will wonder at the limitation to
Hawaii of a genus like Dracæna, that is so well adapted for dispersal
over the Pacific by frugivorous birds; and in endeavouring to explain
the presence in the Hawaiian forests of the gigantic Rumex, R.
giganteus, he will remember that the small group of Tristan da Cunha,
equally isolated in the South Atlantic, possesses an endemic species of
the same genus. He will discover in the recognised dispersing agencies
of wild ducks and other waterfowl an explanation of the occurrence in
Hawaii of the aquatic genera Naias and Potamogeton; but he will be
puzzled at their restriction to this group alone of the three tropical
Pacific archipelagoes here especially discussed.

Amidst these various perplexities he will probably look with relief on
the appearance of Phytolacca brachystachys, an endemic species of the
American “pokeweeds”; and he will feel grateful to the American
botanists like Professor Weed when they tell him that in the United
States crows, blackbirds, and other birds successfully disperse these
plants, the seeds of which are sometimes able to pass through the
alimentary canal undigested.

But by far the most significant lesson that the student of distribution
will carry away from his study of the Hawaiian residual genera will be
that which he learns from the genera Embelia and Naias. He perceives
here that not only with a typical land-genus has specific
differentiation occurred to much the same extent in the continental and
insular localities of its range, but that even with a typical genus of
submerged aquatic plants, where the conditions of existence are as
uniform as they are varied in the case of land plants, the process of
differentiation has proceeded on the same broad lines in the interior of
a continent and in an island in mid-ocean.

The following notes on some of the residual genera refer more
particularly to matters connected with distribution and dispersal.

_Osmanthus_ (Oleaceæ).—This genus, according to the _Index Kewensis_,
contains six species localised in their several habitats of North
America, Hawaii, Japan (two), Hongkong, and the Himalayas. Its
representative in this group is the Hawaiian Olive, the Olea
sandwicensis of Gray, a prevailing tree in the lower and middle woods
(1,000 to 4,000 feet) of all the islands, which, like other Hawaiian
plants, such as those of the genera Eurya and Antidesma, indicates that
the group has been sometimes independently stocked from the regions of
the northern hemisphere. The drupe of this tree contains a stone
two-thirds of an inch (17 mm.) in length, and suitable for dispersal by
frugivorous birds; and birds have evidently distributed the tree all
over the group. In fact Mr. Perkins in mentioning the favourite food of
birds of the Hawaiian genus, Phaeornis, refers to the fruits of this
tree as well as of the Opiko (Straussia) and of the Olapa
(Cheirodendron). When, however, we come to consider the feasibility of
the stones of the genus having been thus originally carried to Hawaii
either from Japan or from North America, we meet with the difficulty
presented to us by other Hawaiian genera with stone-fruits, such as
Elæocarpus, or with berries containing large seeds, such as Sideroxylon.

_Sicyos_ (Cucurbitaceæ).—This genus comprises about thirty-five known
species, of which three-fourths are confined to the New World, being
mainly South American, whilst the remainder are restricted to Hawaii,
with the exception of two species in the Galapagos Group and Norfolk
Island, and a widely-ranging species, S. angulatus. The plant just
named, the small fruits of which possess hooked spines, adapting them
for dispersal in a bird’s plumage, occurs in Africa, Australia, New
Zealand, and America, but has only been recorded in the Pacific islands
from the Kermadec Group.

North America was probably the home of the original Hawaiian species.
Hillebrand describes eight species, of which five are not found in more
than one island, whilst one species is spread over most of the islands.
The fruits vary much in size, and only in a couple of species do they
now possess any fitness for attaching themselves to plumage, some of
them being pubescent or even glabrate, so that deterioration in the
capacity for dispersal has here taken place. Their size is usually a
quarter to half an inch (6-12 mm.); but it is noteworthy that the
species with the largest fruit (Sicyos cucumerinus, one to two inches,
or 25 to 50 mm.) is the species most widely dispersed over the group.
This appears to indicate that there is some other means of inter-island
dispersal in this archipelago than by attachment to birds’ plumage. The
isolation of the genus in Hawaii from the rest of the world is, however,
complete, since all the species are endemic; and when, therefore, we
come to ask how Sicyos angulatus, that has been dispersed in the recent
era over America, Australia, and New Zealand, is not found in these
islands, we are brought face to face with the ever-recurring difficulty,
the suspension in later times of the agency of dispersal in the tropical
North Pacific.

_Jacquemontia_ (Convolvulaceæ).—This genus, which is chiefly American,
is represented in Hawaii by a peculiar species, J. sandwicensis. This
species grows occasionally on the sandy beaches associated with
Heliotropium anomalum and Tribulus cistoides; but it is most at home on
rocky ground and on old lava-flows near the sea-border, making its abode
often in the pockets of black sand produced by the disintegration of the
lava. Its small seeds sink in sea-water even after prolonged drying; and
it can perhaps be supposed that the original seeds were brought from
North America in the crevices of a drifting log. According to Ridley,
Fernando Noronha possesses a peculiar species also growing near the sea;
and it may be that the drifting log has here been the agent also: but in
neither case would this explanation account for the endemic character of
the species.

_Cuscuta_ (Convolvulaceæ).—It would seem that with the exception of
Hawaii, where an endemic species, C. sandwichiana, occurs, no other
oceanic group in the globe possesses a peculiar species of the Dodders.
With the exception of an endemic species in New Zealand, and an
introduced species in Fiji which is found usually near the gardens of
the white residents on Viti Levu, the genus takes but little part in the
Pacific floras. The Hawaiian species is a characteristic beach-plant
growing on Ipomœa pes capræ, Scævola Kœnigii, Tribulus cistoides, and on
other plants that find a permanent or a temporary abode on the beaches.
We learn from Ridley and Moseley that Cuscuta americana in Fernando
Noronha finds its host also in Ipomœa pes capræ. Since the seeds of the
Hawaiian plant and of the introduced Fijian species possess no buoyancy,
even after drying for years, we cannot look to the agency of the current
unless we call the drifting log to our assistance, and in that case the
endemic character of the Hawaiian species would present the difficulty
already alluded to in the case of Jacquemontia. The seeds of the
Hawaiian plant are about one-twelfth of an inch (2 mm.) in diameter, and
as far as size is concerned they might have been transported in a bird’s
stomach; but, on account of the rapidity with which the seeds of the
genus absorb moisture and swell up, it is most unlikely that they would
escape injury. This is one of the several difficulties in
plant-dispersal which New Zealand and Hawaii share in common. Further
remarks on the germination of the Hawaiian species are made in Note 69.

_Rumex_ (Polygonaceæ).—Hawaii possesses two peculiar species of Rumex, a
genus not recorded from any other of the Polynesian groups. One of these
species, R. giganteus, is a very remarkable plant, growing to a height
of thirty or forty feet when supported by trees. It is noteworthy that
the small group of Tristan da Cunha in the South Atlantic possesses a
species, R. frutescens, confined to those islands (_Bot. Chall. Exped._,
ii. 154). Both Hawaii and Tristan da Cunha lie in mid-ocean, cut off
from the nearest continent by some 1,800 or 2,000 miles of sea; and we
may have to choose between the bird and the current in selecting the
agency concerned with the transportation of the original seeds; or
perhaps they have co-operated. Birds could disperse the nutlets of Rumex
as readily as they do those of Polygonum, and I have found these fruits
at times in the stomachs of partridges. On the other hand, Rumex fruits
occur amongst the drift stranded on beaches in England and in
Scandinavia; and, as indicated by the observations of Sernander and
myself in these two localities, they float through the winter in ponds
and rivers, germinating afloat in the spring. The nutlets sink, but they
owe their buoyancy to the persistent perianth. In my sea-water
experiments the fruits of Rumex hydrolapathum and R. conglomeratum were
still afloat after from six to twelve months’ immersion, and their seeds
subsequently germinated. It is quite possible, therefore, that currents
can carry these fruits unharmed to oceanic island-groups like Hawaii and
Tristan da Cunha.

_Dracæna_ (Liliaceæ).—This Old World genus, which on account of its
berries is eminently suited for dispersal by frugivorous birds, is
represented in Polynesia by a solitary species (D. aurea) peculiar to
the Hawaiian Group. Attaining a height of twenty to twenty-five feet, it
often forms a striking feature in the vegetation of the open wooded
regions up to altitudes of 3,000 feet. I found it growing in abundance
in the large island of Hawaii between Waimanu and Waipio, and on the
northern slopes of Hualalai. It grows in a variety of stations, and I
came upon it once in the broken-down caverns of an old lava-flow that
were frequented by pigeons which no doubt brought the seeds. Its
conspicuous yellow berries have hard rounded seeds a quarter of an inch
(6 mm.) across and weighing two to three grains when dry, which would
probably withstand injury in a bird’s stomach, the minute embryo being
protected by a very tough albumen. Neither the entire berry nor the seed
could be transported by currents, the last sinking even after drying for
six years.

_Naias_ (Naiadaceæ).—If we except New Caledonia, where two or three
species have been found, Hawaii is the only island-group in the tropical
Pacific from which this interesting world-ranging genus of submerged
aquatic plants has been recorded. Chamisso, the celebrated naturalist of
Kotzebue’s expedition, collected Naias marina in Oahu in the early part
of last century; but apparently it did not come under Hillebrand’s
observation in the group. However, in 1897 I found it in another
locality, namely, just within the mouth of the Waipio, a river on the
north-west side of the island of Hawaii. The mature fruits of this genus
have never been experimented on by me; but there is nothing in the
structure of the fruits to indicate that they have any buoyancy, or to
show that they differ in this respect from the fruits of other
completely submerged aquatic plants like Ceratophyllum, Ruppia, and some
of the Potamogetons. It is to ducks and other waterfowl that we must
attribute the dispersal of this and the other genera just mentioned over
wide tracts of ocean, a subject dealt with in discussing those plants.

The Hawaiian Group probably represents the most isolated locality
occupied by this genus, since none of the other islands from which
species have been recorded, such as New Caledonia, Mauritius, and
Bourbon, are so far removed from continental regions. The source of the
Hawaiian form of Naias marina lies evidently on the Asiatic side of the
Pacific, since it is referred by Mr. Rendle to the variety
“angustifolia,” an Asiatic plant found also in the island of Bourbon and
in West Australia, but not recorded from the New World. The important
little monograph of the genus by Mr. Rendle (“Naiadaceæ,” in Engler’s
_Das Pflanzenreich_, 1901) is full of suggestiveness for the student of
plant-distribution. His interest is excited when he discovers that one
of the most typical genera of aquatic plants displays the same principle
of differentiation at work that is so well illustrated by many of the
land genera of the Pacific islands. I refer to the principle implied in
the existence of a widely-ranging genus comprising “a polymorphic
species occurring over almost the whole area of the genus,” as well as a
number of less widely distributed species, most of which have
“restricted areas and fall for the most part into small geographical
groups.” I have just been quoting Mr. Rendle’s description of the
distribution of Naias, the “polymorphic” species concerned being N.
marina; but it need scarcely be remarked that it would apply just as
well to several of the land genera dealt with in the previous chapter
(XXVI.), such as Alphitonia, Metrosideros, Pisonia, &c.

Although there is such a contrast in the degree of uniformity of their
life-conditions between land and water plants, a strictly aquatic plant
being but slightly affected by changes in the physical conditions that
are accompanied by a complete transformation in the character of the
terrestrial vegetation, yet—and this is the important point—we find the
same principle of differentiation at work with both land and water
plants. If one wished to produce proof of the contention that the
production of new species is largely independent of external conditions,
one could not do better than take the cases of Elæocarpus, Metrosideros,
and Naias. In all cases we see a widely-ranging polymorphous species
settling down and “differentiating” in particular localities or regions,
and forming subcentres for the distribution of the genus.

_Potamogeton_ (Potameæ).—Though well suited for dispersal by waterfowl,
the Potamogetons have been recorded from the Hawaiian and Marianne
Islands alone among the tropical groups of the open Pacific. The genus,
though not so well represented in insular floras as we might have
expected, is still not infrequently to be found. Widely-ranging species
have been observed in the Azores, Madeira, and the Canaries in the
Atlantic, as well as in Hawaii in the Pacific; whilst species have been
recorded that are peculiar to Martinique, the Mascarene Islands, and to
the Marianne Group. Hillebrand gives for Hawaii, Potamogeton fluitans, a
plant of the Old and New Worlds, and P. pauciflorus, a North American
species; whilst in the _Index Kewensis_ a peculiar species, P.
owaihiensis of Chamisso (which is, however, regarded by Hillebrand as a
form of P. fluitans), is also accredited to the group. Owing, however,
to the paucity of streams and rivers this genus takes no prominent part
in the Hawaiian flora, and the species seem to have been recorded alone
from Oahu. As they were discovered by Chamisso in the early part of last
century they are in all probability truly indigenous in Hawaii, even if
none are peculiar to the group.

That ducks and similar birds are the agents in carrying the seeds of
Potamogeton to oceanic islands cannot be doubted. About twelve years ago
I examined the stomachs and intestines of thirteen wild ducks obtained
in the London market. Three of them contained in all forty-one
Potamogeton seeds, or rather “stones,” most of which subsequently
germinated in water. In one of my experiments, carried out in the month
of December, I fed a domestic duck with the fruits of Potamogeton
natans. They appeared in quantity in the droppings, for the most part
divested of their soft coverings, but otherwise uninjured. Sixty per
cent. germinated in the following spring; whilst of those left in the
vessel, from which the duck had been fed, only one per cent. germinated
in the next spring, and another year elapsed before any number did so.
These results were published in _Science Gossip_ for September, 1894.

One often reads in books of travel interesting remarks bearing
indirectly on the dispersal of the Potamogetons. Thus, when Sir Joseph
Hooker (then Dr. Hooker) noted in his _Himalayan Journals_ the
occurrence of P. natans in the Neongong Lake in the Himalayas, and the
presence of coots, he most probably mentioned the bird that brought the
plants, coots being active distributors of the seeds of water plants. It
is of importance to remember that (as shown in my experiment on the
duck) seeds of water-plants are voided in a condition peculiarly
favourable to early germination. Ducks, coots, and other water birds
might often be characterised as “travelling germinators.” My experiment
showed that seven to eight hours at least were occupied by Potamogeton
nutlets in passing through the digestive canal of a duck, and that
probably nine or ten hours would be required after an average full meal.
But this does not represent the possible maximum period, since the bared
“stone” may remain in the gizzard for a long time with ordinary gravel.
Most of the Potamogeton fruits found by me in wild ducks were obtained
from the gizzard, where they were mixed with gravel and other hard seeds
or seedvessels, as described in Chapter XXXIII. Such fruits afterwards
germinated. With regard to the chances, therefore, of the fruits of
Potamogeton being carried by a bird without injury across an ocean, we
may infer that, whether they are retained in its body for only ten hours
or for as long as three or four days, they will preserve in some cases
their germinating power.

HAWAIIAN GENERA FOUND IN TAHITI TO THE EXCLUSION OF FIJI.

Taking only the genera that are strictly indigenous, and excluding
therefore all those introduced by the aborigines, the number available
for establishing an independent connection between the Hawaiian and
Tahitian regions is exceedingly few. Amongst the Hawaiian shore-plants
not found in Fiji proper but occurring in the Tahitian region are
Heliotropium anomalum and Sesuvium portulacastrum. The last-named,
however, has been recorded from Tonga, which lies within the Fijian
area; whilst the first will probably be found in the same region.
Amongst the Hawaiian and Tahitian mountain genera not recorded from Fiji
proper are Nertera, Vaccinium, Cyathodes, and Luzula. As is pointed out
in Chapter XXIII., the absence of these genera from Fiji is connected
with the relatively low elevation of those islands, though it is quite
possible that one or more of them may yet be found on the highest
summits of Fiji; and indeed Nertera depressa and Vaccinium have been
discovered in the more elevated uplands of Savaii in Samoa.

After removing the littoral plants and the mountain genera, there are
probably not more than half a dozen inland genera that connect the
Hawaiian lowlands with the Tahitian region to the exclusion of the
Fijian Group; and Byronia (Ilicineæ), Reynoldsia or Trevesia
(Araliaceæ), Phyllostegia (Labiatæ), and Pseudomorus (Urticaceæ) may be
taken as examples. Of these, Pseudomorus, which has a small drupaceous
fruit suitable for dispersal by frugivorous birds, has been recorded
from New Caledonia, and not improbably it exists in the Fijian area; and
the same may be postulated of Reynoldsia, which is discussed in a later
page, since it has been found in Samoa. We may almost form the same
opinion of Byronia, since it exists in Australia. This genus of small
trees contains only three known species, one in Australia, one in
Tahiti, and one in Hawaii. Its fleshy drupes, about a third of an inch
(8 mm.) in size, would attract birds, and their numerous cartilaginous
pyrenes would probably pass unharmed through a bird’s alimentary canal.
Phyllostegia, a Labiate genus with fleshy nucules that might attract
birds, is, with the exception of a solitary Tahitian species, entirely
confined to Hawaii (see Chapter XXII.).

From these data it may be inferred that the interchange of plants
between the regions of Hawaii and Tahiti to the exclusion of Fiji has
been very slight. The facts of distribution are just such as we might
look for in the case of a general dispersal over the oceanic groups of
the tropical Pacific, with the altitudes of the islands playing a
determining part. In this general dispersal Hawaii has shared; and
except in the case of Phyllostegia it is evident that this group has
kept nearly all it received and has distributed but little.

HAWAIIAN GENERA FOUND IN FIJI TO THE EXCLUSION OF TAHITI.

We shall be able to throw further light on the floral history of Hawaii
by discussing the few tropical genera, not a score in all, that it
possesses in common with Fiji to the exclusion of the Tahitian region.
The following genera offer themselves for treatment:—Eurya
(Ternstrœmiaceæ), Gouania (Rhamnaceæ), Maba (Ebenaceæ), Sideroxylon
(Sapotaceæ), Antidesma (Euphorbiaceæ), Pleiosmilax (Smilaceæ), and
Ruppia (Potameæ).

These seven genera, which with the exception of Ruppia, an aquatic
genus, are only represented in Hawaii by peculiar species, possess in
all cases, except Gouania and the last-named genus, drupaceous or
baccate fruits likely to attract frugivorous birds. Two of them, Eurya
and Antidesma, have their home in Malaya and in the Asiatic continent;
three of them, Gouania, Maba, and Sideroxylon, are found on both the
Asiatic and the American sides of the Pacific Ocean; whilst Pleiosmilax
should, strictly speaking, be regarded as a Polynesian subgenus of
Smilax, a world-ranging genus; and Ruppia is a cosmopolitan brackish-
and salt-water genus.

It is highly probable that Fiji received almost all these genera from
the Old World through Malaya; and in some cases the resemblance between
the Malayan and the Fijian species is so close that, as in Gouania, Dr.
Seemann questioned if they were not forms of the same species. In other
instances, as with Maba, we find a widely-ranging Asiatic and Malayan
species, like Maba buxifolia, extending into Western Polynesia, where it
is accompanied by other species peculiar to that region. But if the
genera were able subsequently to extend their range thence to Hawaii, it
is difficult to understand why they have not reached the Tahitian
region. It is therefore likely that Hawaii received most of these genera
by a northern route and not through the South Pacific; and it is
legitimate to suppose that when Old World genera like Eurya and
Antidesma occur in north-eastern Asia, as in Japan and in the
neighbouring mainland, Hawaii received the genus by that route. In the
case of Eurya it is noteworthy that Fijian and Samoan forms, regarded by
Seemann and Gray as distinct species, are viewed by Reinecke as forms of
E. japonica, an extremely variable species found in Japan. With genera
like Gouania and Maba, that exist on both sides of the Pacific, it is
possible that they may have originally reached Hawaii from America.

A noticeable feature in the instance of genera like Maba and Sideroxylon
is that hard seeds or pyrenes 3/4 to 1 inch (18 to 25 mm.) in length
have seemingly been transported by frugivorous birds across the ocean to
Hawaii. This at first sight seems improbable; but it is known that
fruit-pigeons can swallow very large drupes, as in the case of those of
Canarium, Dracontomelon, and Elæocarpus, afterwards disgorging the
“stones.” They have carried such stones to Fiji, across some 500 or 600
miles of ocean; and unless we impute a continental origin to Hawaii we
must assume that in some cases, as with Elæocarpus, Maba, and
Sideroxylon, they have been able to transport these large stones or
pyrenes to that group. The extent of ocean to be crossed is no doubt
much greater, but this area of the Pacific is not without some small
half-way groups that would serve as resting-places.

That fruits of the order Sapotaceæ are much appreciated by fruit-pigeons
is already known. We learn from Kirk that the fruits of Sideroxylon
costatum (Sapota costata) are a favourite food of the New Zealand
fruit-pigeon, the fruits, about an inch long, containing three hard
crescentic bony seeds nearly as long as the fruit. The natives of Vanua
Levu informed me that a Fijian species of Sideroxylon with hard seeds
about an inch long was much appreciated on account of its fruit by the
pigeons. I found the hard, sound seeds of a species of Sapota,
two-thirds of an inch (or 16 mm.) in size, in the crop of a Fijian
fruit-pigeon. The similarly large seeds of a species of Achras were
identified by Mr. Charles Moore, of Sydney, amongst a collection of
seeds, &c., found by me in the crops of fruit-pigeons shot in the
Solomon Islands (Guppy’s _Solomon Islands_, p. 293). It may be added
that the difficulty concerned with Sideroxylon in Hawaii is the
difficulty concerned with other large-seeded Sapotaceous trees in Fiji
and New Zealand, and the same explanation must be applied to all. Some
further remarks on the Sapotaceæ in the Pacific are given below.

The mode of dispersal of some of these genera is illustrated in other
regions. The berries of Pleiosmilax, a subgenus of Smilax, are well
suited for aiding the dispersal of the genus by frugivorous birds; and
we learn from Prof. Barrows (Weed, p. 42) that in the United States
crows feed on the fruits of Smilax rotundifolia and disperse the seeds.
On the other hand, it is not at first sight easy to understand how a
genus like Gouania has been distributed over the tropics of the globe,
since it possesses dry capsular fruits about half an inch across,
separating into three woody cocci that appear most unlikely to attract
birds. The same difficulty exists, however, with other dry-fruited
widely-ranging genera like Alphitonia and with many of the Euphorbiaceæ.

Amongst these genera found in Hawaii and Fiji to the exclusion of Tahiti
we can at times detect indications of the operations of a polymorphous
species as described in Chapter XXVI., when a widely-ranging highly
variable species is associated in some groups with peculiar species. We
see some evidence of this in the genera Gouania, Maba, and Eurya,
alluded to on a previous page. (See also _Bot. Chall. Exped._, iii. 134,
under “Gouania.”)

One of the mysteries of the Pacific is concerned with the distribution
of the Sapotaceæ, the dispersal of which by frugivorous birds has been
dealt with above. It is strange that whilst the order seems to have
found a _rendezvous_ in Tonga, no one except Horne appears to have
recorded any of the genera from Samoa. They are fairly well represented
in Fiji; but it is in Tonga that we especially note the gathering
together of several Sapotaceous trees with large heavy seeds, of the
genera Bassia, Mimusops, and Sideroxylon. Besides owning one or two
species of Sideroxylon in common with Fiji (Burkill), this small group
possesses Bassia amicorum and Mimusops kauki, both of which were found
there by Forster at the time of Cook’s visit. In a list of a small
collection of plants made by him in Upolu in the Samoan Group about
1879, Horne includes two species of Sideroxylon (_Year in Fiji_, p.
286); and according to Seemann there is a Sapotaceous tree in Wallis
Island. A species of Bassia exists in Rarotonga, the seeds of which,
from Mr. Cheeseman’s description of the fruit, must be almost an inch
long. Drake del Castillo refers to an endemic Tahitian tree near
Mimusops; but its fruit was not known to him.

As already indicated, the difficulties connected with the Sapotaceæ
affect the whole Pacific from New Zealand north to Hawaii and from Fiji
east to Tahiti. We are driven to appeal to the agency of frugivorous
birds, at least in the case of Sideroxylon, since some fruits
experimented on by me in Fiji sank at once or in a day or two, the seeds
having no buoyancy. That birds actually disperse the seeds of this and
other genera of the order has been already pointed out, yet it is
possible that currents have at times aided in the dispersal of some of
the genera. This is indicated by the circumstance that, as we learn from
Schimper, some Sapotaceous trees are to be included in the Malayan
strand-flora, namely, Sideroxylon ferrugineum, Mimusops kauki, and M.
littoralis, all occurring as well on the Asiatic mainland, the first
growing also in the Liukiu Islands, and the last in the Andaman and
Nicobar Groups.

_Ruppia maritima_ (Potameæ).—This cosmopolitan aquatic plant has only
been recorded in Polynesia from Hawaii, Samoa, and Fiji. It had not been
collected in Fiji before my discovery of it in 1897. Amongst other
oceanic islands where it occurs may be mentioned the Bermudas, where,
according to Hemsley, it exists as an indigenous plant in the lagoons.
Chamisso first noticed it in Hawaii, and Hillebrand remarks that it
grows in shallow waters along the coasts. Amongst other localities where
I noticed it in this group may be mentioned the north-west coast of the
large island of Hawaii between Kailua and Keahole Point. Here in 1896 it
was thriving in brackish-water ponds, with Sesuvium portulacastrum
growing at the edges. Reinecke observes that it occurs in similar ponds
in Samoa. In 1897 I found it in abundance in the Rewa estuary (Fiji),
both in the creeks and in the main channel. In the following year it was
not to be found in this locality, a circumstance noticed both by the
natives and by resident whites. The fruits of this plant possess no
floating power, sinking, even after prolonged drying, in a few hours. It
is to ducks and to birds of similar habit that its dispersal must be
attributed.

THE ABSENTEES FROM HAWAII.

It has been before remarked that of the 330 or 340 genera of
flowering-plants recorded from Fiji some 200 are not known in Hawaii. It
will only be possible to deal with the absent genera in a cursory
manner; but enough will be done to show that we are face to face here
with a multitude of the seeming inconsistencies that so often beset the
study of plant-distribution.

A host of plants are unrepresented in Hawaii, of which it may be said
that their seeds or fruits are not less suited for being carried across
the Pacific than those of many that are now in that group. On the other
hand, a number of genera exist there which we should never expect to
have been endowed with the capacity, and to have received the
opportunity, of crossing nearly 2,000 miles of ocean. Yet perhaps when
Nature acts in a wholesale fashion and excludes entire orders we may be
able to perceive the dim outlines of a principle of exclusion at work.
But even here much caution and some clearing of the ground are needed.

For example, having regard to the several modes of dispersal possessed
by the great variety of fruits and seeds of the Sterculiaceæ, it would
be almost meaningless to remark that the order so well represented in
Fiji is practically non-existent in Hawaii as far as truly indigenous
plants are concerned. It is true that two species of Waltheria are here
present, but one of them W. americana, is a weed probably introduced by
the aborigines whilst the other, W. pyrolæfolia, recorded from a
solitary locality by the Wilkes Expedition, has seemingly never been
found since. From the standpoint of dispersal the genera Sterculia,
Heritiera, Kleinhovia, Melochia, and Commersonia, that are represented
in Fiji but not in Hawaii, cannot be discussed together. With Sterculia
is concerned the dispersal by birds of large seeds, an inch in length,
not particularly well protected, the genus being confined to Fiji alone
of all the oceanic Pacific groups. Heritiera is only represented by a
littoral species, the large fruits of which are carried great distances
by the currents; and no other agency of dispersal is here possible. The
last three genera are distributed over the South Pacific, their
relatively small seeds being probably in the main dispersed by
granivorous birds; whilst the setose fruits of Commersonia may have been
at times transported in birds’ plumage.

It is more legitimate, perhaps, to speak collectively of the orders
Meliaceæ and Melastomaceæ as absent from Hawaii; but even here the issue
raised is one concerned rather with opportunities than with capacities
for dispersal. Several years ago, M. Casimir de Candolle remarked that
“it is hardly credible that the Meliaceæ should be entirely absent from
the Sandwich archipelago” (_Trans. Linn. Soc. Bot._, vol. i. 1880). Yet
it can scarcely be said that this is a matter connected with means of
dispersal. Amongst the Meliaceous genera represented in Fiji, Vavæa and
Aglaia have a berry, Melia has a drupe, and Dysoxylum has a capsule. So
again with the Melastomaceæ; it possesses at least six genera in Fiji,
two in Tahiti, and none in Hawaii. Whilst the genera Melastoma and
Medinilla have baccate fruits with minute seeds, Astronia has a capsule
with similar seeds, and Memecylon has a single-seeded berry. Since,
however, minute seeds are most typical of the order, those of Melastoma
denticulatum being about one-fiftieth of an inch or ·5 mm. in size, it
would seem that this character has not aided its dispersal in the
Pacific so far as Hawaii is concerned. From the circumstance that
berries, drupes, and capsules are represented in these two Fijian orders
we may form the opinion that their non-occurrence in Hawaii is due not
so much to lack of capacities for dispersal as to failure of
opportunities.

This opinion is much strengthened when we come to deal with the
individual genera, where the predominant cause of the absence of so many
Fijian genera from Hawaii is concerned with the failure of the agencies
of dispersal. It is not a question of a difference in size between the
groups, since, although the surface-area is approximately the same in
both groups, Hawaii possesses only two-thirds of the number of genera
occurring in Fiji. It is not a question of capacity for dispersal across
an ocean, since birds have transported across the Pacific to Hawaii the
“stones” and large seeds of genera like Elæocarpus and Sideroxylon, a
feat that would have been deemed impossible by many botanists. It is no
lack of capacity for dispersal that has excluded Loranthus from Hawaii
and has admitted Viscum.

Few genera, indeed, would seem to be better fitted for dispersal by
frugivorous birds in the Pacific than that of Ficus. Its fruits are
known to be eaten by birds all over the area of the genus; and we find
the species distributed over the South Pacific from Fiji to Tahiti, but
they are quite absent from Hawaii. This is the more remarkable on
account of the occurrence of a species of Ficus resembling a banyan in
Fanning Island about 900 miles south of the group (_Bot. Chall. Exped._,
iii. 116, 194), and because the Hawaiian Islands possess the Meliphagidæ
or Honey-eaters, which are widely distributed in Polynesia and are known
to feed on these fruits—a matter further discussed in my treatment of
Ficus later on in this chapter.

Of several Rubiaceous genera with fleshy fruits that are represented
both in Fiji and Tahiti, such as Stylocoryne and others, and of those
Rubiaceous genera with minute seeds that, like Ophiorrhiza, are
distributed over the South Pacific, none occur in Hawaii. Here we find
represented other genera of the order, like Gardenia, Plectronia, and
Coprosma, that do not appear to be better fitted for dispersal by
frugivorous birds than many of the genera not existing there. If birds
have carried to Hawaii in their plumage the fruits of Pisonia and
Sicyos, it cannot be merely a question of capacity for dispersal that is
concerned with the restriction to the South Pacific of genera with hairy
seeds, such as Trichospermum, Alstonia, and Hoya.

It is unnecessary to dwell longer here on the subject of the Hawaiian
absentee-genera, since many of the absent plants will be discussed when
dealing with the peculiarities of the Fijian flora. The data there given
all go to show that mere lack of capacity for dispersal over the Pacific
often counts for little in supplying us with an explanation of the
absence of so many likely genera from the Hawaiian flora. Hawaii has
only been stocked with those genera common to Fiji and Tahiti that could
have reached it during each age of general dispersal over the Pacific.
In later eras the dispersing agencies have been mainly active in the
tropical South Pacific; and thus it is that, as will be pointed out in a
later page, the bulk of the plants of the Malayan era are confined to
the region between Fiji and Tahiti. In a still later period the
dispersing agencies have confined their operations mainly to Western
Polynesia and the last immigrant genera have not reached beyond the
Fijian region.

The whole story of plant-life in the tropical Pacific is bound up with
these successive stages of decreasing activity of the dispersing
agencies. The story of plant-distribution in this region is well
illustrated in its earlier phases of general dispersion in the floral
history of Hawaii, in its later phase by those Asiatic genera that have
only crossed the South Pacific to Tahiti, and in its last phase by those
genera that have never extended beyond the groups of the Fijian area.
The area of active dispersion, that first comprised the whole of the
tropical Pacific, was afterwards restricted to the South Pacific, and
finally to the western portion of that area. It can scarcely be doubted
that these successive stages in the contraction of the area of active
dispersion of plants in the Pacific were accompanied by a corresponding
diminution in the general distribution of birds in the same ocean, to
which it stood in the relation of an effect to a cause.

TAHITI.

The peculiarities of the Tahitian flora as compared with Hawaii and Fiji
may be discussed by treating first those genera that are alone
represented in Tahiti, the “residual” genera; then those that it
possesses in common first with Hawaii and then with Fiji; and lastly by
pointing out the more noticeable gaps in the flora. By Tahiti is
typically signified the whole Tahitian region, which includes the
Austral and Cook Groups, the Society Islands, the Paumotus, and the
Marquesas.

THE TAHITIAN RESIDUAL GENERA.

The non-endemic genera occurring alone in the Tahitian region and not
found either in Hawaii or in one or other of the three groups of the
Fijian region (Fiji, Tonga, Samoa) are not more than half a dozen. These
six genera are exceedingly interesting; but since each tells a different
story and gives its own independent indication they cannot be treated in
a collective sense. Nor are they all to be regarded as anomalies in
plant-distribution, since with a single exception there is scarcely one
concerning which it is not in some way possible to give an explanation
of its isolation without coming into conflict with the principles of
plant-dispersal. The exception is Lepinia tahitensis, which, without
presenting any very evident capacity for dispersal, has not been
recorded from any other localities in the Pacific than the far-separated
Solomon and Tahitian Groups. There is a suspicion that, as in the case
of the residual genera of Hawaii, America may have contributed some of
the original plants, since three of the genera, Buttneria, Coriaria, and
Bidens, occur in that continent, and in the case of Coriaria Tahiti
possesses a species found in South America as well as in New Zealand.

One of the trees in question is Cratæva religiosa, an Asiatic species,
which may be placed among a group of trees, including Cananga odorata
and Fagræa Berteriana, which, whilst they are much esteemed by the
inhabitants of the South Pacific for their fruits or their flowers, and
are often planted in and around their villages, possess fruits that
attract birds, and in the case of Cananga are known to be dispersed by
fruit-pigeons. Probably the aborigines and the birds have worked
together in the distribution of these trees.

The genera Buttneria of the Sterculiaceæ and Berrya of the Tiliaceæ are
represented in this region by species that must owe their dispersal to
birds, though I have no data relating to the matter of their dispersal,
their fruits being capsular, in the first case prickly. Coriaria is a
mountain genus in Tahiti and will be found discussed in Chapter XXIV. in
connection with the Tahitian mountain-flora. Its absence from the West
Polynesian groups is no doubt to be connected with their insufficient
altitude. In addition to the introduced Bidens pilosa, a common tropical
weed, Tahiti possesses two other truly indigenous species of Bidens, of
which one at least is peculiar to the region. The achenes of this genus
are well known to be adapted for dispersal in a bird’s feathers; and
since the genus has its principal home in America, no other indigenous
species having been recorded from South Polynesia, it is not unlikely
that the parent species was American.

One of the numerous enigmas of the Pacific floras is concerned with the
presence in the islands of Tahiti and Moorea (Eimeo), in the Society
Group, of the Apocynaceous tree, Lepinia tahitensis. The genus contains
this solitary species, which has been collected only in one other
locality, namely, in the Solomon Group, where it was obtained by the
Rev. R. B. Comins. Such an instance of disconnected distribution is rare
in the Pacific Islands, and undoubtedly it represents one of the
difficulties of the Tahitian flora. The fruits, which are indehiscent
and five or six inches in length, possess a fibrous pericarp and a
single seed. No data are to hand relating to the capacities for
dispersal possessed by this plant, but it is certain that it has had
some means of crossing the sea between the adjacent islands of Tahiti
and Moorea. (See Hemsley, _Journ. Linn. Soc. Bot._, xxx. 165.)

TAHITIAN GENERA FOUND IN HAWAII TO THE EXCLUSION OF FIJI.

This subject has been already discussed in this chapter in dealing with
the genera restricted to Hawaii and Tahiti.

TAHITIAN GENERA FOUND IN FIJI TO THE EXCLUSION OF HAWAII.

Excluding the orchids, sedges, and grasses, as well as the few endemic
genera, between sixty and seventy genera, or rather less than half of
the genera of the flowering-plants of Tahiti, are found in Fiji to the
exclusion of Hawaii. Of these, rather over a half are Old World genera;
about a third occur in both the Old and the New World; four are confined
to Polynesia, and not one is exclusively American. One-third are genera
now possessing in the Tahitian region endemic species either entirely or
in part, and in such cases we may consider that the agencies of
dispersal are now inactive or partially suspended; the others belong
entirely to the present era of dispersal. About half have more or less
fleshy fruits fitted for dispersal by frugivorous birds. About a fourth
have capsular or other dry fruits that must have been also dispersed by
birds preferring a drier diet. Three only possess hairy seeds or fruits
suitable for being carried in a bird’s plumage, namely, Commersonia,
Weinmannia, and Alstonia. There remain about a fourth of the total that
are shore-plants dispersed by the currents, being in two cases (Ximenia
and Kleinhovia) assisted by birds; whilst Triumfetta, another littoral
genus, is probably distributed by birds alone.

There are no cases of special difficulty from the standpoint of
dispersal in these sixty and odd non-endemic genera that Tahiti
possesses in common with Fiji to the exclusion of Hawaii. The lack of
difficulties connected with the dispersal of all these Tahitian genera
is worthy of note, because there are very few difficult genera amongst
the rest of the Tahitian flora. Excluding Lepinia tahitensis, which has
been already referred to, there are scarcely any “impossible” plants in
the Tahitian region; and even in this case, when the modes of dispersal
of Lepinia come to be investigated, it is likely that much of the
difficulty will disappear. Hawaii, as we have before seen, abounds with
perplexing questions of this nature. When dealing with the absentee
Tahitian genera, later on in this chapter, it will be shown that “size”
has played a prominent determining part in the exclusion of genera from
Tahiti, genera with seeds or “stones” exceeding half an inch or twelve
millimetres in dimension being, as a rule, unrepresented amongst the
truly indigenous plants.

My further remarks on these Tahitian genera found in Fiji but not in
Hawaii will be limited to some general observations from the standpoint
of dispersal. I will first discuss some of those genera that possess
only peculiar species. They belong to an era of dispersal that, as far
as Tahiti is concerned, is passing or has passed away. Here we have the
species of each genus more or less localised in the various South
Pacific archipelagoes; but, as with Meryta, Alstonia, and Loranthus, it
is often apparent that, although the Tahitian region is mainly outside
the zone of present dispersal, the different groups of the Western
Pacific are kept in touch by the possession of species in common. This
testifies to the activity of dispersal in that region after it had
become suspended in Eastern Polynesia. The connection between the
isolated endemic species of Eastern Polynesia and a species ranging over
the Western Pacific can sometimes be shown, as in the case of Loranthus,
where a species confined to the Society Islands and to the Marquesas is
very closely related to L. insularum, a widely-ranging West Polynesian
species that reaches eastward as far as Rarotonga.

We next have those genera like Grewia, Nelitris, Melastoma, Randia,
Geniostoma, Tabernæmontana, Fagræa, Bischoffia, Macaranga, and Ficus,
that possess in Polynesia one or more widely-ranging species. The agency
of the polymorphous species, which I have described in the preceding
chapter in connection with the general dispersal of Malayan plants over
the whole of Polynesia, is evidently also active when the work of
dispersal is restricted to the South Pacific. Its operation is to be
distinctly traced in all the genera above named except in Fagræa and
Ficus. Thus, in the genera Grewia, Melastoma, Randia, Geniostoma, and
Macaranga we find a single variable species ranging over the South
Pacific from Fiji to Tahiti, keeping all the groups in touch, but
associated in each, as a rule, with one or more peculiar species. A yet
earlier stage in the process is to be seen in those genera which, like
Nelitris, Tabernæmontana, and Bischoffia, possess only a solitary
species ranging over the South Pacific, varying in each group, but not
usually associated with endemic species. As with Melastoma, Macaranga,
and others, we can often trace the widely-ranging species of Polynesia
back to its home in Malaya, and with these and other genera the
connection between a species confined to a group and a variable species
ranging through all the archipelagoes of the South Pacific can sometimes
be detected in the affinity of their characters.

It is thus seen that one of the principal determining causes of the
differentiation of species in Polynesia lies in the failure of the
dispersing agencies, a widely-ranging species becoming in consequence
gradually isolated in the various groups. With some genera, as with
Ophiorrhiza, it is possible to show that the resulting endemic species
pass into each other by intermediate forms.

My further remarks on the Tahitian genera occurring in Fiji but not in
Hawaii will be devoted mainly to those with which I was most familiar
from the standpoint of dispersal.

The Tiliaceous genus GREWIA offers a good example of those Polynesian
genera which possess in the South Pacific a single widely-ranging
species associated often with endemic species in the individual groups.
It is likely that a polymorphous form, including most of the Polynesian
species, could be here constituted. The fruits are dryish drupes,
becoming black and moist when over-ripe, and containing three or four
pyrenes suitable for distribution by birds and five or six millimetres
in size.

The berries of NELITRIS, a genus of the Myrtaceæ, contain a few hard
seeds that are well fitted for dispersal by frugivorous birds. I am
inclined to follow Drake del Castillo, who considers that there is only
one varying species, N. vitiensis (Gray), which is distributed over the
whole of the South Pacific from the Solomon Islands to Tahiti. The
tendency of this widely-ranging species to vary in different groups is
indicated in the fact that some botanists have distinguished other
species within these limits. It is noteworthy that N. paniculata in
Indo-Malaya and N. vitiensis in the Pacific cover the whole range of the
genus. It would be interesting to establish a connection between them.

MELASTOMA, an Old World genus of forty and more species, has one very
variable species, M. denticulatum, which, as defined by Bentham, has the
range of the genus from tropical Asia across the Pacific to Tahiti. This
plant is associated in some groups, as in Fiji, Tonga, and Samoa, with
other more or less localised species, and it affords a good example of
the principle of polymorphism in species-making. The berry-like fruits
contain an abundance of minute seeds, half a millimetre in size, which,
when rendered adhesive by adherent pulp, might readily stick to
feathers, or they might pass unharmed through the alimentary canal of a
bird. It is noteworthy that amongst the plants regarded by Prof. Penzig
as introduced by frugivorous birds into Krakatoa since the eruption is a
species of Melastoma.

Few genera in these islands would better repay a careful study of their
species with regard both to the influence of station on specific
characters and to the question of “mutations” than OPHIORRHIZA. I found
the three Fijian species of this Rubiaceous genus so often in close
association, that I cannot doubt there is some connection between them.
Seemann and Gray, indeed, characterise two of them as confluent species.
The island of Tahiti alone possesses five peculiar species, and it is
evident that this island has been a centre of development for species of
Ophiorrhiza, just as Samoa has become the birthplace of many species of
the Urticaceous genus Elatostema. The minute angular seeds of
Ophiorrhiza would probably be transported in a bird’s feathers or in
adherent soil. As my experiments showed, they do not become adhesive
when wet.

The genus LORANTHUS as distributed in the South Pacific has already been
briefly noticed. There is a species confined to the Tahitian region, and
there is another peculiar to Samoa, whilst one widely-ranging species,
L. insularum, that connects these regions together, reaching east to
Rarotonga, is closely related with the Tahitian species. There was no
doubt originally a single polymorphous plant that ranged over the
tropical South Pacific. With regard to the mode of dispersal of the
seeds of this genus of parasites, I should at once refer to the
systematic and careful observations made by Mr. F. W. Keeble in Ceylon
(_Trans. Linn. Soc._, v. 1895-1901). He formed the opinion that the
seeds of Loranthus usually reach their host without passing through the
alimentary canal of a bird, being merely wiped off its bill. This method
would never carry the seeds to Tahiti or even to Fiji; and since this
observer remarks that, although most of the seeds in the droppings were
completely rotten, some of them “possibly pass through the gut
uninjured,” we may accept this possibility as sufficient for the purpose
of dispersal in the Pacific Ocean. Mr. Keeble notes the observation in
Teil 3 of Engler’s _Die Natürlichen Pflanzenfamilien_ that the seeds may
germinate after passing through a bird’s intestine; and we may therefore
infer that whilst the method he describes is typical of local dispersal,
the other method is required in the instance of oceanic dispersal.

ALSTONIA, an Apocynaceous genus of tropical Asia and Australia, yields
the caoutchouc of Fiji. Besides possessing in Fiji and Samoa peculiar
species, the islands of Western Polynesia have in A. plumosa a species
common to Fiji, Samoa, and New Caledonia. Another species, A. costata,
is restricted to Eastern Polynesia, occurring in the different islands
of the Tahitian Group as well as in Rarotonga. It is possible that the
Pacific species may be connected with A. scholaris, a species possessing
the range of the genus with the exception of Polynesia. The long
ciliated or hairy seeds, six to nine millimetres in length, are fitted
for transport by the winds and in birds’ plumage. The follicles dehisce
on the tree, and, according to Horne, the light seeds are distributed
locally by the wind. It is probable that the thick white juice oozing
from a broken branch would at times aid the adhesion of the seeds to a
bird’s feathers.

GENIOSTOMA, a genus of the Loganiaceæ, is found in Malaya, Australia,
and New Zealand. It possesses in G. rupestre a species that ranges
across the South Pacific from New Caledonia to Tahiti, being associated
with one or more endemic species in most of the groups. The fruit is a
dehiscent capsule containing numerous small seeds imbedded in a
yellowish pulp; and from the standpoint of dispersal it may be placed in
the same category with Pittosporum and Gardenia (see pages 310, 313).

The same principle involved in the occurrence of a species ranging the
South Pacific from New Caledonia to Tahiti, and associated with one or
more endemic species in most of the principal groups, is illustrated in
the Euphorbiaceous genus MACARANGA. It is specially noteworthy that M.
tanarius, which ranges from India to East Australia and the New
Hebrides, comes in touch in the group just named with M. harveyana, the
widely-ranging plant of the South Pacific above alluded to, and itself
an Asiatic species (see Burkill; _Bot. Chall. Exped._, iii. 191; _Index
Kewensis_). The connection between M. harveyana, the widely-ranging
species of the South Pacific, and the endemic species in the various
groups is indicated by its affinity with M. reineckei, a Samoan species.
The Macarangas in Fiji grow in a variety of situations, on the borders
of estuaries, in the mountain forests, and on the isolated mountain
peaks. It is to birds that we must look for the dispersal of the genus.
In the case of a species, apparently M. seemanni, common in the Rewa
delta, the seeds, which soon fall out of the cocci, are not infrequently
found in the drift of the estuary, but they sink in a week or two. Other
species examined showed no capacity for dispersal by currents. The fruit
of M. harveyana is provided with a few prickles, but since it breaks up
into the cocci, from which the seeds soon fall out, these appendages
could scarcely aid its dispersal.

Like many other genera, TABERNÆMONTANA, an Apocynaceous genus
distributed through the tropics, is represented in Polynesia by a
widely-ranging species, T. orientalis, which extends from Malaya and
Eastern Australia through all the large groups of the South Pacific from
the New Hebrides to Tahiti, and is associated in Fiji with one or two
peculiar species, one of which, according to Mr. Burkill, is nearly
related to it. This genus therefore seems to illustrate the earliest
stage in the Pacific of that process by which a widely-ranging species
takes on a polymorphous habit and through its variations gives rise to
different species in various groups. Prof. Schimper ranks T. orientalis
amongst the Malayan strand-flora; but in Fiji the Tabernæmontanas are
only littoral where the soil is rich as in alluvial regions; and they
have no capacity for dispersal by currents that is worth speaking of,
the seeds in the case of T. orientalis and another species sinking after
drying for years, whilst the follicles soon open in water and go to the
bottom in a few days. The observations of Gaudichaud and Moseley
indicate that some Malayan species are dispersed locally by the currents
(_Bot. Chall. Exped._, iii, 279, 293); but the fruits of the genus are
evidently quite unfit for oceanic dispersal by this agency. We find in
the bird the agent that has carried the genus to the distant
island-groups of the Pacific; and from the standpoint of dispersal the
fruits may be placed with those of Pittosporum and Gardenia, being
follicular, and in the Fijian plants possessing seeds, 5 to 10
millimetres in size, embedded in a pulp.

FAGRÆA, an Asiatic and Malayan genus of the Loganiaceæ, is represented
in the Pacific by F. berteriana ranging through all the groups and
islands of the South Pacific from the Solomon Islands and New Caledonia
to Tahiti and the Marquesas, and by one or two other species in Fiji. It
is with Fagræa berteriana that we are entirely concerned. The tree is
often planted by the Pacific islanders near their villages; and since
they value its timber and use its large fragrant flowers for personal
decoration and for other purposes, it is probable that they have aided
in its dispersal. But, as shown below, it behaves in most localities as
an indigenous plant; and its berries are well fitted for promoting its
dispersal by frugivorous birds.

I was familiar with Fagræa berteriana both in the Solomon Islands and in
Fiji; and in the last-named locality I especially studied it from the
standpoint of dispersal. All over the South Pacific, whether in the
Solomon Islands, in Fiji, in Rarotonga, or in Tahiti, this tree, though
thriving also in the lower levels, especially frequents rocky scantily
vegetated or open-wooded hill-tops and crests up to 2,000 or 2,500 feet
above the sea. In the rich alluvial soil of the Rewa delta in Fiji it
attains a height of 25 or 30 feet or more, whilst in the poor, dry soil
of the “talasinga” plains in this group it is much dwarfed, and often
does not exceed 10 feet, and may be only 6 feet high. It is in these
“talasinga,” or “sun-burnt,” plains of Fiji, especially in the Mbua
province of Vanua Levu, that the tree, although dwarfed, seems most at
home. Here it flowers and fruits abundantly whilst associated with
Acacia, Casuarina, and Pandanus trees, and it is in such dry localities
that this tree reflects in its choice of station the behaviour of
different species of the genus in the Malay Peninsula, where they grow
in open heath-country and sometimes on sandy heaths (Ridley in _Trans.
Linn. Soc. Bot._, iii, 1888-94). The fruits and seeds of F. berteriana
have little or no capacity for dispersal by currents. On the Fijian
plains the berries partially wither and rot on the tree. In the western
part of its area this tree almost comes in touch with the Asiatic
species, F. obovata, that ranges from India and Ceylon to the Malayan
region, a species that must be indebted to frugivorous birds for its
wide distribution.

The Euphorbiaceous genus BISCHOFFIA seems to offer another example of
polymorphism in a wide-ranging species. Following Drake del Castillo, I
take the genus as including only a single species, B. Javanica, a tree
distributed over tropical Asia, Malaya, and Polynesia as far east as
Tahiti. The variable character of the species is indicated by the
different views held by the several botanists who have discussed the
South Pacific species. Whilst it is a common forest-tree in Indo-Malaya,
it affects in the Pacific islands the open-wooded districts of the lower
levels, and it is not uncommon on the dry “talasinga” plains of Fiji.
The fruits and seeds displayed in my experiments little or no capacity
for dispersal by currents; nor do these dryish berries, with seeds four
or five millimetres long, seem to be especially attractive for
fruit-eating birds; and it is likely that the same birds that distribute
Macaranga seeds also disperse those of this genus. The tree bears the
same name over the South Pacific, “koka” in Fiji and Rarotonga, and “oa”
in Samoa. Like many other Polynesian trees, it has its uses, but there
is no reason to believe that the natives have aided materially in its
dispersal.

FICUS, a large genus comprising several hundred species, attains its
greatest development in tropical Asia and in Malaya. It is well
represented in the Western Pacific from the Solomon Islands to Fiji and
Samoa; but in Eastern Polynesia the species are very few, and the genus
is altogether absent from Hawaii, although a species has been found in
the North Pacific in Fanning Island, about 900 miles south of the
Hawaiian group (see page 377).

The Polynesian species are for the most part restricted to the Pacific
islands, but there are only two species that range over the South
Pacific as far east as Tahiti, namely, Ficus prolixa, the Tahitian
banyan, and F. tinctoria. Some species are confined to Western
Polynesia, such as F. obliqua, the Fijian banyan, F. scabra, and F.
aspera, the last occurring in East Australia. Among the individual
groups Fiji possesses probably fourteen or fifteen species, of which,
perhaps, a third would be peculiar. According to Dr. Warburg, as cited
in Dr. Reinecke’s paper, Samoa owns eight species, of which six may be
endemic. In Rarotonga and Tahiti we find only F. prolixa and F.
tinctoria. The species in the groups where they are best represented
belong to three or four sections of the genus.

The banyans of the South Pacific are represented by three or four
species, namely, Ficus prolixa, the Tahitian banyan, found all over the
tropical groups of the South Pacific from the New Hebrides and New
Caledonia to Tahiti, the Marquesas and Pitcairn Island (Maiden); F.
obliqua, the Fijian banyan, confined to the islands of the Western
Pacific from the New Hebrides to Tonga; and two new banyans in Samoa, as
described by Dr. Warburg in Dr. Reinecke’s paper. In my paper on
Polynesian plant-names it is shown that the banyans possess two names in
the Pacific, one being “aoa,” the Polynesian name, found in all the
groups from Samoa eastward, and connected linguistically with the
Malayan and Malagasy banyan-words; the other, the Melanesian name
typified in the Fijian “mbaka,” and represented in a variety of forms in
the New Hebrides and neighbouring groups.

It is probable that the Pacific islanders have assisted in the dispersal
of one or two of the species of Ficus, such as F. tinctoria, which they
employ for different purposes, but, generally speaking, birds are active
agents in distributing the genus. I need scarcely say that the agency of
the currents is quite insufficient to explain the distribution of Ficus.
When in Fiji I experimented on three or four different species of Ficus
belonging to the sections of the genus there represented. The fruits may
float at first, but within a week or ten days they break down, and the
seeds escape and sink. Beneath a tree of F. scabra growing on the banks
of the Wai Tonga in Viti Levu, I noticed a number of its fruits floating
in a sodden condition among the reeds at the river-side.

It is with the banyans that the dispersal of the seeds by frugivorous
birds becomes most evident. This is at once indicated by the frequent
occurrence of these trees in the interior of coral islets in the Western
Pacific, as in Fiji and in the Solomon Islands. Fruit-pigeons roost in
their branches, and birds shot on these islets often contain the fruits
in their crops (_Bot. Chall. Exped._, iv, 310). The process may also be
seen in operation in Krakatoa. Professor Penzig found in 1897 that three
species of Ficus had established themselves there since the eruption of
1883 through the agency of frugivorous birds. Besides pigeons, we find
that parrots, hornbills, honey-eaters, &c., feed on these fruits, and I
possess a large number of references to this subject. The Messrs. Layard
in New Caledonia, Dr. Meyer in Celebes, Mr. Everett in Borneo, Dr.
Forbes in Sumatra, and several other contributors to _Ibis_ might be
here mentioned. Dr. Beccari, in his _Wanderings in the Great Forests of
Borneo_, speaks of “the facile dissemination of the various species of
Ficus through the agency of birds,” and he arrives at certain important
conclusions which are discussed in Chapter XXXIII.

I have before alluded to the absence of Ficus from Hawaii. This group
possesses the Honey-Eaters (Meliphagidæ), birds well suited for
dispersing species of Ficus over Polynesia; but this family of birds is
only represented by peculiar genera in Hawaii, and therein lies the
explanation. At the time when the Honey-Eaters roamed over Polynesia,
the genus Ficus had not arrived from Malaya. The connection between the
bird and the plant is well shown on Fernando Noronha, which possesses a
peculiar species of Ficus and a peculiar species of dove, the only
fruit-eating bird in the island (Ridley).

THE ABSENTEES FROM TAHITI

Generally speaking, all the “difficult” genera which puzzle the student
of plant-dispersal in Fiji and Hawaii are absent from the Tahitian
region. Those with stone-fruits and with large seeds, where the stone or
seed is an inch in size and over, are absent from Tahiti. Thus the
genera Canarium, Dracontomelon, Myristica, Sterculia, and others, of
which the three first-named are known to be dispersed by fruit-pigeons,
have not advanced into the Pacific eastward of the Fijian region. We
miss in the Tahitian islands the large-fruited palms of Fiji, such as
the Veitchias with fruits two to two and a half inches (5 to 6 cm.)
long, and we find in their place a Ptychosperma, evidently very rare,
and the widely spread Pritchardia pacifica, that may have been
introduced by man, both with drupes not far exceeding half an inch (1·2
cm.) in size. The islands of the Tahitian region also lack the Coniferæ;
and genera like Dammara, Dacrydium, and Podocarpus that give such a
character to the Fijian forests are not to be found. In this region we
do not find many of the large-seeded Leguminous genera, such as
Cynometra, Storckiella, and Afzelia, that occur in Fiji, the only
large-seeded genera that it possesses being such as are brought by the
currents, namely, Mucuna, Strongylodon, Cæsalpinia. The difficulties
presented by the occurrence of the inland species of Canavalia and
Mezoneuron in Hawaii do not offer themselves in Tahiti (see Chapter XV).
Tahiti also lacks, as often before observed, the mangroves and most of
the plants of the mangrove-formation.

As above remarked, the Fijian trees with large “stones” and heavy seeds
an inch in size are not to be reckoned amongst the indigenous Tahitian
plants, “size” being an important determining factor in the exclusion.
The occurrence of Elæocarpus in Rarotonga presents no real difficulty,
as I have explained in Chapter XXVI. An apparent exception is presented
by the existence in Tahiti of Calophyllum spectabile, where the stones
are about an inch across; but since its fruits can float in sea-water
for nearly a month, and on account of the value placed on its timber by
the Polynesians, we cannot altogether exclude the agencies of man and
the currents. One seeming exception is also offered by the presence of
Serianthes myriadenia, a tree which in Fiji grows both in the forests
and on the banks of the tidal estuaries. Its seeds, which are six to
seven-tenths of an inch (15 to 18 mm.) in length, have no buoyancy, and
the pods float only two or three weeks. The case of Lepinia tahitensis
is alluded to elsewhere, but it may be added that these and other
difficulties await further investigation.

A great many Fijian plants are not found in the Tahitian region, such as
Micromelum, those of the order Meliaceæ, the Melastomaceous genus
Medinilla, Myrmecodia, Ophiorrhiza, &c., which are often quite as well
fitted for over-sea transport as are several of the plants already
established there. But it should be remembered that crowding out would
often come into play in such a contracted region. The area, however, has
been very generously dealt with as regards plant genera. Though the
total land-surface cannot be more than one-fourth or one-third that of
Fiji or Hawaii, it possesses more than half the number of genera found
in Fiji, and four-fifths of the number found in Hawaii.

FIJI

_The Fijian Genera not found in either the Tahitian or Hawaiian Regions_

We have already in some degree dealt with Fiji in so far as the partial
dispersal of genera over the Pacific islands is concerned. We have seen
that it possesses very few genera (not a score in all) in common with
Hawaii that are not found in the Tahitian region, and it is assumed that
in most cases such genera reached Hawaii independently and not through
the South Pacific. On the other hand, excluding the grasses, sedges, and
vascular cryptogams, Fiji owns in common with Tahiti between sixty and
seventy genera that do not occur in Hawaii. This shows unmistakably the
trend of plant migration in the Pacific islands. Several interesting
features in plant-distribution have been already brought out, and
notably the fact that Indo-Malayan genera with large seeds or “stones”
an inch in size have been arrested in the Fijian region in their passage
into the South Pacific. Thus Canarium, Dracontomelon, Myristica, and
Sterculia have not extended eastward of the Fijian area.

Yet a very large proportion of the Fijian genera, quite half of the
total number, are not represented either in the Tahitian or in the
Hawaiian region; and of many of them it is obvious that they are as well
fitted to be carried over the Pacific as are those that have actually
reached Tahiti and Hawaii. Take, for instance, Begonia, which has not
extended east of Fiji, though Hillebrandia, a genus of the order, is
peculiar to Hawaii. Nor can we explain why with three genera like
Geissois, Dolicholobium, and Alstonia, possessing seeds dispersed by the
winds, only the last-named has passed beyond Fiji. However, as before
remarked, it is probable that lack of opportunity rather than capacity
for dispersal has determined the matter, and we must, therefore, assume
that many of the genera have halted in the Fijian region because they
entered the Pacific after the age of active general dispersal over that
ocean.

Occasionally we notice in this region that which we have observed in the
case of Cyrtandra in different Pacific groups, namely, a sudden
development of what Hillebrand terms “formative energy” in a genus, such
as we find in the case of Elatostema in Samoa, and in that of Psychotria
in Fiji and Samoa. The principle of polymorphism in the development of
species is also illustrated by Micromelum and by Limnanthemum. In the
last case we possess a typical polymorphous species in Limnanthemum
indicum that has played in this respect the _rôle_ of Naias marina in
the warm waters of the globe.

With several genera that like Gnetum, Myristica, and Sterculia occur
both in the Old and the New World, it is evident that in explaining
their distribution we are dealing with something more than questions of
means of dispersal. With these genera, and with others like Lindenia, it
seems almost futile to talk of means of dispersal, when to all
appearance their existing distribution is but the remnant of an age of
general dispersion over the greater part of the warm regions of the
world. These genera, with others, might be cited in favour of the
continental hypothesis relating to the islands of the Western Pacific.
Trees with stone-fruits, such as Canarium, Couthovia, Dracontomelon, and
Veitchia, where the stones are an inch and more in length, might be also
adduced by some in evidence of this theory. But in these cases the
lesson of Elæocarpus (Chapter XXVI) should always be remembered, since
the “stones” of drupes may vary greatly in size amongst the different
species of a genus, and species seemingly “impossible” from the
standpoint of dispersal in one group may be represented in other groups
by species where the size of the “stone” presents no difficulty in
attributing the dispersal of the genus to frugivorous birds.

_Sterculia_

The problem connected with the presence of this genus in Fiji is but a
part of the still more difficult problem connected with the dispersal of
the genus over the tropics. The riddle presented by the Fijian species
seems, indeed, difficult enough; but it merely presents in miniature the
great mystery surrounding the whole genus. According to the _Index
Kewensis_ no other species have been found in oceanic islands except
those occurring in the Western Pacific, as in Fiji, the New Hebrides,
and New Caledonia, and most of these seem to be confined to those
islands. We have here a genus that repeats the Dammara difficulty of the
Western Pacific.

The trees are common in places in the Vanua Levu forests, where the
large, woody, open follicles may be seen lying in numbers on the ground,
empty and in all stages of decay. The seeds of one species, near
Sterculia vitiensis, were nearly an inch long and sank like stones. The
unopened follicles will float for weeks; but it is evident that Nature
does not disperse the genus in this fashion, since the fruits before
dehiscence remain on the tree. It is also noteworthy that Gaudichaud,
when describing the floating drift of the Molucca seas, refers to the
open follicles of two or three species of Sterculia (_Bot. Chall.
Exped._, iii, 279). The fruits never came under my notice in the drift
of Fiji. The seeds of a Fijian species examined by me were four-fifths
of an inch (2 cm.) long. They had a thin, brittle, outer skin and
crustaceous inner test, and, being edible, might attract birds; but such
birds would be ground feeders, like the Megapod, and the Goura pigeon of
New Guinea, and the Nicobar pigeon, birds of this habit being rare in
Fiji. I should doubt whether the seeds are sufficiently protected to be
preserved from injury in a bird’s stomach during a long sea-passage; and
they may thus be placed in the same category with the seeds of
Myristica, a genus that has also failed to reach Tahiti and Hawaii.

But the distribution of Sterculia raises other more important questions
than that connected with its occurrence in Fiji, which involves an
over-sea passage of only 500 or 600 miles. As in Podocarpus amongst the
Coniferæ, which has a similar distribution in the Western Pacific, we
have to explain the existence of the genus in the three great
continental masses of Africa, Asia, and America, now separated by oceans
several thousands of miles across. Here also we must look far back into
the ages for a common centre of diffusion in the extreme north, such as
is in a sense suggested by the occurrence of the order in the Eocene
beds of Europe.

As showing unmistakably that Fiji received its species from the Old
World, it may be observed that one of its trees, Sterculia vitiensis, is
very closely allied to S. fœtida, widely spread in tropical Asia, in
Malaya, and Australia, as well as in Africa.

_Trichospermum_ (Sterculiaceæ)

There are only two species of this tree recorded in the _Index
Kewensis_, one in Java, and one in Fiji as well as in Samoa. The fruit
is a capsule with small, flat seeds, margined by long hairs, that might
possibly attach themselves to a bird’s feathers.

_Micromelum_ (Rutaceæ)

This small genus of tropical Asia, Malaya, tropical Australia and the
islands of the Western Pacific, has one species, Micromelum pubescens,
possessing the range of the genus with other species that are restricted
to different localities. We thus have apparently another illustration of
the part played by a wide-ranging polymorphous plant in providing new
species. The red berries would easily attract frugivorous birds; but the
seed-tests seem too delicate to allow the seeds to remain more than a
few hours in a bird’s stomach without injury.

_Cananga odorata_ (Anonaceæ)

This tree, which is cultivated in many places in tropical Asia and
Malaya, but is certainly indigenous, according to the authors of the
_Flora Indica_, in Ava and Tenasserim, has apparently extended into the
Pacific by cultivation. But though much valued by the natives on account
of its fragrant flowers, and in consequence often planted by them near
their villages, it grows in some localities in Fiji and Samoa as an
indigenous plant. The berries are especially suited for dispersal by
frugivorous birds, their flat seeds, 8 mm. in length, possessing hard
crustaceous tests that would enable them to pass unharmed in a bird’s
droppings. According to Reinecke the fruits are sought after by pigeons,
and particularly by Didunculus strigirostris, the Samoan Tooth-Billed
Pigeon. The tree has not travelled eastward of Tonga and Samoa, with the
exception of its occurrence in Rarotonga; and according to Mr. Cheeseman
the Rarotongans received it from Samoa several years ago.

_Geissois_ (Saxifragaceæ)

This genus of seven or eight known species is found in Australia, New
Caledonia, the New Hebrides, and Fiji. Since New Caledonia possesses
four species, it may be considered the home of the genus. To the Fijian
endemic species, G. ternata, I paid special attention. The capsules
dehisce on the tree and allow the small seeds to escape. These seeds,
which are very light, 150 to 200 going to a grain, are 3 to 4 mm. long
and are winged at one end. They could no doubt be carried some distance
by strong winds; but they possess no buoyancy. Large bats probably aid
in their dispersal. The Fijians assert that these animals are in the
habit of visiting the trees for the sake of the honey furnished by the
conspicuous red flowers. When they see a bat flying towards these trees,
they are wont to remark that it is going to drink the “se ni vota,” that
is, to suck the flowers of the Vota tree. It is very likely that seeds
would sometimes be carried in their fur for considerable distances.

_Begonia_

Before the discovery of Hillebrandia, a new genus of the Begoniaceæ, in
Hawaii, the order was not known from Polynesia. However, in 1878 Mr.
Horne collected a species of Begonia in Fiji, and it was probably this
species that frequently came under my notice in the rain-forests of the
Vanua Levu mountains. In 1883 I collected a Begonia in the Solomon
Islands, which I gave to Baron F. von Mueller, who informed me that it
was the first record of the genus east of New Guinea, the description of
Mr. Horne’s Fijian plant apparently not having been published (see
Guppy’s _Solomon Islands_, p. 288). It is not easy to explain why a
genus with such minute seeds, which are apparently as well fitted for
dispersal as those of the orchids, should have such a limited
distribution in the Pacific.

_Dolicholobium_ (Rubiaceæ)

In the _Index Kewensis_ this genus, containing five species, is
restricted to Fiji. It must, however, be more generally distributed in
the Western Pacific, since the genus was identified at Kew among my
Solomon Island collections, and it is recorded in the list given in my
book on that group (pages 283, 288, 297).

The showy, large, white, fragrant flowers of these small trees recall
those of Lindenia, with which Dolicholobium is often associated in Fiji
by the sides of streams and rivers. As Horne observes, the Fijian
Dolicholobiums range from the sea-shores and the heads of the estuaries
to the tops of the highest mountains. As noticed by me in the Solomon
Islands they affected the same station, being especially common on the
banks of streams. The genus has a long, narrow capsule six inches or
more in length. The linear seeds, though very light, are an inch or more
long, the coats being drawn out into a long tail at either end, and thus
differing greatly from those of Lindenia, the other Rubiaceous genus,
with which these plants are so frequently associated at the river-side.
I can only suppose that the seeds are transported by the winds. The
history of the genus is suggested in my remarks on Lindenia.

_Lindenia_ (Rubiaceæ)

Respecting its distribution in the Pacific, this genus of showy
river-side shrubs takes the same place amongst the plants that Galaxias
takes among the fishes. It is full of mystery. Of the four species
known, two grow on the river-banks of Central America and two in similar
stations in the islands of the Western Pacific. Of the last-named both
occur in New Caledonia, one of them being endemic, whilst the other,
Lindenia vitiensis, is found also in Fiji and Samoa. Reinecke seemingly
records no Samoan species, but in the list of additions at the end of
his _Flora Vitiensis_, Seemann refers to the Fijian species as having
been found in Samoa by Dr. Graeffe.

Lindenia vitiensis, as Horne aptly remarks, adorns the rocky banks of
many Fijian streams with its cream-coloured flowers, which impregnate
the air with their sweet odour. I found it in Vanua Levu, both at the
heads of the estuaries and beside the stream and the torrent in the
heart of the mountains. It was often associated with a species of
Dolicholobium, which it resembled strangely in its large, showy, scented
flowers and in the form of the leaf. Seemann says it is also accompanied
at the river-side in Viti Levu by Ficus bambusæfolia and Acalypha
rivularis. It is noteworthy that all the four plants here mentioned as
being associated river-side plants in Fiji possess the long, narrow
leaves of the willow type, a subject that is discussed in note 79.

The capsules of Lindenia vitiensis contain numbers of small, angular
seeds about 1·5 mm. across, some 400 of them when well dried going to a
grain. The seeds float buoyantly by reason of their outer covering of
crisp, air-bearing, cellular tissue. When this outer covering is
stripped off, the minute nucleus, or seed proper, which is barely a
millimetre across and is but slightly protected, sinks at once. As the
seeds float on the surface of a stream they might readily get on the
plumage of an aquatic bird; but they have no special means of
attachment; though, if they dried on the feathers they might adhere to
some extent. That they could be carried in mud adhering to a bird across
an ocean’s breadth I think most unlikely; and it should be remembered in
this connection that only the dead or sickly seeds would be found at the
bottom of a stream.

The most reasonable explanation of the extraordinary distribution of
Lindenia is that it was in a past age found over the tropical regions of
both America and the Old World, and that it has died out over the
greater part of its original area. To study the means of dispersal of
plants with such a distribution seems almost futile. I am inclined to
think that the limited range of Dolicholobium, so frequently its
station-companion in Fiji, may be similarly explained.

_Limnanthemum_ (Gentianaceæ)

This interesting genus of aquatic plants is dispersed over the tropical
and temperate regions of the globe, but with the exception of Fiji and
the New Hebrides it is not found in oceanic groups, though it occurs in
large continental islands like New Caledonia and Cuba. About twenty
species are enumerated in the _Index Kewensis_, but it is stated in the
_Genera Plantarum_ that they can probably be reduced to ten, the
reduction being chiefly applicable to the tropical species, nearly all
of which are reducible to varieties of L. indicum, the temperate species
being often very distinct. It would thus appear that although dispersal
is still active in the tropics, it is in part suspended in the temperate
zone, and we seem to possess in L. indicum a typical polymorphous
species that has played the _rôle_ of Naias marina in the warm, fresh
waters of the globe (see page 368).

Although some of the temperate species, like Limnanthemum nymphæoides in
Europe and Northern Asia, have a wide range, it is probable that this is
connected not so much with means of dispersal, as with its relation to
present and past drainage-areas. Rivers in the lapse of ages change
their courses and carry their aquatic floras with them, leaving,
however, a few of their plants around the springs and in the lakes which
serve still as centres of dispersal. Rivers may even exchange their
plants in flood-time in extensive level districts. Nor is the occurrence
of the genus in the Old and New Worlds in the northern hemisphere to be
connected with questions of dispersal across an ocean. Except in the
case of small-seeded plants, like Nasturtium and Lythrum, where the
dispersal could be carried on by water-fowl, the plant-species being
often identical on both sides of the Atlantic, it is probable that most
of the large-seeded river-side genera common to Europe and North
America, such as Iris and Acorus, had in past ages their home in the
extreme north, whence the plants spread as from a focus into the
continents of America and Eurasia. It is also to be doubted whether even
in the tropics there has been much over-sea dispersal of Limnanthemum
without the aid of man, and reasons will be given for the belief that
probably in Fiji, in the New Hebrides, and in New Caledonia the seeds of
the first plants were unintentionally introduced by the aborigines.

Following Bentham we may regard the species of the Western Pacific
Islands as a form of the wide-ranging Limnanthemum indicum. These plants
in Fiji do not play the part in river-vegetation that they do in the
temperate regions, as for instance in the Upper Thames. They are not
common except in places, and seem to be chiefly confined to Viti Levu,
particularly to ponds in the Rewa delta, where their _rôle_ is that of
an Indian tank plant. In the Rewa delta they may be sometimes seen
thriving in brackish water having a density of 1·005.

Looking at the mode of dispersal to which the Limnanthemums owe their
existence in the Western Pacific, we cannot disregard, especially in
Fiji, the possibility of the seeds having been unintentionally
transported by the natives when they carried in their migrations their
edible tubers, such as Colocasia antiquorum, Alocasia indica, and
Cyrtosperma edulis, that are cultivated in wet places. It is in the
ponds around which these plants grow that the Limnanthemums thrive. The
Chinese, with their peculiar methods of cultivation, are now carrying
with them strange water-plants over the warmer regions of the globe; and
it would be surprising if the Pacific islanders in their migrations did
not do the same. If such an introduction, however, took place, it must
have happened before the time of Captain Cook, when the plant was found
in New Caledonia. (It may be remarked in this connection that the seeds
of the genus will germinate after being kept dry for years. Seeds of the
British species which I had kept dry for two and a half years germinated
healthily when placed in water.)

Some years ago I ascertained that the seeds of the British plants were
enabled, by means of their fringe of hairs, to attach themselves firmly
to the downy plumage of a bird’s breast. This could not happen with the
Fijian plant as the seeds are naked, and the same may be said of some
species described by Gray and Chapman as widely spread over the United
States. The seeds of the genus appear quite unsuited for safe transport
inside the body of a bird. The Fijians give the plants a variety of
names, nearly all of which are associated with the word for a duck, and
none of them bear an ancient impress. Thus we find such names as
“Ndambe-ndambe-ni-nga” and “Vothe-vothe-ni-nga,” meaning respectively
“the duck’s seat” and “the duck’s paddle.”

_Ceratophyllum demersum_

This wonderful aquatic has been dispersed over most of the globe; but I
will only mention its occurrence in oceanic islands, such as Fiji,
Samoa, the Bermudas, and the Azores, to indicate the necessity of
attributing its distribution in islands to birds. Several years ago I
made a careful study in England of the habits and mode of germination of
this plant, the results of which are given in _Science Gossip_ for
November, 1894; but reference can only be made here to such points as
bear on the occurrence of the plant in the Pacific islands.

It is well known that in our English ponds and rivers the plant
propagates itself, as a rule, by budding; and that it is only in
unusually hot and dry summers, such as that of 1893, when many ponds
became very low and were excessively heated, that the fruits mature in
any quantity. My observations clearly showed that a higher temperature
is required for the completion of maturation than for the early stage of
the fruiting process and for the flowering. After a comparison of my
river and pond temperatures, I formed the conclusion that whilst in
water 12 to 18 inches deep this plant requires for a week or more an
average daily maximum water temperature of 70° F. to produce its
flowers, a warmth of 80° and over is necessary to mature its fruit, a
condition to be found in England only in shallow ponds, where the plants
may fruit abundantly, but not in rivers, where they flower and rarely
mature the fruit (see also for the thermometric conditions my paper in
_Proc. Roy. Phys. Soc. Edin._, xii, 296). Since a yet lower temperature
(an average maximum water temperature of 66° for a week or more) is
sufficient for germination, it follows that the thermal conditions of
our English climate will allow Ceratophyllum to germinate and to flower,
though but rarely to mature the fruit.

Even in Fiji we can notice the distinction between the cooler river and
the superheated ponds and swamps of the Rewa delta as regards the
maturation of the fruit. In 1897 I found Ceratophyllum thriving in the
main channel of the Lower Rewa where the water was quite fresh; whilst
lower down where the water was often brackish its place was taken by
Ruppia maritima. In the main river, where the water unmixed with
sea-water rarely acquires a temperature of 80° F., the reading being
usually 78° to 79°, I never found the plants in fruit, and it is only in
the superheated shallow waters of the swamps and back-waters that they
mature their fruits.

Since Ceratophyllum even in tropical climates would probably only mature
its fruits in the superheated waters of shallow ponds, tanks, and
ditches, it follows that its dispersal by birds is confined to warm
regions. In the cold waters of the Siberian lakes and rivers it would
never mature its seeds, and could only be propagated by budding. If it
existed in the head-springs of the sources of a river in these
latitudes, it would be distributed by means of its floating shoots and
fragments along the length of the river basin, and in the times of flood
it might pass in the lower plains from one river system to another. When
rivers changed their courses it would be left behind in the lakes and
ponds and springs, and would also be carried away to the new region. In
this manner it would in the course of ages be distributed over a
continent without the aid of seed, propagating itself in a vegetative
fashion.

In the case of oceanic islands, however, we have to appeal to the seed.
Since the fruits sink in sea-water even after prolonged drying, and
since a few days’ immersion in sea-water, as I found, kills the floating
plant, we are driven to the agency of birds. The fruits, which without
appendages are a quarter of an inch (6 mm.) in length, are too large and
heavy to be carried in dry mud adhering to birds. The chances of their
becoming entangled in a bird’s feathers by means of their basal spines
and terminal style seem small, since they would be lying usually on the
mud under the water. They are quite fitted for safe transport in the
stomach and intestines of birds, such as is established in Chapter
XXXIII for Potamogeton and Sparganium in the case of ducks. As my
experiments show, drying for a period of three months does not injure
the germinating capacity of the seeds.

_Dracontomelon_ (Anacardiaceæ)

This is a genus accredited in the _Index Kewensis_ with eight species,
of which three belong to Borneo, one to Sumatra, one to Java, one to the
Philippines, and two to Fiji, all the species being restricted in their
range. My observations were confined to D. vitiense, Engler (D.
sylvestre in Seemann’s work), the Tarawau of the Fijians, who regard it
as a tree that is planted by the dead in Naithombothombo, the place of
departed spirits, according to the legend given by Hazlewood in his
Fijian Dictionary. Its method of dissemination in the Fijian forests is,
however, far more prosaic. Pigs and fruit-pigeons assist in the
dispersal of the seeds in these islands. Pigs are often found in the
vicinity of a Tarawau tree; and evidently they much appreciate the
fallen fleshy fruits, which are about 1-1/3 inch (3·3 cm.) across and
inclose a large stone 7/8 inch (2·2 cm.) in diameter. The entire fruit
and the detached stone sink in sea-water, the last floating only a few
hours, even after drying for four years. Mr. Hemsley regards the genus
as probably dispersed by the currents, since a stone was found amongst
the floating drift collected by the _Challenger_ Expedition off the
coast of New Guinea. The stone, however, is described as seedless, which
may explain its buoyancy. It is, however, to the fruit-pigeon that we
must look for the dispersal of this genus. In the crop of one of these
birds shot in Fiji I found the entire fruit of a Tarawau tree.

_Canarium_ (Burseraceæ)

This genus of trees, to which nearly a hundred species are referred in
the _Index Kewensis_, belongs mainly to tropical Asia and Malaya, a few
species occurring in tropical Africa, Madagascar, the Mascarene Islands,
and Polynesia. Its great home is in Malaya, to which two-thirds of the
species are confined; but its distribution in the oceanic islands of the
Indian and Pacific Oceans is especially interesting, Mauritius, Bourbon,
Fiji, Tonga, and Samoa (Horne) each possessing a species.

The large drupes of the genus, as I found in Fiji, have no capacity for
dispersal by currents; and we are, therefore, compelled to appeal to the
agency of the frugivorous bird. Yet to a person unaccustomed to the ways
of fruit-pigeons the transportation across a broad tract of ocean of
large heavy “stones,” an inch and more in size, would seem impossible;
and even to a student of dispersal improbable. Unless, however, we
prefer to accept the Lemurian theory for the Indian Ocean and the theory
of a Melanesian continent for the Pacific we are compelled to appeal to
these birds; and it can scarcely be said that our appeal is without some
justification. Both in the Solomon Islands and in the Fijis I was
familiar with the dispersal of the stones of these trees by
fruit-pigeons; and Wallace, amongst other writers, observed the same
long ago in the Malayan Islands (_Malay Archipelago_). Stones obtained
from the crops of Fijian pigeons measured 1-2/10 × 1 inch (3 × 2·5 cm.).
In the Solomon Islands these birds stock the interior of the coral
islets with trees of the genus, and the ground below the trees is often
strewn with the disgorged stones (_Bot. Chall. Exped._, iv, 310; Guppy’s
_Solomon Islands_, p. 85).

Although the difficulty concerned with the transport of the seeds across
a broad tract of ocean seems very great, it is quite possible that
further investigation will enable us to overcome this objection, just as
we have done in Chapter XXVI when explaining how the genus Elæocarpus
may have reached Hawaii. It is, indeed, not unlikely that, as with
Elæocarpus, the stones of the drupes may in some species be much smaller
and far more fitted for being carried in a bird’s body over several
hundred miles of ocean.

_Couthovia_ (Loganiaceæ)

Reference is here made to this genus because its mode of dispersal is
known, and because I was familiar with it in Fiji. Seemann gives two
species for Fiji, C. corynocarpa and C. seemanni, and the few other
species known seem to be confined to the Western Pacific. Solereder
gives a third species, C. densiflora, for Kaiser-Wilhelmsland in New
Guinea (Engler’s _Pflanz. Fam._ teil 4, abth. 2); and a Solomon Island
species, nearly allied to, if not a variety of, the Fijian species, C.
seemanni, is referred to in the list of plants from that group given in
my book on those islands. I found C. corynocarpa not infrequently
growing on the banks of small rivers in the heart of Vanua Levu. Its
drupes, which float for a few days in sea-water, are, according to
Seemann, eaten by fruit-pigeons. The “stone” varies from 2 to 4
centimetres (3/4 - 1-1/2 inch) in length; and from the standpoint of
dispersal the genus ranks with Canarium and Dracontomelon. Seemann
describes and figures this species, which was constituted by Gray, in
his _Flora Vitiensis_; but, apparently through an error, it is in the
_Index Kewensis_ accredited to Hawaii. Hillebrand makes no reference to
the genus in his book on the Hawaiian flora.

_Veitchia_ (Palmaceæ)

This genus of palms is closely allied to Ptychosperma, a Malayan genus
also represented in Fiji. The _Index Kewensis_ names four species, one
New Hebridean, and three Fijian. The fruits of two of the last-named
species tested by me had no floating power. The seed is about an inch
long, and the genus would be likely to be spread by fruit-pigeons. From
the standpoint of dispersal the genus would be placed with Canarium and
Couthovia; but possibly its presence in the Pacific may be indicative of
an ancient Western Pacific continent.

_Hibbertia_ (Dilleniaceæ)

This genus of some eighty known species is almost entirely Australian,
with the exception of a few species found in New Caledonia, Tasmania,
and apparently also in the Mascarene Islands. Horne was the first to
record a species from Fiji, where it grows commonly in the “talasinga”
plains on the lee sides of the islands, and also on the scantily
vegetated mountain summits. In Vanua Levu I often found these plants
growing on the rocky peaks of the highest mountains of the island, as on
Mbatini, 3,500 feet, and on Mariko, 2,900 feet. Their presence on these
isolated peaks can only be attributed to birds. The carpels contain one
or two seeds, which have a membranous aril; but in the plains the seeds
are usually destroyed by grubs.

_Myrmecodia_ and _Hydnophytum_ (Rubiaceæ)

These two genera of epiphytes, distributed over Malaya and extending to
the islands of the Western Pacific, possess tuber-like stems, which are
extensively chambered by ants that find a home in the interior. They
were familiar to me in the Solomon Islands, where they frequently grow
on the mangroves and on other littoral trees. They do not form such a
feature in the shore vegetation of Fiji, and judging from the
observations of Dr. Seemann and myself they occur most often on the
wooded mountain-peaks. The berries of these plants would attract
frugivorous birds; and their pyrenes, which in a Fijian Myrmecodia I
found to be 4 millimetres long, appear quite suitable for dispersal
through this agency. It would seem that germination may occur in the
berry on the plant. A specimen of Myrmecodia in fruit, that had been
lying overlooked for a fortnight between newspapers during one of my
mountain journeys, displayed on examination the pyrenes in a germinating
condition, the process being subsequently completed. The reader will
find these interesting plants described and illustrated in the English
edition of Schimper’s work on _Plant-Geography_, pp. 149, 150.

_Myristica_

The Nutmeg trees, though principally at home in Indo-Malaya, are found
also in the warm regions of Africa and America, as well as in the
islands of the Western Pacific from the Solomon group eastward to Fiji,
Tonga, and Samoa. The Tongan and Samoan groups possess two species in
common, whilst Fiji seems to possess its own species, four or five in
number.

The seeds of this genus have long been known to be dispersed by
fruit-pigeons. Mr. Moseley, in his _Notes of a Naturalist_, and in the
_Journal of the Linnean Society_ (vol. xv), tells us how at one time
these birds in their dissemination of the seeds in the Banda Islands
were active opponents of the policy of the Dutch Government in
preserving their monopoly of the cultivation of the nutmeg of commerce.
He found numbers of wild nutmegs in the crops of these birds in the
Admiralty Islands, some of which were partially digested and others
seemingly sound; and Mr. Hemsley includes the genus as amongst those
dispersed in the Western Pacific by birds (_Bot. Chall. Exped._, Introd.
46; iv, 229, 308). In my book on the Solomon Islands I refer to the
occurrence of these seeds in the crops of fruit-pigeons; and I found
that the seeds were similarly dispersed by these birds in Fiji. It is
likely that the absence of the genus from Eastern Polynesia is to be
partially connected with the insufficient protection of the seeds
against injury during such a long ocean passage in a bird’s body.

Gaudichaud, as quoted by Hemsley, refers to the occurrence of the fruits
of three or four species of Myristica in the drift floating in the
Molucca Sea. When in the Solomon Islands I noticed that the unopened
fruits of a species floated in sea-water. In later years in Fiji I
tested this point, and found that whilst the fruits just before
dehiscing will float between three and seven days in sea-water, the
seeds sink. As I have pointed out in the chapter on Drift, rivers carry
down to the sea an abundance of seeds and fruits that can float a few
days but do not imply dispersal by currents.

Although, as I have above remarked, the localised range of the genus in
Polynesia may be in part connected with the insufficient protection of
the seed, it is apparent that in the case of a genus found in Asia,
Africa, and America we are brought into contact with questions other
than those of means of dispersal. No one would pretend that Myristica
seeds could be carried by birds uninjured across the Pacific Ocean; and
to explain the present distribution of the genus we must recall cases of
a similar kind, such as Podocarpus, where the genus in past ages had a
home in the north, from which, as from a focus of dispersion, it
extended into the continents of the Old and the New World (see p. 302).

_Rhaphidophora_ (Araceæ)

This genus of climbing aroids, which gives a character to the forests of
Indo-Malaya as well as to those of the Western Pacific, is represented
in the New Hebrides, Fiji, Tonga, and Rarotonga by a variety of the
widely spread R. pertusa that ranges over Indo-Malaya and Eastern
Australia. The ripe berries would readily attract birds; and the seeds,
4·5 millimetres long in the case of a Fijian plant, appear hard enough
to pass unharmed through a bird’s digestive canal. We seem here to have
evidence of a somewhat recent connection between Indo-Malaya and
Polynesia through the agency of frugivorous birds. That the genus has
been long established in Polynesia is, however, indicated by the
occurrence there of a species seemingly peculiar to Fiji. We are
disappointed that in Engler’s recent contribution to the _Pflanzenreich_
(in his volume on the Araceæ-Pothoideæ) he has not been able to include
this genus in the field of his studies.

_Gnetum_ (Gnetaceæ)

This Gymnospermous genus, which is found in the warm regions both of the
Old and the New World, is represented in Fiji by a Malayan species,
Gnetum gnemon, which exists also in the Solomon group with other species
of the genus (Guppy’s _Solomon Islands_, pp. 288, 301). I was familiar
with this species in both Fiji and the Solomon group; but in the
first-named locality it is seemingly restricted to the borders of
Wainunu Bay on the south side of Vanua Levu, where Dr. Harvey first
found it. It grows there abundantly in young wood.

It seems almost idle to discuss the mode of dispersal of a genus that is
placed in a class apart with the African Welwitschia and the European
Ephedra, possessing with them a history of which we know nothing. Yet it
is ranked by Mr. Hemsley amongst those genera that are dispersed in
Polynesia by birds, and he produces better evidence in support of this
view than we possess for many other plants. Thus a fruit of a species of
Gnetum, perhaps G. gnemon, has been found in a New Guinea fruit-pigeon;
and the fruits of two species of the genus were found in the crops of
fruit-pigeons shot by Mr. Moseley in the Admiralty Islands (_Bot. Chall.
Exped._, Introd. 46; iv, 308). The red drupes of Gnetum gnemon of Fiji
would readily attract birds, and their nut-like stones, about 8
millimetres long, are well suited for this mode of dispersal. My
experiments in Fiji show that neither the drupe nor the stone of this
species floats in sea-water; and it is probable that the fruits of this
genus referred to by Mr. Hemsley as having been picked up on the beach
in the Aru Islands possessed only a temporary buoyancy.

This genus presents us with the same puzzling question put to us by
several Fijian genera, such as Myristica and Podocarpus, that occur in
both Asia and America; and until we answer that query it seems almost
futile to study modes of dispersal.

_Elatostema_ (Urticaceæ)

This genus of annual and perennial herbs belongs to the tropical regions
of the Old World. It is represented in Samoa by fifteen known species
and by at least four or five in Fiji, whilst with the exception of a
solitary Tahitian species it is not recorded from East Polynesia.
Reference is here made to it particularly on account of its great
development in Samoa. We have here a genus that, like Psychotria in Fiji
and Cyrtandra in Fiji, Samoa, and Hawaii, runs riot in respect to the
production of species (see p. 317). Dr. Reinecke describes fifteen
Samoan species, of which, with the exception of two found in Malaya, all
seem to be described for the first time. So sensitive, he remarks, is
the genus to external conditions that station-forms abound; and he
points out that if we were to follow the dividing lines usually
recognised between species, we should account every station-form a new
species. It is, of course, obvious that the polymorphism of the Samoan
Elatostemas depends primarily not on the varying influence of station
but on their sensitiveness to external conditions. One might put the
question to the Samoan Elatostemas that Hillebrand put to the Hawaiian
Cyrtandras, and ask why nature in this particular genus in this
particular locality thus luxuriates in formative energy. Almost every
Pacific group in respect of some of its plants presents the problem so
well stated by Dr. Reinecke for this genus in Samoa. It is noteworthy
that Schimper, in his work on _Plant-Geography_ (English edition, pp.
291, 297, 299), especially singles out Elatostema and Cyrtandra as
growing socially in the tropical rain-forests of Java and of the Asiatic
mainland.

_Scirpodendron costatum_ (Cyperaceæ)

As far as I can gather, this giant-sedge has not been previously
recorded from Fiji; but it is included in the Samoan flora, and has also
been found at Penang and Singapore, as well as in Borneo, Java, and
Queensland. In Samoa, as we learn from Reinecke, it grows both in the
coast swamps and on dry ground. In Fiji it is very common in the
mangrove-swamps at the mouths of rivers, especially in the Lower Rewa;
but in Vanua Levu it is also frequent in the marshy localities of inland
plateaux, 700 to 800 feet above the sea, as well as by the side of
streams in swampy districts on the lower hill slopes. This double
station in the salt-water swamp of the coast and in the fresh-water
marsh of the interior seems to be repeated in Java, where the plant was
first discovered by Zippelius on the banks of torrents in mountainous
regions and in swampy places.

The genus comprises, according to the _Index Kewensis_, only this
species, though variations are to be observed in plants from different
localities. The species was described by Kurz in the _Journal of the
Asiatic Society of Bengal_ (vol. 38, 1869) and by Bentham in his _Flora
Australiensis_; and an illustration is given by Miquel in his
_Illustrations de la Flore de l’Archipel Indien_ (1871). The plant is so
common in Fiji that one can only suppose that its resemblance to a
stemless Pandanus, from which, as Kurz observes, it is with difficulty
distinguished except when in flower or fruit, led to its being
overlooked by both Seemann and Horne. Its leaves, from 9 to 12 feet in
length, are commonly used for making mats and for thatching, both in
Fiji and Samoa. The plant usually attains a height of 3 to 5 feet.

The fruits occur abundantly in the floating and stranded river and sea
drift in Fiji, a circumstance that led to my discovery of the parent
plant in the swamps. The fruit, which is about half an inch (12 mm.)
long, consists of a hard, stony nut invested by a thick ribbed cork-like
covering, to which it owes its buoyancy, since the nut sinks. The
detached fruit is perforated at the base through both coverings, and
only a little soft tissue closes the aperture in the inner shell, the
protection against the entry of sea-water in the case of floating fruits
being quite inadequate. This explains also why the stranded fruits were
so frequently found by me germinating on the beach, where, as my
observations showed, they never established the plant. This early
germination would prove to be an advantage in the case of fruits
stranded in a suitable locality.

But though the perforation in the fruit favours its early germination,
it lessens its ability to withstand a long sea-passage without injury to
the embryo. I found in different experiments on fruits of plants growing
in the mangrove swamps, that when placed in sea-water 40 per cent. sank
during the first fortnight, whilst 15 per cent. floated after five or
six weeks, but all were at the bottom in two months. On the other hand,
fruits from plants of the swamps of the inland plateaux displayed much
feebler floating power, in some cases sinking at once, in others
floating for a few days, and in others again floating for a week or two.
In this case the outer cork-like covering proved to have lost most of
its floating power.

From the number of empty seed-vessels found, both in the floating and
stranded drift, it appeared evident that the seed had often rotted away
during the flotation. It is apparent from these observations and
experiments that Scirpodendron costatum is not suited for dispersal by
currents over wide tracts of ocean. The fruits might be able to float
unharmed for a few weeks, but they would be unable to accomplish much
more than the 500 or 600 miles intervening between Fiji and the nearest
groups to the west.

_Lemnaceæ_

This order, judging from the writings of Hegelmaier, Schenck, and
Hemsley, is represented by one or other of the common species, Lemna
minor, L. gibba, L. polyrrhiza, in various Atlantic islands, as in the
Bermudas, the Azores, Madeira, the Canary Islands, and St. Helena; but
doubts frequently arise as to their being truly indigenous. Lemna
trisulca is regarded by Hemsley as indigenous in the Bermudas. Lemna
minor has been introduced in recent years into Hawaii, where I observed
it flowering and sometimes fruiting abundantly in the heated waters of
the ponds. Two species found in other regions were recorded by Seemann
from Fiji, and I have come upon few other records of the occurrence of
the order in the tropical islands of the open Pacific. I am inclined to
the opinion, based not only on the facts of distribution, but also on
the results of numerous experiments on the means of dispersal, that this
order has in most cases reached oceanic islands with man’s assistance.

Some years ago I made a systematic study of the habits of the British
Lemnæ, most of the results being published in the _Linnean Society’s
Journal_ (vols. xxix and xxx), as far as concerned Lemna minor, L.
gibba, and L. polyrrhiza. During this inquiry I ascertained that with
these species, as well as with L. trisulca, the chances of a bird’s
carrying their fronds uninjured in its plumage over a wide extent of
ocean were small. None of them survived twenty-four hours’ drying in
fine weather, whether in the sun or in the shade; but in rainy weather
they withstand an exposure of one or two days. It is, therefore,
unlikely, even if the fronds were entangled by their rootlets in a
bird’s feathers, that they would be able under ordinary conditions to
reproduce the plants after a day’s flight of some five hundred miles
across the sea. It must also be remembered that the drying capacity of
the air when a bird is in full flight in ordinary weather would be that
displayed during a gale of wind with a velocity of at least thirty to
forty miles an hour. For this reason I do not think with Kerner that
under usual conditions drops of water would be a factor of importance in
causing the adherence of minute seeds of any kind to birds’ plumage.
Where the seeds are not available, it is most probable that birds
disperse the duckweeds by their fronds over short distances, but not
across broad seas. This would certainly apply to temperate latitudes,
where these plants rarely seed. Thus with Lemna, as with Ceratophyllum,
it would seem that the dispersal of the seeds by birds takes place
normally only in warm latitudes. Those of the duckweeds could be
transported in adherent mud over land-areas.

According to Hegelmaier, the two species of Lemna found in Fiji are L.
paucicostata, an Asiatic species, and a variety of an Australian
species, L. oligorrhiza, possessing dark root-sheaths. These plants
mostly came under my notice in the Rewa delta. They were rarely seen in
Vanua Levu, where in one locality I found the typical Lemna minor. The
first species is also Samoan.

In 1897 and in 1899, in a pool near Notho in the Rewa delta, in Viti
Levu, Fiji, I found a great abundance of a species of Wolffia, specimens
of which were sent to Prof. Schimper with my mangrove collections, but
his death intervened, and I have not been able to follow up the matter.
On comparing the specimens with Hegelmaier’s descriptions and plates, it
would seem that the species is near W. arrhiza and W. brasiliensis, but
differs from both in the greater length of the fronds. As concerning the
means of dispersal of the genus, I may add that the fronds were killed
after being allowed to dry for eighteen hours.

_Marsilea_ (Marsileaceæ)

A species of this genus, apparently near Marsilea villosa, was common in
the ditches and ponds around Notho, in the Rewa delta, Fiji, in 1897-99.
The genus is included by Horne in his list of Fijian plants; but is not
given by Seemann. The villous sporocarps, when dry, are very light and
readily catch in cloth and in feathers. Hillebrand includes in the
Hawaiian flora M. villosa and M. crenulata. The first-named, which was
collected by Chamisso and Gaudichaud, finds (he says on the authority of
Braun) its nearest relative in a species from Oregon and California. The
other has been collected in the Liukiu Islands, the Philippines,
Mauritius, and Bourbon. It is very probable that the occurrence of the
genus in oceanic islands is due to the agency of birds.

_Summary of the Chapter_

(1) We are here concerned with the more restricted distribution of
non-endemic tropical genera over the Pacific. The general trend eastward
of these genera is well brought out in the fact that whilst Fiji
possesses some sixty or seventy genera in common with Tahiti to the
exclusion of Hawaii, it does not possess a score in common with Hawaii
to the exclusion of Tahiti. The grasses and sedges and the mountain
genera are not here included; and we are comparing the flora of the
Hawaiian lowlands below 4,000 feet with the floras in mass of Fiji and
Tahiti.

(2) Hawaii possesses very few genera (less than thirty) that are not
found either in Fiji or in Tahiti, or in both; and of these quite a
third are to be traced to America.

(3) From two of these genera, Embelia, a land genus, and Naias, an
aquatic genus, we obtain two important indications, namely, that
specific differentiation has taken place to much the same extent in a
water plant as in a land plant, whether in a continent or in an island.
In other words, new species have been developed or are developing
independently of the immediate environment and of isolating influences.

(4) The interchange of plants between the regions of Hawaii and Tahiti
to the exclusion of Fiji has been very slight. Probably not half a dozen
genera belong to this category.

(5) Excluding plants brought by man and by the currents, Tahiti
possesses very few that present any difficulty from the standpoint of
dispersal, plants with seeds or “stones” an inch in size being, as a
rule, absent.

(6) With the genera (60-70) common to Fiji and Tahiti, and distributed,
therefore, over the South Pacific, the wide-ranging highly variable
plant is an important factor in the development of peculiar species in
the different groups, just as it has been shown to be in the previous
chapter in the case of genera dispersed over the whole Pacific. The
_rôle_ of the polymorphous species has always been an important one in
this region.

(7) In the case of several Fijian genera it seems almost futile to talk
of existing means of dispersal, since the present distribution of genera
like Sterculia and Gnetum, that occur on both sides of the Pacific, in
America and in Asia, is not to be thus explained.

(8) On account of the large size of their seeds and “stones” it might be
argued that certain of the Fijian plants afford evidence of a previous
continental condition of the islands of the Western Pacific, since it is
not easy to understand how such large seeds and “stones” could have been
transported over broad seas by birds. It is, however, pointed out that
in these respects the species of a genus may vary greatly, and that the
seeds and stones may be large in some species and small in others.

(9) The greater number of the genera that have entered the Pacific from
the Old World have not advanced eastward of the Fijian region, half of
the Fijian genera not occurring in the Hawaiian and Tahitian regions;
and the explanation of this is to be found not in any lack of capacities
for dispersal, but in a want of opportunities. The story of
plant-distribution in the Pacific is bound up with the successive stages
of decreasing activity in the dispersing agencies. The area of active
dispersion that at first comprised the whole of the tropical Pacific was
afterwards restricted to the South Pacific, and finally to the Western
Pacific only. The birds that in an early age carried seeds all over this
ocean became more and more restricted in their ranges, probably on
account of increasing diversity of climatic conditions. The plants of
necessity responded to the ever narrowing conditions of bird-life in
this ocean, and the differentiation of the plant and of the bird have
taken place together.

CHAPTER XXVIII

THE POLYNESIAN AND HIS PLANTS

MAN AND THE SEED

MAN in his distribution in the islands of the Pacific reproduces in a
minor degree nearly all the difficulties presented there by plants,
birds, and other forms of animal life. Like the plant he entered the
ocean from the west; and as with the plants, so with the aborigines,
there was an era of general dispersion over this ocean, followed by an
age in which Polynesian man, ceasing to migrate, tended to settle down
in the several groups, there undergoing differentiation in various
respects, as in physical characters, in language, and in manners. Just
as we can now recognise the type of a plant, of a bird, or of an insect,
that belongs to a particular group of islands, so we can distinguish
between the Hawaiian, the Tahitian, and the Maori, whether in physical
characters, in his speech, or in his customs. Fiji possesses in the
Papuan element of its Melanesian population the earliest type of man in
the Pacific, just as it also possesses in the Coniferæ the most ancient
types of trees in this region. Divesting his mind of all previous
conceptions, the ethnologist, as I have remarked in my discussion of the
distribution of Freycinetia in Chapter XXV, might profitably study _de
novo_ the dispersion of man in the Pacific from the standpoint of
plant-dispersal.

Man and the seed have battled their way over the Pacific apparently in
defiance of the prevailing winds and currents, and both have failed to
reach the New World. Man in the Pacific is almost as enigmatical as the
plant. As a denizen of this region he is by no means a recent
introduction; and though his food-plants are mainly Asiatic, they belong
to distinct ages in the history of man’s occupation of these islands.

I venture to think that a great deal lies behind the Indo-Malayan mask
of the Polynesian, and that there is a story concerned with his origin
that has yet to be told. We have by no means solved the riddle when by
following the evidence we assign to him a home in Asia. It is only then
that the real difficulties begin. It required many centuries of European
civilisation for the discovery of America; but the voyages of Columbus
sink into insignificance when we reflect on what had been dared and
accomplished by uncivilised man when he first landed on the shores of
Hawaii and Tahiti.

The problem of man in the Pacific bristles with difficulties differing
in degree but not in kind from those relating to the flora. Whenever a
particular theory seems on the point of being well established, some
disturbing question arises, and as with the plant, we are never able to
push our facts quite home. Since I first visited the Solomon Islands,
now twenty-four years ago, the Pacific islander and his flora have
deeply interested me. The history of man and of the plant cannot be
separated in the Pacific; and the same determining principles of
distribution have affected both.

THE FOOD-PLANTS OF THE POLYNESIANS AND PRE-POLYNESIANS

One can imperfectly distinguish two sets of food-plants in this region;
the first comprising such plants as Pachyrrhizus trilobus, Tacca
pinnatifida, Amorphophallus campanulatus, the Mountain Bananas, the Wild
Yams, and several others that grow wild, and, as a rule, only serve as
food in times of scarcity; the second including the plants that are
extensively cultivated by the present islanders, such as the Breadfruit,
the Banana (Musa paradisiaca), the Taro (Colocasia antiquorum) and the
two Yams (Dioscorea alata and D. sativa), &c. Those of the first set
probably formed the food of the earliest inhabitants of the Pacific
Islands, pre-Polynesian peoples that practised only a rude sort of
cultivation, as with the present “bush-men” of the islands of the
Western Pacific. Those of the second set belong to the later occupants
of these islands, the Polynesians.

(a) _The Pre-Polynesian food-plants._—In addition to those above named
one may mention Cycas circinalis, Cyrtosperma edulis, Lablab vulgaris,
Pandanus odoratissimus, Saccharum officinarum, Sagus vitiensis, &c.
Inocarpus edulis is probably to be here included, and amongst the Wild
Yams should be named Dioscorea nummularia and D. pentaphylla. Some of
them are now occasionally cultivated; but most of them only occur in the
wild condition, either as weeds or as larger plants growing
spontaneously in uncultivated localities. Even the knowledge of them as
food-plants has sometimes been altogether lost, the present inhabitants
of the Fijis, for instance, knowing nothing of Lablab vulgaris and Sagus
vitiensis as sources of food. The question of the antiquity of the
Coco-nut Palm in Polynesia was discussed at length by Seemann; but for
various reasons we cannot be absolutely certain whether or not it is an
older denizen of the Pacific islands than the Polynesian. It is,
however, to be inferred that it came originally from the home of the
genus in America, perhaps as a gift brought by the Equatorial Current
from the New World to Asia. Several chapters might be devoted to the
discussion of the earlier food-plants of these islanders; but here only
a brief reference can be made to a few of them.

Perhaps the oldest of the earliest aboriginal food-plants are those
that, like Cyrtosperma edulis and Sagus vitiensis, are apparently
confined to Fiji. Here we seem to possess indications of the development
of new species since that group was first occupied by man. Others, like
Pachyrrhizus trilobus and Cycas circinalis, that are restricted to the
groups of the Western Pacific may come next in relative antiquity.

Although most of the early food-plants hail from the Old World, the home
of Pachyrrhizus is in America. One may indeed wonder how a plant with
such a history ever reached the Western Pacific. It seems to be
generally distributed in this part of the ocean, having been recorded
from New Caledonia, the New Hebrides, Fiji, Tonga, and Samoa. Although
its edible roots are only used in times of scarcity, the plant grows
wild all over Fiji, being especially frequent in the “talasinga” plains.
Though I searched diligently, it never presented me with its seed. In
Tonga, according to Graeffe (as quoted by Reinecke) the plant is much
employed in preparing the land for yam-cultivation, since it restrains
the growth of weeds and keeps the soil moist.

Amongst the food-plants of this early period that are distributed over
the South Pacific as far east as Tahiti may be mentioned the Wild Yams
(D. nummularia and D. pentaphylla), the Mountain Bananas, Tacca
pinnatifida, Amorphophallus campanulatus, and others. Of these Tacca
pinnatifida and Dioscorea pentaphylla are alone found in Hawaii. I will
only now refer to the Mountain Bananas.

The Mountain Bananas of the tropical South Pacific, distinguished by
their erect fruit bunches and their seeded fruits, present us with one
of the mysteries connected with aboriginal man in this ocean. Whether in
New Caledonia, Fiji, Samoa, Rarotonga, or in Tahiti, they grow wild in
the interior, and form often a conspicuous feature of the vegetation in
the mountains and at the heads of the valleys. They are occasionally
cultivated. Their Fijian and Samoan names of “Soanga” and “Soa’a”
reproduce the names of the banana, “Saguing” and “Saing” in the Tagalog
language of the Philippine Islands. The Tahitian appellation is “Fehi”
or “Fei,” and this reappears in Samoa in the form of “Fa’i,” the word
for the common cultivated banana, Musa paradisiaca. The Rarotongan name
of “Uatu,” as given by Cheeseman, is suggestive of the Micronesian form
(Ut, Uut, &c., in the Carolines) of a widely spread banana word in
Malaya, Melanesia, and West Polynesia (Fudi, Vundi, Undi, &c., &c.). It
is not unlikely that all these South Pacific mountain bananas with erect
inflorescences and seeded fruits belong to one species, variously
designated by botanists as Musa fehi, M. uranoscopus, M. troglodytarum,
&c., and confined to this region. Under the name of Musa fehi Schumann
includes the New Caledonian and Tahitian plants, and he views the Samoan
plant as probably identical with them. This botanist, in his monograph
on the Musaceæ (Engler’s _Das Pflanzenreich_, 1900), establishes the
home of the bananas in tropical Asia, and considers that their
occurrence in America before the time of Columbus has not been proved.
Birds have no doubt often assisted in the dispersal of the wild, seeded
plants; but it is likely that man is responsible for the occurrence of
the mountain forms in the Pacific, and probably their fruits formed when
cooked one of the principal articles of diet of the earliest immigrants.
(There evidently exists in Vanua Levu a plant very like the African Musa
Ensete. Its presence was only indicated by the occurrence of its empty
seeds in the stranded beach-drift, and reference is made to it in that
connection in Chapter XXIX.)

(b) _The Polynesian food-plants._—The cultivation of the yams, the
taros, the breadfruits, and the bananas in later ages all over the
Pacific islands cannot here be dealt with. My readers will already know
that a very ancient cultivation is in each case indicated by the
occurrence of a great number of varieties. Much has been written upon
this matter, and amongst the recent contributions to the subject may be
reckoned Mr. Cheeseman’s interesting paper published in the
_Transactions and Proceedings of the New Zealand Institute_ (vol.
xxxiii).

I may here mention in connection with the Sweet Potato (Batatas edulis),
a plant that may have an American origin, though much mystery surrounds
its home, that it rarely seeds in Fiji except when it is grown in poor,
sandy soil, and in dry, rocky situations. The Fijians were quite
incredulous as to its maturing seed; but after much searching I found a
solitary plant in seed and removed their doubts.

THE POLYNESIAN WEEDS

Some curious questions are raised in connection with the weeds of this
region. Polynesia, says Dr. Seemann, presents a most interesting problem
with regard to its weeds. It is, however, necessary to point out that
these plants arrange themselves into two groups, the aboriginal weeds
comprising those existing in the islands at the time of Captain Cook’s
expeditions in the latter half of the eighteenth century, and the white
man’s weeds that have been since introduced.

As concerning the Fijian Islands, Dr. Seemann remarked that although the
majority of the non-endemic plants of the flora is Asiatic, “the bulk of
the weeds is of American origin, or, at all events, is now found in
America.” His principal point was to show that American weeds displayed
a greater disposition than Asiatic weeds to spread in Fiji, because Fiji
was to American plants altogether virgin ground. This is a purely
botanical matter, and we are not in a position to oppose a conclusion
formed by such a careful observer of plant life. But to the ethnologist
it is a very different matter whether most of the Fijian weeds are of
American origin or merely now exist in America. His interest lies
entirely in the aboriginal weeds. To the student of plant-dispersal this
distinction is also a very important one; and his interest again is all
on the side of the aboriginal weeds.

Dr. Seemann enumerates 64 Fijian weeds, of which at least 37 were in the
Pacific islands when Captain Cook’s botanists made their collections
(see Note 82). Of these 22 occur in continental regions on both sides of
the Pacific, 13 are natives of the Old World alone, and two only are
seemingly American exclusively, namely, Waltheria americana and Teucrium
inflatum. The first is claimed to be American because most species of
the genus are American, but it is now widely distributed in the Old
World as well as in America. The second, though widely distributed in
tropical America, has strangely enough only been found in the islands of
the Western Pacific.

The important point is thus brought out that although in Captain Cook’s
time the food-plants cultivated by the Polynesians, such as the banana,
the breadfruit, the taro, and the yams, were almost exclusively Asiatic
in origin and bore Malayan names, a large proportion of the weeds were
not exclusively Asiatic, but occurred in America as well as in the Old
World. The inference to be drawn, however, is not, as Dr. Seemann
implies, that the Polynesians derived several of their weeds from
America (since with few exceptions all the aboriginal weeds named in
Note 82 occur in the Old World, and in more than a third of the plants
in the Old World only), but that many so-called cosmopolitan weeds were
distributed very much as they are now, when the Polynesians brought
their food plants from Indo-Malaya into the Pacific.

Weeds follow the cultivator in all climates; and it is very natural
that, as Mr. Hemsley points out, plants which seem to owe their wide
dispersal to cultivation are not found in Australia (_Bot. Chall.
Exped._, iii, 142). The Australian native as a rule cultivates nothing.
Yet I fancy that man’s share in weed dispersal is as often as not merely
restricted to producing the conditions favourable to the growth of
weeds, and that the seeds are often brought by birds and other agencies.
Many weeds of the genera Atriplex, Polygonum, and Ranunculus are
dispersed by partridges in England, and I have often found the uninjured
fruits of the plants in the stomachs of these birds. Many weeds, like
Prunella vulgaris, Plantago major, Capsella bursa-pastoris, Luzula
campestris, and several others named in Note 43, possess seeds or fruits
that become “sticky” when wet, and would readily adhere to a bird’s
plumage.

We can also say of tropical weeds that many of them are distributed by
birds. In the crop of a dove in Hawaii I found a number of the small dry
fruits of Waltheria americana, the widely spread tropical weed before
mentioned, and of another weed of the order Compositæ. On the bare rocky
peak of one of the Vanua Levu mountains the only plants found growing
were Oxalis corniculata and a species of Peperomia, both of them
evidently growing from seed dropped by birds. The fruits of Urena lobata
and of species of Sida, as well as those of Bidens pilosa and Ageratum
conyzoides, could be readily dispersed, entangled by their appendages in
the plumage of birds, whilst the sticky achenes of Adenostemma viscosum
would easily adhere to feathers. Weeds with drupes or berries like
Geophila reniformis and Solanum oleraceum would attract frugivorous
birds, and I have often seen berries of the last-named pecked by birds.
Man has doubtless often been the agent in dispersing the seeds of
Leguminous weeds like Lablab vulgaris. On the other hand, we know from
the observations of Focke (see page 150) that birds can distribute the
seeds of a plant like Vicia faba; and in the Pacific islands it is
evident from the frequent occurrence of Tephrosia piscatoria on bare
rocky hill tops that its seeds are dispersed through the same agency.
Birds also probably carry about the seeds of Cardiospermum halicacabum.

If we based our conclusion solely on the distribution of weeds without a
previous study of their means of dispersal we might, as students of the
distribution of man, acquire some startling and very erroneous notions
on the history of the races of man, especially in the New World. Lacking
such an acquaintance with existing modes of dispersal it would not be
prudent to attach too much importance to the occurrence of Asiatic weeds
in America and of American weeds in Asia. Mr. Hemsley, in his work on
the botany of Central America (_Biologia Centrali-Americana_), gives a
list of ten British plants of world-wide range, which we will designate
plants of waste places rather than weeds. They are plants often found
not only in the Old and New Worlds, but also in the southern hemisphere,
and I will here name them: Radiola millegrana, Alchemilla vulgaris,
Cotyledon umbilicus, Lythrum salicaria, Convolvulus sepium, Sibthorpia
europæa, Prunella vulgaris, Lycopus europæus, Aira cæspitosa, Luzula
campestris. According to this authority these plants “are most unlikely
to have been aided, intentionally or unintentionally, by man” and
“possess no special means of dispersion by animals or birds or the
elements” in the way, as is implied, of appendages like hooks, hairs, a
pappus, &c.

Five of these plants are referred to in various connections already in
this work. In all I have tested the means of dispersal of six or seven
of them; and although my results are not always conclusive, I venture
here to indicate some of them. The nutlets of Prunella vulgaris and the
seeds of Luzula campestris emit mucus when wetted and adhere firmly to
feathers on drying, whilst the nutlets of Lycopus europæus are sticky in
the dry state and adhere to the fingers on handling. This last-named
plant is occasionally to be noticed on rubbish heaps growing with other
waste-plants. No such adhesive qualities, whether in the wet or dry
condition, came under my notice with Alchemilla arvensis or with Lythrum
salicaria. With Alchemilla the seed-like fruits fall from the plant,
inclosed in the dried-up calyx. The seeds of Cotyledon umbilicus are so
minute (1/3 mm. or 1/75 inch) that they can be compared with Juncus
seeds from the standpoint of dispersal. They are naturally a little
sticky and tend to adhere to feathers, but more probably they are
transported in adherent soil. The case of Convolvulus sepium is a very
remarkable one, and I have referred to it on page 29 and in the notes
there indicated. The species of Radiola, Sibthorpia, and Aira have not
been tested by me. Dispersion, however, would be favoured by the small
size of the seeds in the first two species and by the awned glumes in
the case of Aira.

The distribution of aboriginal weeds might be expected by some to supply
data of profound interest to the student of the races of mankind; and I
think the botanist rarely realises how often he tantalises the
ethnologist by the remark that certain weeds have been spread by
cultivation all round the tropics. De Candolle many years ago, in his
_Géographie Botanique_, gave a list of nearly 100 plants, made up of Old
World species naturalised in America and of American species naturalised
in the Old World, and quite half of them were classed as plants
distributed in one way or another through man’s agency. Now this is
either a subject of supreme importance or it is of no interest to the
student of man’s history. If it should prove that birds have done most
of this dispersal, then the story of the aboriginal weed would be of
little interest in connection with the races of man in the New World.

I will now refer briefly from the standpoint of dispersal to a few
interesting Polynesian plants in which man has been in most cases more
or less concerned in their distribution.

ALEURITES MOLUCCANA (THE CANDLE-NUT TREE)

Much interest is attached to this tree, which is found in India, Malaya,
and North-east Australia, and occurs all over the Pacific, extending
north to Hawaii, south to the Kermadec Islands, and east to Tahiti and
Pitcairn Island (Maiden). In the Hawaiian Islands it is often so
frequent as to form whole forests, or at all events to give a character
to the forest zone up to 2,000 feet above the sea. Its prevalence in
Hawaii might be regarded as evidence of its indigenous character; but
its predominance there is due to the circumstance that it is one of the
few forest-trees that the cattle and other animals avoid, most other
trees falling victim to their depredations by the loss of the bark. In
Fiji, though frequent in places, it does not form such a conspicuous
feature in the vegetation as in Hawaii. In Samoa it is abundant in the
coast-bush. In Rarotonga it forms with Hibiscus tiliaceus, as we learn
from Cheeseman, the major portion of the lower forests, a circumstance
which seems to indicate, since both these trees were probably introduced
by the natives, that this island like Hawaii has lost or is losing many
of its original forest-trees. In Tahiti, according to Nadeaud, it is
common from the sea-level up to 3,000 feet above the sea.

As a Polynesian tree, Aleurites moluccana presents itself to me as an
intruder which has often taken the place of trees of the primeval
forests of these islands. That the natives usually employ the oily seeds
for illuminating purposes is well known; and its prevailing name of
Tuitui (Kukui in Hawaii) is derived from the Polynesian custom of
threading the seeds before using them for lighting purposes. One of the
Fijian names, “Sikethi,” is suggestive of “Saketa,” a name for the tree
in the Ternate dialect of the Indian Archipelago. To the modes of
dispersal of this tree, I have devoted much attention.

The more or less empty seeds of this tree are to be commonly found
floating in rivers and stranded on beaches. I have found them in numbers
on the beaches of Fiji and Hawaii in the Pacific, and of the south coast
of Java and of Keeling Atoll in the Indian Ocean. In all I have examined
many hundreds of these seeds, whether stranded on the beaches in the
localities above named, or floating in the Fijian rivers and at sea
amongst the islands of that archipelago. The seeds were always either
empty or contained a kernel in an advanced stage of decay. A sound seed
has no floating power under any condition; and sound seeds are only to
be found in beach drift near the mouths of estuaries, where they have
been freed by the decay of the fruits brought down by the rivers. During
some dredging operations in the harbour at Honolulu several years ago,
quantities of old Aleurites fruits and seeds were brought up. It is only
by means of the floating fruit that the sound seed can be carried any
distance by the currents; but even in this case the opportunities of
wide dispersal are very limited. If one places in sea-water a number of
well-dried fruits, most of them will sink within a week, and all will be
at the bottom in a fortnight.

The seeds stranded on a beach are often found cracked. This I think
arises from long exposure to the scorching rays of the sun. On account
of the hardness of the shell it is very difficult to obtain the kernel
entire; but the Mangaians get over this difficulty, as we learn from the
Rev. Wyatt Gill, by slightly baking the seeds; whilst the Fijians,
according to Mr. Horne, effect the same object by throwing the heated
seeds into cold water. On one occasion I placed an empty seed on a tin
plate kept at a temperature 115° to 120° F., a temperature near that
which the seed would acquire when lying exposed to the sun on a tropical
beach. After five days I found it had reproduced the cracks noticed in
another empty seed from the Keeling beach.

Facts are not wanting with regard to the dispersal of the seeds by
birds; but since the kernel alone is sought for by birds, and as there
is no means of cracking the shell in their stomachs, such an agency is
only available for local distribution. The Messrs. Layard inform us that
in New Caledonia a small crow (Physocorax moneduloides) and a parrot
(Nymphicus cornutus) are very partial to these seeds (_Ibis_, 1882).
They were told that the crow cracked them by carrying them to a
considerable height and letting them fall on a stone. We are not told
how the parrot cracks the seed, which has a shell so hard that the
Malays, I may remark, term the seed “bua kras,” or “the hard seed,”
whilst a hammer is required to break it. However, since Indian parrots,
according to Mr. J. Scott, are able to split open with their beaks the
hard beans of Adenanthera pavonina (_More Letters of Charles Darwin_,
ii, 349), they evidently possess ingenuity in seed-cracking.

My general conclusion with reference to this tree in Polynesia is that
it could not have been distributed, except locally, by birds and
currents; and that it owes its dispersion there principally to man. A
contrary indication seems to be offered by the occurrence of the tree in
the uninhabited Kermadec group; but since Cordyline terminalis also
exists there, a cultivated plant widely dispersed by the Polynesians, it
would appear that these islanders have formerly visited the group. It is
also contended by Canon Walsh that the Cordyline of the Maoris was
introduced into New Zealand by that race. (See Cheeseman in vols. xx and
xxxiii, _Trans. N.Z. Inst._, for papers on the Kermadec flora and on the
food-plants of the Polynesians.)

INOCARPUS EDULIS (THE TAHITIAN CHESTNUT)

Like Aleurites moluccana this tree presents a _primâ facie_ case for
dispersal by currents. As the result of inquiries in this direction I
have formed the opinion, however, that it has been mainly distributed by
man. Though occurring in all the South Pacific groups, as far east as
Tahiti and the Marquesas, it does not occur in Hawaii. With its home in
Malaya it possesses a range closely resembling that of the breadfruit
tree; and yet, although its fruits are often a common article of food in
Polynesia, it requires no cultivation, and reproduces itself so
abundantly in favourable situations that, as Dr. Seemann observes, only
the dense shade of the parents checks the occupation by the seedlings of
all the adjacent ground. It possesses in the Pacific two sets of names,
neither of which I have been able to identify with any Malayan names,
and both occur over much of the region. Thus the Fijian “Ivi” and the
Samoan and Tongan “Ifi” are represented by “Ii” in Rarotonga, “Ihi” in
Tahiti and the Marquesas, “Hi” in Ualan in the Carolines, “Ifi” in
Futuna in the New Hebrides, and “If” in a New Guinea dialect. Then we
have the Tahitian “Mape,” the “Marap” of Ponape in the Carolines, and
the “Mamape” of Fate in the New Hebrides, besides other forms found in
Melanesia.

In the South Pacific islands, as in Fiji, Samoa, Rarotonga, and in the
Tahitian group, it flourishes in low, moist localities at and near the
coast, by the side of streams and estuaries, and in the rich soil of the
lower valleys. In the Rewa delta in Fiji it is especially abundant,
often bordering the creeks in the mangrove swamps, and occupying
stations that are under water when the river is in flood. It may extend
inland in the various groups, but it is in the low-lying, moist, coast
regions that it mostly thrives; and in Fiji it presented itself to me as
essentially a tree of the estuaries, a station strongly suggestive of
dispersal by currents. Schimper, it may be remarked, includes it amongst
the shore vegetation of the Indian Archipelago.

When in Fiji I paid especial attention to the dispersal by currents of
these large fruits, the agency of birds being, of course, negatived by
their size. They are to be commonly observed floating in the rivers when
in flood, as well as at sea between the islands, and stranded on the
beaches. Of those found afloat in the Rewa River not more than a fourth
had a sound seed. Of those stranded on the beaches two-fifths were
empty, two-fifths displayed a rotten seed, and one-fifth had sound
seeds. Of those picked up at sea all were empty. These fruits, unlike
many others in the drift of the Fijian rivers, do not germinate afloat.
They soon lose in the water their outer, fleshy, non-buoyant coat;
whilst the inner fibrous coat, to which the floating power of the fruit
is due, the seed having no buoyancy, is not water-tight, and moisture
soon enters and leads to the decay of the seed. In order to test their
floating power, I placed in sea-water ten mature fruits. Five of them
floated after forty-five days, having then lost most of the outer,
fleshy coat. Two were afloat after sixty days, but their seeds were
rotting. One fruit that sank after five weeks had a sound seed. Most of
them were sown out afterwards in a place where the trees were thriving,
but none germinated, and of two or three examined all had a decaying
seed. The empty fruits may float a long time after the decay of the
seed. Forty days would probably be the extreme limit for the flotation
in sea-water of a fruit with a seemingly sound seed, though a very small
proportion would reach this limit, and I much doubt whether such a fruit
would germinate afterwards.

I, therefore, inferred that currents are only available for the local
dispersion of the fruits of Inocarpus edulis. It is to man that the tree
owes its existence in Tahiti and other groups of the open Pacific; and
it is to be concluded that the occurrence of this tree on Christmas
Island in the Indian Ocean marks an early Malayan occupation of the
island.

GYROCARPUS JACQUINI

The cosmopolitan distribution of this seemingly useless tree, growing,
as Hemsley remarks, in maritime districts throughout the tropics, in
America, Australia, Asia, and Africa, presents one of the puzzles of
plant-distribution. It is by no means universally spread in the Pacific
islands, and I find reference to it only in Fiji and Tahiti. Seemann
says that in Fiji it is common on the beaches of Taviuni and other
islands. I found it to be a rare coast tree on Vanua Levu. It does not
seem to have been recorded by the botanists of the 18th century in the
Pacific. It, however, has evidently been long established there. Nadeaud
does not speak of its littoral station in Tahiti, and says that it grows
best in the regions of the interior up to elevations of 2,000 feet,
where it attains a great size; and its abundance is implied by his
remark that he had to fell many trees to collect the fruits.

The singular fruit, which has two long wings and looks like a
shuttlecock, dries up on the tree; and in course of time it is detached
and falls to the ground. The falling fruit in its descent twists round
like a screw, and hence the Fijians call the tree the Wiri-Wiri tree,
the same name in the form of Wili-Wili being given for a similar reason
to Erythrina monosperma in Hawaii. Schimper (p. 157) truly remarks that
the fruits are too heavy to be carried by the wind across a wide extent
of sea; and I ascertained by experiment that in an ordinary trade-breeze
they would only be carried a few paces. Birds are quite out of the
question as agents of their transport to oceanic islands. We are driven
then either to the agency of man or to that of the current. The trees
grow rapidly and the timber is soft and perishable. The fruits are not
edible, and as far as I could ascertain the tree is of little or no
value to the Pacific Islander, there being at all events no reason to
believe that he has distributed it.

We appeal lastly to the currents, the agency which Mr. Bentham selected
on _a priori_ grounds (Presidential address, _Linnean Society_, 1869).
My experiments in Fiji showed that the fruits, when dried on the tree
and afterwards detached, are able to float over long distances in
sea-water. After two months they were still afloat, the seeds inside
being dry and unharmed. The fruit’s buoyancy was tested in different
conditions, either without the wings, or with both wings, or with but
one wing, and it was found that the wings, which float for only a day or
two by themselves, lessen the buoyancy of the fruit. Of fruits with both
wings attached forty per cent. floated after two months, whilst of those
deprived of the wings all floated after two months. In the ordinary
course of flotation the wings in most cases break off during the first
few weeks, and in the rough-and-tumble of current-transport this would
occur sooner, so that the floating power of most of the fruits would not
be much affected. The cause of the buoyancy in a structural sense
belongs to the Convolvulaceous type. The kernel has no buoyancy, but it
incompletely fills the cavity of the seed-vessel, the coats of which are
quite waterproof, but have no independent floating power.

It is thus evident that like many other shore-trees Gyrocarpus Jacquini
is distributed by the currents. It is not unlikely that its present
sporadic occurrence in the Pacific islands may be due to the gradual
extinction of the tree in this region, either on account of some insect
pest introduced since Cook’s time or from the use of the timber for
fire-wood by the aborigines.

SERIANTHES MYRIADENIA

This is a striking looking Acacia-like tree that might have been fitly
discussed in the chapter on the enigmas of the Leguminosæ. Only four or
five species are named in the _Index Kewensis_, of which one occurs in
Malacca and in the Philippines, a second in New Caledonia, a third in
Fiji, and the fourth, S. myriadenia, over the South Pacific groups of
Fiji, Tonga, and Tahiti. Reinecke does not include the genus in the
Samoan flora; and it is merely assigned to that group by Seemann on the
authority of Mr. Pritchard, the British Consul in Fiji. Though common in
the forests of the larger islands of Fiji, S. myriadenia is most at home
on the banks of the estuaries, usually behind the mangrove belt, but not
beyond tidal influence. The peculiar species, S. vitiensis, I found on
the banks of the estuary of the Mbua River in Vanua Levu, the locality
from which Gray described it. According to the French botanists, S.
myriadenia, in Tahiti, ranges from near the sea to an elevation of 800
metres. The Fijian name of the trees is “Vaivai,” the name also of
Leucæna Forsteri, and of some other introduced trees of the Acacia
habit. The Tahitians apply the same name in the form of “Faifai” to S.
myriadenia.

The Fijians value the trees on account of the wood; but unless the
Polynesians were in the habit of transporting the seeds of their
numerous timber trees, which is most unlikely, it seems at first sight
useless to look to man’s agency for an explanation of the wide dispersal
of a tree like S. myriadenia in the South Pacific. The tough, woody,
indehiscent pods, from 3-1/2 to 4 inches long, floated in my sea-water
experiments in the case of both S. myriadenia and S. vitiensis between
seven and twenty-five days, after drying for some months. The seeds,
about two-thirds of an inch (17 mm.) in length, are only freed by the
decay of the fallen pod, and have no buoyancy. The agency of birds is
evidently excluded; and it is, therefore, to the currents that we must
make our final appeal; but their powers of dispersing the species appear
quite insufficient to explain the occurrence of these trees in Tahiti.
Perhaps, as in the case of Calophyllum spectabile, another Polynesian
timber-tree found in Tahiti (see p. 136), man and the currents have
worked together.

LEUCÆNA FORSTERI

This bush of the Mimoseæ frequents maritime sands in the South Pacific,
and is confined to this region. It has been found in New Caledonia,
Fiji, Tonga, Rarotonga, and Tahiti. The seeds sink and the pods dehisce
on the plant, so that the agency of currents, unless we invoke the
intervention of the drifting log, bearing the seeds in its crevices,
seems to be excluded. Sea-birds might carry the seeds unharmed in their
stomachs, but there is no evidence bearing on birds as agents in the
dispersal of the species. Since the plant has not been recorded from
localities outside the Pacific islands, and since it was collected by
Cook’s botanists in Tonga and Tahiti, it cannot be placed amongst plants
of recent introduction. Although growing on maritime sands in Fiji,
Rarotonga, and Tahiti, it may grow inland, and according to Cheeseman is
particularly abundant in Rarotonga. In Fiji it is apt to occupy
newly-formed alluvial land at the mouth of the rivers, as in the case of
the Rewa; but the “how and why” caused me much fruitless speculation,
and I abandoned the plant in despair. The Fijians sometimes give it the
native name of Serianthes myriadenia, which they then term “Vaivai ni
Viti,” or the Fijian Vaivai. In Tahiti it is named “Toroire,” and in
Tonga “Toromiro.”

MUSSÆNDA FRONDOSA

Mussænda frondosa is the only one of the sixty species of this tropical
Asiatic and African genus that extends into Polynesia. This beautiful
shrub, which is easily recognised by its conspicuous white, leaf-like
calyx lobe, is common everywhere in Fiji, decorating, as Horne fitly
remarks, in the contrast presented by its golden flowers, its large
white calyx leaf, and its green foliage, many an acre of waste, grassy
land, where the orange-coloured doves and the red and the green parrots
flit to and fro. With its home in India, China, and Malaya, it ranges
all over the South Pacific, from the Solomon Islands to Tahiti. Its
berries contain an abundance of small, minutely-pitted seeds, 0·7 mm. or
1/35 of an inch in size, and weighing when well dried about 600 to the
grain. The seeds retain after years of drying the property of clinging
to passing objects by means of a few microscopic, thread-like fibres,
that are attached to their surfaces. In this manner they will fasten
themselves to the point of a knife, and the observer is astonished to
see them dangling in the air from a pin’s point. I suppose that this is
connected with some hygroscopic quality. At all events, it would enable
these light seeds to be carried about not only by birds and bats but
also by insects. It is possible that man has aided in the dispersal of
this interesting plant; but birds, bats, and insects have, I think,
mainly done the work.

LUFFA INSULARUM

This is regarded as a maritime form of Luffa cylindrica, a plant
commonly cultivated throughout the tropics. The South Pacific plant,
which occurs also in Australia and Malaya, has been found in New
Caledonia, Fiji, Tonga, Rarotonga, and Tahiti. In Fiji it grows chiefly
on the “talasinga” plains and in places once under cultivation. I
noticed it in one locality climbing over the branches of an Inocarpus
tree on the banks of the Rewa. In Rarotonga it is common in the lower
regions. It is, according to Nadeaud, fairly frequent on the shore and
in the lower valleys of Tahiti, where it was collected by Banks and
Solander, the companions of Cook. The Pacific islanders, as far as can
be gathered, make little or no use of the plant; and unless it was
introduced accidentally with their cultivated plants, they could
scarcely have been concerned in its dispersal.

In Fiji I made a special point of investigating the mode of dispersal of
this plant. The fruits, which ultimately become dry and fibrous, are to
be seen hanging vertically from the plant as it climbs among the
branches of a tree. The apical disk usually falls off, and many of the
seeds drop out through the hole thus produced; but a few remain
entangled in the fibrous material occupying the interior of the fruit. I
have noticed such fruits floating down the stream of the Rewa River; but
my experiments showed that they do not float more than a week, whether
in fresh or salt water. The seeds, however, possess a hard, impervious
shell, and are well adapted to withstand unharmed prolonged immersion in
the sea. They will evidently float for months. Out of one hundred
selected seeds placed in sea-water, sixty were found afloat and sound
after two months. The cause of the seed-buoyancy is purely mechanical.
Neither the shell nor the kernel has any floating power, the buoyancy
arising, as with Convolvulaceous seeds, from the unfilled space in the
seed-cavity. When in Fiji, I tested the seeds of the ordinary cultivated
tropical form of the plant which had been introduced into a garden from
Australia. They all sank in a few days, and on being cut across the seed
displayed but little unoccupied space in its cavity. I have no doubt
that the Pacific form of this plant has been at times dispersed by the
currents, not, however, through the fruits, but through the seeds. It is
also quite possible that it may have been introduced by a pre-Polynesian
people into the Pacific.

_Summary of the Chapter_

(1) Man in his distribution over the Pacific islands reproduces, but in
a less degree, nearly all the difficulties presented by the plant in its
dispersal. In both we have the age of general dispersion followed by a
suspension more or less complete of the migrating movements; and in both
we have differentiation associated with the isolation.

(2) The Pacific islanders possess two sets of food-plants. In addition
to those commonly cultivated in our own time, such as the yam, the taro,
the banana, &c., there are a number of food-plants now growing wild, but
rarely cultivated, and only used when the others fail. These plants,
which include the wild yams, the mountain bananas, Tacca pinnatifida,
Pandanus odoratissimus, and several others, are regarded as older than
the Polynesians in the Pacific, and as having probably formed the food
of a pre-Polynesian race that practised only a rude sort of cultivation.

(3) The weeds of Polynesia also fall into two groups. In the first place
there are the aboriginal weeds, of which those found in this region by
Captain Cook’s botanists in the latter part of the 18th century are
taken as examples. These include species of Urena and Sida, besides
Waltheria americana, Oxalis corniculata, Bidens pilosa, and many other
weeds. In the second place, there are the numerous weeds that are known
to have been introduced by the white man since the voyages of the
English and French navigators of Captain Cook’s time.

(4) There is reason to believe that many weeds now cosmopolitan in the
tropics had obtained their present distribution in America and in the
Old World before the Polynesians entered the Pacific. It is thus that we
can explain how there existed in these islands at the time of their
discovery by Cook, Bougainville, and other navigators of that period, a
number of weeds that have their homes in America.

(5) It is not considered that the distribution of aboriginal weeds can
materially aid the ethnologist in his study of the early history of man,
since birds are regarded as the chief distributors of their seeds and
fruits. Whilst man has prepared the conditions for the growth of weeds,
the bird has usually brought the seeds.

(6) Amongst interesting plants concerned with man in the Pacific are
Aleurites moluccana and Inocarpus edulis, which are regarded as in the
main distributed through man’s agency. Gyrocarpus Jacquini is viewed as
a tree originally widely dispersed by the currents in the Pacific, but
now becoming extinct.

CHAPTER XXIX

BEACH AND RIVER DRIFT

THE BEACH DRIFT OF TEMPERATE LATITUDES

DISPERSAL by currents seems to be mainly restricted to warm latitudes.
Whilst in the tropics seed-drift is abundant on the beaches, in the
cooler regions of the globe it is usually very scanty and often masked
by other vegetable _débris_.

Let us take, for instance, a beach in the south of England. We can find
by careful searching amongst the stranded drift the seeds and
seed-vessels of various littoral plants of the buoyant group, such as
Arenaria (Honckeneya) peploides, Cakile maritima, Crithmum maritimum,
Convolvulus soldanella, Euphorbia paralias, &c., and such sundries as
bits of stems of Salsola kali bearing fruits; but their amount is
scanty; and they are often difficult to find on account of the great
amount of rubbish with which they are associated, such as empty stones
of cherries, plums, and peaches; empty seeds of grapes; hazel-nuts,
beech-nuts, chestnuts, acorns, all either empty or with decaying seed;
the spiral pods of Medicago; besides quantities of leaves, sticks, and
bark. Although the occasional shell of a Spirula, or the horny skeleton
of a Velella, or a genuine pumice pebble (see Note 76), may tell us of
long wanderings in mid-ocean, we find little that is not English or
derived from neighbouring coasts on a beach in the south of England. I
have examined numerous beaches on the coasts of Devon and Cornwall, and
have never come upon any indubitable tropical seed-drift.

On one occasion I examined many of the beaches between Ilfracombe and
Padstow with the object of finding tropical seeds, but to no purpose.
Portions of bark, generally 2 to 4 inches across and much water-worn,
together with a quantity of steamer-slag or cinders, often largely
composed the stranded drift. No doubt this bark is stripped off by the
waves from floating trees, which are generally stranded in a bare
condition after a long ocean voyage. This is the case with the timber
brought in the Oregon drift to Hawaii; and Sernander (p. 117) remarks
that bark seldom occurs on the trees washed ashore with the Atlantic
drift on the coasts of Scandinavia. Modern marine deposits ought to
contain much bark _débris_.

On the beaches in the vicinity of estuaries we find a certain amount of
river drift, and amongst it fruits or seeds of Sparganium ramosum, Iris
pseudacorus, Alnus glutinosa, Rumex, and many other river-side plants,
such as I have mentioned in my paper on the Thames drift (_Journ. Linn.
Soc. Bot._, xxix). Most of them are capable of reproducing the plant,
but not on the sandy beach where the waves have stranded them; and we
thus see here one of the limits of the efficacy of currents as
seed-dispersers.

From the labours of Lindman and Norman, the results of which are summed
up by Sernander (p. 116) we can learn what are the components of the
“Atlantic Drift” on the Scandinavian coast; and a strange assortment we
here find, in which it is difficult to detect much indication of
effective seed dispersal. Besides the seeds of Cæsalpinia Bonducella,
Entada scandens, and Mucuna urens, familiar to us as occurring in the
drift of tropical beaches, there is a quantity of vegetable drift
hailing sometimes from North America, sometimes from the Canary Islands,
and sometimes from the West Indies, mingled with much local drift in
which the larch and steamer-slag or cinders predominate. The seed-drift
derived from the proper beach-plants of the coast plays a subordinate
part, though it is stated by Norman and others that seeds and
seed-vessels, as the case may be, of Arenaria peploides, Cakile
maritima, Convolvulus soldanella, and Lathyrus maritimus, with those of
other plants, are also to be found.

The Mediterranean beach drift, as illustrated by the results of my
examination of numerous beaches in Sicily as well as in the islands of
Stromboli and Lipari, and of the beach at Cumæ, is of a scanty nature.
If we eliminate the various evidences of cultivation which seem to occur
over much of the temperate regions of the globe, very little remains of
an interesting character. As in the south of England and in other
regions, the empty stones of the cherry, plum, and peach, the empty nuts
of the oak, hazel, &c., together with the spiral pods of Medicago figure
largely in the drift; and here and there we come upon the seeds of
littoral plants, such as Convolvulus soldanella and Euphorbia paralias.

I have found Medicago fruits in all these localities on the beaches.
They often contain seeds, which, it may be added, have no buoyancy, the
seeded pods themselves floating from two to five days. The pods of
several kinds of Medicago form the great feature of Sicilian drift and
are often indications in other places of the vicinity of cultivated
districts. A small hairy species thrives on Letojanni beach near
Taormina, and I observed its seeds together with those of Euphorbia
paralias germinating in the drift stranded on the same beach. Arcangeli,
in his _Flora Italiana_, enumerates as many as thirty-three species of
Medicago. Many of the species grow in maritime districts, and their
fruits must often get into the beach drift independently of cultivation.
I noticed the pods amongst the drift brought down by the Alcantara, a
river near Taormina, a fact which goes to explain their presence in
beach drift.... On the beach of Trogilus Bay, near Syracuse, I gathered
several fruits of a Vitex, apparently V. agnus castus. After being kept
afloat for six weeks in sea-water some were placed in soil, when they
soon germinated and reproduced the plant.

The beach drift of temperate Chile is described in Chapter XXXII. There,
as in other beaches of cool latitudes, it is not easy to find seeds
amongst the rubbish; but amongst the scanty seed-drift may be recognised
much of what we are familiar with in the Old World, such as the seeds of
Convolvulus soldanella, bits of the fruiting stems of Salsola kali, as
well as the rubbish indicating the white man’s presence, such as empty
stones of cherry, plum, and peach, Medicago pods, &c. In addition, we
find the seed-vessels of plants like Franseria and Nolana that are
peculiar to American beaches; and now and then, the seeds of Sophora
tetraptera, a tree of the immediately adjacent hill-slopes, come under
our notice.

Before quitting this subject of the beach seed-drift of temperate
latitudes, it may be observed that when at San Francisco I visited the
beach running south from the Golden Gate. With the exception of the
fruits of Cakile maritima, a plant growing on the beach, few other seeds
or fruits were observed in the drift.

The inference that there is very little effective dispersal by currents
in temperate regions is of some importance, and Sernander arrived at a
similar conclusion when discussing the origin of the Scandinavian flora.
The few plants with buoyant seeds and fruits, such as Arenaria
peploides, Cakile maritima, Crithmum maritimum, Convolvulus soldanella,
Euphorbia paralias, and Lathyrus maritimus, are no doubt thus dispersed,
and Norman is quite right in attaching some value to the distribution by
currents of certain plants within the region of the Arctic flora; but
after all it amounts to little, and geographical and climatic conditions
have often had a predominant influence in determining the distribution
in the temperate latitudes of littoral plants possessing buoyant seeds
or fruits.

Nowhere is this shown more plainly than with the littoral plants with
buoyant seeds or seed-vessels that are found on our English beaches.
Some have evidently acquired their present distribution before ice and
snow reigned supreme in the extreme north. Though it may be possible, it
seems highly improbable, that either Arenaria peploides or Lathyrus
maritimus, both of which occur on beaches in high northern latitudes in
the Atlantic and Pacific Oceans (as in Arctic Norway, Spitzbergen, and
Behring’s Straits), could possess in our own day any means of
communication between their areas of distribution on the borders of
these two ocean-basins.

So again with Cakile maritima, the occurrence of this or of two closely
allied species on both sides of North America cannot be attributed to
any present working of the currents for two reasons. In the first place,
as is remarked in Note 18, the results of two independent experiments
made by me show that the fruits will not float more than a week or ten
days in the sea, a capacity that will not admit of their transportation
by the currents over tracts of ocean more than one or two hundred miles
across. In the second place, this species is not an Arctic plant like
Arenaria peploides and Lathyrus maritimus; and the possibility of
inter-communication between the Atlantic and Pacific Oceans having any
effective value from the standpoint of dispersal, shadowy as it is with
the two Arctic species, is still more so in the case of Cakile maritima.
Norman’s observations on the coast of Norway, as quoted by Sernander
(page 123), indirectly indicate how hopeless it would be for this plant
to attempt to traverse the Arctic region. Just as I have noticed on the
north coast of Devonshire, the fruits occur plentifully in the beach
drift and germinate freely in the upcast wrack as far north as Senjen in
latitude 69°. Further north the plant has been recorded from only eight
localities, and since it is there sterile and but a summer annual, the
seed-vessels, it is argued, must have been brought by the currents from
the south.

The reference to Cakile maritima as a summer annual on the north coast
of Norway is of interest; but I may point out that it displays a similar
behaviour in England on the north coast of Devonshire. Here, during the
latter half of July, 1903, I found the fruits common in the stranded
drift, and often in a germinating condition, whilst numerous seedlings
one to two inches high with the fruit-shell still attached were growing
out of the sand. From this arises the curious reflection that an annual
which germinates in the end of July could scarcely be expected to mature
its fruit before the winter. It would seem that this beach plant hampers
its own dispersal by its misdirected efforts; and the idea suggests
itself that we have here the explanation of its sterility in the north
of Norway. Had it been a perennial like Arenaria peploides and Lathyrus
maritimus it might have had a similar distribution within the Arctic
Circle.

Quite other considerations seem to be suggested by the perennials
Crithmum maritimum and Euphorbia paralias. In these cases, although the
seeds or fruits, as the case may be, will float for months in sea-water
without apparently sustaining any injury, the species are confined to
the warmer parts of the European region.

From Convolvulus soldanella we obtain another story. Its occurrence in
the temperate regions of both the northern and southern hemispheres,
great as the floating powers of the seeds may be, is concerned with
something more than with questions relating to modes of dispersal. The
circumstance that in its distribution in the temperate regions it is
practically coterminous with Ipomœa pes capræ in the tropics is very
significant (see Note 49).

Each one of the English beach plants with buoyant seeds and fruits has
its own story of the past to tell. Time has indeed gathered on our
beaches current-dispersed plants, which, if they could speak, would tell
us strange stories of many latitudes, stories of change within the
Arctic Circle, and stories of great events within the temperate regions,
and, as in the case of Convolvulus soldanella, stories of a past within
the tropical zone. It cannot be said that investigators lack clues
leading to lines of inquiry into the age that immediately precedes our
own.

Yet valuable as our British plants would be for this purpose, they do
not afford any indication that currents have played an important part in
plant distribution in temperate and arctic latitudes. Ekstam strikes the
true note for these regions when discounting the agency of currents in
the instance of the Spitzbergen flora, he regards the wind as the
greatest factor in seed-dispersal and after that the bird. The several
interesting points raised by this botanist are discussed in Chapter
XXXIII.

THE BEACH-DRIFT OF TROPICAL LATITUDES.

Tropical beaches, as a rule, present a much greater abundance and
variety of stranded seeds and fruits than we find on beaches in
temperate latitudes. Observers in different parts of the tropics have
alluded to the enormous amount of vegetable drift floating in the sea
off the coasts, particularly in the vicinity of estuaries. Though much
of it is brought down by rivers, a good proportion is also derived from
the luxuriant vegetation that lines the beaches. Gaudichaud speaks of
the immeasurable quantity of drift (trees, branches, leaves, flowers,
fruits, and seeds) floating amongst the islands of the Molucca Sea; and
Hemsley, who quotes this author, gives other facts illustrating the same
point. Moseley tells us that seventy miles off the coast of New Guinea,
H.M.S. _Challenger_ found the sea in places blocked with drift (_Bot.
Chall. Exped._ iv. 279, 284). When the author of this book was in the
Solomon Group, long lines of vegetable drift were frequently observed
floating among the islands. The Rewa River in Fiji carries down a great
amount of drift to the sea; and as described in Chapter XXXII, the
Guayaquil River in Ecuador bears seaward an enormous quantity of these
materials.

When we come upon this floating drift out at sea off an estuary, we
find, as Mr. Moseley pointed out, that the leaves have gone to the
bottom, whilst the floating islets, composed of the matted vegetation
lifted up from the shallows of a river channel, which form such a
feature in the Guayaquil River, have been dispersed or sent to the
bottom. However, a very large proportion of the seed-drift brought down
by a river from the interior has no effective value for the purposes of
dispersal. Many of the fruits and seeds brought down from inland owe
their presence in river-drift entirely to the buoyancy acquired by the
decay of the seeds. It is in its lower course when it traverses the
mangrove belt that a river picks up most of the material that is of
service in distributing the species; and this is mingled out at sea with
the numerous buoyant seeds and fruits of littoral plants that are swept
off the beaches by the currents.

A description is given in Chapter XXXII of the enormous amount of
vegetable drift brought down by the Guayaquil River to the coast of
Ecuador. Besides the huge tree-trunks and the floating Pistias, we
observe large islets formed mainly of Pontederias and Polygonum,
together with a host of seeds and seed-vessels, both large and small,
including those of Anona paludosa, Entada scandens, Erythrina, Hibiscus
tiliaceus, Ipomœa, Mucuna, Vigna, &c., accompanied by the empty seeds of
Phytelephas macrocarpa and of many other strange plants from the slopes
of the Chimborazo mountains. In addition, we notice the seedlings of
Avicennia and of Rhizophora mangle together with the seeded joints of
Salicornia peruviana and the germinating fruits of Laguncularia.

When in Fiji I made an especial study of the drift of the Rewa Estuary
within tidal influence, the results of which are incorporated in various
parts of this work. In the rainy season, when the drift is most
abundant, the following would be its most characteristic components:

Seedlings of Bruguiera and Rhizophora.
Fruits of Barringtonia racemosa and B. speciosa, the first-named most
abundant and often germinating.
Seeds of Carapa obovata, most of them far advanced in germination.
Fruits of Lumnitzera coccinea.
Fruits of Cerbera odollam, abundant.
Fruits of Inocarpus edulis, with the seed generally rotten.
Fruits of Heritiera littoralis, Parinarium laurinum, and Pandanus.
Empty seeds of Aleurites moluccana.
Fruits of Scirpodendron costatum, abundant.
Fruits of Clerodendron inerme and Smythea pacifica, both of them in
some cases germinating.
Pyrenes of Morinda citrifolia.
Small fruits of Vitex trifolia and Premna taitensis, both sometimes
abundant.
Seeds of Entada scandens, Mucuna, and Vigna lutea.
Pods of Dalbergia monosperma and Derris uliginosa, the last sometimes
in a germinating condition.
Seeds of Hibiscus tiliaceus and of different species of Ipomœa, such as
I. peltata and I. pes capræ.

Amongst other seeds and fruits brought down by the Fijian rivers and
stranded with a large amount of miscellaneous vegetable _débris_ on the
beaches in the vicinity of the estuaries are the seeds of Dioclea,
Strongylodon lucidum, and Afzelia bijuga; the empty seeds of Musa Ensete
(as identified with a query at Kew); the empty stones of the Sea tree,
apparently a species of Spondias; the seeds of Colubrina asiatica; the
fruits of an inedible indigenous Orange (Citrus vulgaris?) referred to
in Chapter XIII; the cocci of Excæcaria Agallocha and Macaranga; and
Coco-nuts.

The occurrence in Fijian beach-drift of the seeds of Musa Ensete, or of
a wild banana much like it, is very remarkable. This species is found in
the mountains of Abyssinia and on the slopes of Kilima-njaro in
Equatorial Africa; but according to the monograph by Schumann on the
Musaceæ (Engler’s _Pflanzenreich_, 1900) the species is confined to
Africa, whilst all the other species of the subgenus are mostly
restricted to the same continent with the exception of one or two in
Further India. The empty seeds are frequent on the beach at Duniua at
the mouth of the Ndreke-ni-wai in Savu-savu Bay, Vanua Levu, and are
doubtless brought down by that river. Strangely enough the natives could
give me no information about the parent plant which I never discovered.
The seeds did not come under my notice in any other locality in Fiji.
They answer to the description and to the figure given by Schumann for
Musa Ensete; and their presence in the drift is one of the mysteries of
the Pacific floras.

To enumerate the seeds and fruits found stranded on beaches in Fiji
would be to give a list of all the littoral plants with buoyant seeds or
fruits that are included in the list given in Note 2. I may here allude
to the fact that the Coco-nut, whether brought down by a river or
transported by a current, is able to germinate and establish itself when
washed up on the Fijian beaches. I have found these fruits germinating
amongst the drift stranded on the beaches near the mouths of rivers,
some just beginning to germinate and others already striking into the
sand and showing the first leaves. White residents living for years in
one locality were quite convinced that this frequently happens. One of
them pointed out to me some newly formed land at a river’s mouth, not
over two years old, on which were growing young plants three or four
feet high of Barringtonia speciosa, Calophyllum Inophyllum, and several
other plants including young Coco-nut palms, all growing from fruits
washed up by the waves and therefore self-sown.

Like the littoral flora the beach-drift proper to the Hawaiian Islands
is very scanty. This is due to the scarcity of rivers, to the absence of
the mangrove-formation from which much of the drift is derived in other
tropical regions, and to the paucity of shore-plants with buoyant seeds
or fruits. As is observed in Note 30, where the composition of the beach
drift is described, the presence of a large amount of timber and of
other materials brought by the currents from the north-west coast of
America masks much of the local drift.

Remarks on the beach-drift of the Panama Isthmus, and of the Ecuadorian,
Peruvian, and Chilian coasts of South America will be found in Chapter
XXXII. I have examined beach drift in other tropical regions, as in the
Solomon group, on Keeling Atoll, and on the south coast of West Java;
whilst there are at my disposal the data supplied by Schimper and Penzig
for the Malayan region including Krakatoa, and by Hemsley for tropical
regions generally. It will, I think, be best, if instead of describing
in detail the composition of the drift for each locality, I refer
briefly to the features that distinguish the tropical beach-drift of the
Old World from that of the New World.

The beach-drift reflects the characters of the coast flora; and since
tropical littoral floras belong to two great regions, the Asiatic
including Polynesia and the African East Coast, and the American
including the African West Coast, the seeds and fruits stranded on the
beaches may be similarly referred to the same two regions.

All over tropical Asia, as well as in the tropical islands of the Indian
and Pacific Oceans, the drift stranded on the beach presents the same
general character, and as a rule possesses seeds and fruits of the same
species that range over the whole or the greater part of this region.
Almost everywhere we find seeds or fruits of the same plants of the
beach formation, such as Barringtonia speciosa, Cæsalpinia Bonducella,
Calophyllum Inophyllum, Canavalia obtusifolia, Cerbera Odollam, Cordia
subcordata, Entada scandens, Guettarda speciosa, Hernandia peltata,
Hibiscus tiliaceus, Ipomœa pes capræ, Mucuna, Scævola Kœnigii, Sophora
tomentosa, Terminalia Katappa, and Tournefortia argentea. In those
localities where mangrove-swamps occur we find generally diffused in the
stranded drift of this region the seedlings of Bruguiera and Rhizophora,
the seeds of Carapa moluccensis, the fruits of Heritiera littoralis and
Lumnitzera coccinea, and the pods of Derris uliginosa. Amongst sundries
found over much of this region may be mentioned, the drupes of Pandanus,
the seeds of Erythrina, Vigna lutea, and Hibiscus tiliaceus, and the
“nuts” of Aleurites moluccana. With the exception of the last-named all
the fruits and seeds here enumerated are effectively dispersed by
currents over great areas. The sound nuts of Aleurites have no buoyancy;
and the nuts only acquire their floating power through the decay of the
kernel (see p. 419).

The beach drift of the American region, a region which comprises both
the Pacific and Atlantic coasts of tropical America as well as the
African West Coast, has some features in common with the Asiatic
beach-drift and other features peculiar to itself. The plants, however,
that are represented in the drift of both regions are comparatively few,
and none of the large fruits of the Asiatic region are here to be
noticed. We observe, however, that the drift of the two regions possess
in common the seeds of Cæsalpinia Bonducella, Canavalia obtusifolia,
Entada scandens, Erythrina, Mucuna, Sophora tomentosa, and Vigna lutea,
all belonging to the Leguminosæ; and to these we must add the seeds of
Hibiscus tiliaceus and of Ipomœa pes capræ, and the seedlings of
Rhizophora and Avicennia. (Avicennia occurs in tropical Asia, but not in
Polynesia.) The distinctive characters of the beach-drift of both coasts
of America and of the west coast of Africa would be shown in the
presence of seeds of Anona paludosa, the fruits of Laguncularia
racemosa, Conocarpus erectus, Spondias lutea, and other plants. But the
beach-drift of the American region is much more scanty. Of the shore
plants generally dispersed in this region there could not be more than a
couple of dozen that are indebted for their wide dispersal to the
currents, and these alone figure in the effective beach drift. In the
Asiatic region these plants would number at least seventy or eighty.

_Summary._

(1) Effective dispersal by currents is mainly restricted to warm
latitudes, as is indicated by the scanty character of the seed-drift
stranded on the beaches of the south of England, Scandinavia, the
Mediterranean, and Southern Chile.

(2) The present distribution in temperate latitudes of littoral plants
possessing buoyant seeds or seed-vessels is to be attributed more to the
influence of geographical and climatic conditions than to the agency of
currents. With some of them, such as those that occur on both sides of
North America, it is evident that their distribution antedates the
present climatic conditions within the Arctic Circle.

(3) Time has gathered on an English beach current-dispersed plants that
could tell us strange stories of many latitudes.

(4) The seed-drift that is often found in such abundance in tropical
seas is partly brought down by rivers and partly swept off the coast.
Very little of the seed-drift brought down by the rivers from the
interior is of any service for plant-dispersal, nearly all the floating
seed-drift found at sea which has any effective value being derived from
the plants of the beach and of the mangrove belt.

(5) The tropical beach drift of the Old and New Worlds reflects the
characters of the littoral floras of those regions, more especially with
regard to the plants provided with buoyant seeds or seed-vessels. The
plants represented in the beach drift common to both these regions
belong mostly to the Leguminosæ. The large fruits so characteristic of
Old World beach-drift are not found in the New World. The number of
shore plants with buoyant seeds or seed-vessels that are widely
dispersed in the American region are only one-quarter or one-third of
those in the Old World region; and this difference is reflected in the
scanty character of tropical American beach-drift.

CHAPTER XXX

THE VIVIPAROUS MANGROVES OF FIJI

RHIZOPHORA AND BRUGUIERA

_Rhizophora._—Represented by Rhizophora mucronata, Rhizophora
mangle, and the Selala, a seedless intermediate form.—Their
mode of association and characters.—The relation of the
Selala.—Polyembryony.—The history of the plant between the
fertilisation of the ovule and the detachment of the seedling.—Absence
of a rest period.—Mode of detachment of the seedling. Capacity
for dispersal by the currents.—_Bruguiera._—The mode of
dispersal.—Peculiar method of fertilisation.—Length of period between
fertilisation and the detachment of the seedling.—Mode of detachment
of the seedling.—Summary.

BETWEEN 1897 and 1899 I made numerous observations on the Fijian species
of Rhizophora and Bruguiera (mostly around the coasts of Vanua Levu and
in the Rewa delta); and these were supplemented in the early part of
1904 by observations on the first-named genus in Ecuador. I did not make
any collections in Fiji until Prof. Schimper asked me to obtain
specimens; and a fair-sized collection containing specimens dried, and
preserved in spirit, was sent to him. His illness and death shortly
followed, and I lost the advantage of his great experience in these
matters. In a letter written to me in 1898 he expressed the hope that I
would publish my notes on the mangroves of Fiji. Years have since passed
by, and as I read again his words of encouragement I take up once more
the interrupted task.

RHIZOPHORA

Of the three species of this genus, two of them, Rhizophora mucronata
and R. conjugata, are Asiatic and are unknown in America; whilst the
third, R. mangle, was until recently regarded as peculiar to the
American and West African regions.

When Mr. Hemsley wrote the Report on the Botany of the Challenger
Expedition he remarked (iii, 149) that the American Rhizophora (R.
mangle) appeared to be restricted to that region, and he questioned its
existence in the Pacific Islands as indicated by Jouan for New
Caledonia. The same view was taken by Prof. Schimper in his work on the
Indo-Malayan strand-flora published in 1891. There was, in fact, much to
support this view, since Dr. Seemann, one of the most accomplished
botanists who have explored the Pacific, describes only the Asiatic
Rhizophora (R. mucronata) in Fiji, and nothing is said of any other
species collected by the United States Exploring Expedition under Wilkes
in Fiji and Samoa.

However, in a paper on the flora of Tonga, read before the Linnean
Society in 1893, Mr. Hemsley includes the American mangrove, Rhizophora
mangle, amongst the collections made there by Mr. Lister; and he refers
to its occurrence also in Stewart Island (I suppose in the Solomon
Group), but he suggests that it was accidentally introduced with ballast
in both these localities. In 1897 I found a species of Rhizophora, to
all appearances identical with the American species, in great abundance
in the Rewa delta in Fiji. Subsequently the same mangrove came under my
notice as the prevailing species in Vanua Levu in the same group; and on
sending photographs of a branchlet in flower and fruit and of the
germinating fruit to Prof. Schimper he expressed the opinion that they
belonged to the typical Rhizophora mangle.

There are four typical mangroves in Fiji, namely (1) Bruguiera rheedii,
the “Dongo” proper of the natives; (2) Rhizophora mangle, usually known
as “Tiri-wai,” that is to say, the Tiri of the river, or rather of the
estuary; (3) Rhizophora mucronata, the “Tiri-tambua” of the Fijian,
signifying the Whale’s Tooth Tiri in allusion to the form of its fruit;
and (4) a seedless form intermediate between the two species of
Rhizophora, which the Fijians designate “Selala,” a name signifying “the
tree with empty flowers.”

Bruguiera rheedii and Rhizophora mucronata were alone recorded by Dr.
Seemann and his predecessors; but he significantly refers to the natives
speaking of four mangroves. Mr. Horne, who spent twelve months in the
group some years later, also overlooked the American Rhizophora; but it
is apparent that both these botanists were naturally more interested in
the vegetation of the inland regions than of the coast swamps, and we
have before observed that they failed to record Scirpodendron costatum,
a giant-sedge very common and conspicuous in the swamps. It is not easy
to understand Dr. Seemann’s remark that “mangroves are restricted to but
few parts of the larger islands.” Horne, who was in the islands eighteen
years afterwards, makes frequent allusion to them. The natives whom I
questioned closely on this subject scouted the idea that any of the four
mangroves above named were recent arrivals. The coasts, as they said,
had always been extensively fringed by mangroves; and the reader has
only to refer to my remarks in the second chapter of my volume on the
geology of Vanua Levu to convince himself that mangrove swamps of
considerable extent existed in the time of Commodore Wilkes (1840).

_The Relative Abundance and Mode of Association of the three Fijian
forms of Rhizophora._

Stated in their order of frequency, we have first Rhizophora mangle, the
American species, then Rhizophora mucronata, the Asiatic species, and
lastly the Selala. The first is equally at home at the sea-border and on
the banks of brackish estuaries. The second is, as a rule, more
exclusively at home on the sea-coasts; and the same may be said for the
Selala. Usually all three kinds occur in the lower part of an estuary;
but as we ascend the river and the water freshens, the Asiatic
Rhizophora and the Selala disappear, and the American plant is alone
found in the higher reaches, where the density of the water ranges
according to the state of the tide between 1·000 and 1·010. I examined
the distribution of these three forms of Rhizophora in numerous
estuaries of Vanua Levu, as well as in the Rewa estuary in Viti Levu;
and it was ascertained that in all cases they followed the rule above
indicated. When the estuary receives but few streams and the water is
mostly salt, the three Rhizophoras may extend miles inland; but when it
contains a large body of fresh-water, Rhizophora mangle may be the only
form observed from the mouth of the river to the head of the estuary,
and it may monopolise the adjacent coasts. On the other hand, Rhizophora
mucronata may occupy almost exclusively a long extent of coast; or the
Selala may prevail in certain localities, as on parts of the Mathuata
coast of Vanua Levu.

The manner of association of these three Rhizophoras is of interest in
connection with the origin of the seedless Selala. They very rarely
occur mingled together, but grow gregariously in contiguous colonies;
and not uncommonly all three may occur on the same line of coast within
a distance of a few hundred yards. The colonies pass into each other
without a break, and there is no fixed rule of association. Whilst on
the south side of Vanua Levu the Selala is generally associated with the
American Rhizophora, on the north side it is usually in touch with the
Asiatic species. In other localities all three occur in contiguous
colonies. The Selala colony may be exposed on the line of a river-bank
or along the sea-coast, or it may lie in the heart of an extensive
mangrove tract. The most extensive mangrove region in Fiji, that of the
Rewa delta, is in great part occupied by Rhizophora mangle; but all
three forms grow together in the eastern part of the delta; and here,
strangely enough, as at Daku, the Selala may grow sporadically, and all
three may grow mixed together with their branches intercrossing.

_The Characters of the Selala or Seedless Rhizophora compared with those
of the American Mangrove (R. mangle) and the Asiatic Mangrove (R.
mucronata)._

The three kinds of Rhizophora, when seen at the same time along a tract
of coast, may be readily distinguished by the different shades of green
of their foliage, that of Selala being dark green, that of Rhizophora
mucronata light green, and that of Rhizophora mangle intermediate in
shade. The Selala is usually the tallest of the three, and attains a
height of from 20 to 30 feet or even 40 feet and over, the aerial roots
dropped from the higher branches giving it a characteristic aspect.
Rhizophora mangle is generally the shortest, and at the coast is from 10
to 12 feet high; but where the mangrove vegetation is most luxuriant, as
in the great swamps in the interior of the Rewa delta, it forms tall
trees as much as 40 feet in height, displaying the aerial roots hanging
from the higher branches. Rhizophora mucronata is, as a rule,
intermediate in height, and is distinguished by its stout, reddish trunk
and reddish aërial roots.

The trunks of Selala are often in an inclined position and supported
entirely by the trestle-like aërial roots, the lower end raised some 5
or 6 feet above the ground with the rest of the trunk inclined upwards.
They then look like gigantic walking-stick insects. The same habit may
be sometimes observed with the larger trees of Rhizophora mucronata, and
in fact all three may present at times the same habit of growth. The
taller trees of Rhizophora mangle may resemble the Selala in habit, and
the smaller trees of the Selala may approach the habit of Rhizophora
mangle.

The distinctive characters of the Selala are given in the table
opposite. It will be there seen that this form is intermediate between
the other two species as regards the form and size of the petioles and
peduncles; the size of the bracts and bracteoles; the colour, form, and
size of the flowers; and in the length of the style. Its leaves are
smaller than in the case of the other two species, but pointed and
semi-aristate like those of Rhizophora mucronata. There are, however,
two varieties of the Selala, both with larger foliage than that
belonging to the prevailing type of the tree, and from 10 to 15 feet in
height. In one the flowers are more numerous, each flowering stem
branching four or five times and bearing at least twenty-four flowers,
the first branch being trichotomous and the rest dichotomous. In the
other, which is the prevailing form on the Mathuata coast, there is a
nearer approach to Rhizophora mucronata in the rounding of the peduncles
and in the length of the style. Then, again, there are divergent
varieties of Rhizophora mangle which in the larger bracts and bracteoles
and in the greater size, form, and paler hue of the flowers come nearer
to the Selala. Taking all the characters together, the Selala, though
intermediate between the Asiatic and the American species, comes in the
most critical diagnostic points, as in the inflorescence, in the
individual flowers, and in the form of the apex of the leaf, nearest to
Rhizophora mucronata, the Asiatic species.

The seedless character of the Selala is well known to most Fijians of
the coast districts, the native name signifying empty (lala) flowers
(se). Now and then they aver that it produces fruit, but the numerous
offers of rewards in money never resulted in their bringing me the
fruits. During my residence of two years in the group I examined the
Selala trees in a great number of localities and never succeeded in
finding them in fruit.

With all three kinds the anthers burst in the bud before it begins to
open, and we may ask why the process of self-fertilisation, which is
effectual with the other two kinds, produces no result with the Selala.
In all three cases the flower-buds and expanded flowers hang downwards,
and the expanded flowers retain their parts for the first twenty-four
hours, the pollen being caught in quantity on the hairy edges of the
petals. During the next day the withering stamens fall out, and on the
following day the petals fall too. With the Selala, the style soon
begins to blacken and wither, and in a few days the flower becomes
detached and drops off. With Rhizophora mucronata and Rhizophora mangle,
the style preserves its healthy condition, and shortly evidences of
fertilisation appear in the altered shape of the ovary. It is apparent,
therefore, that in the case of the Selala fertilisation has not
occurred, although the mechanical processes connected with it have been
carried out. The cause of this is not far to seek.

+-----------------+--------------------------------------------------------+
| | FIJI |
| CHARACTERS +------------------+------------------+------------------+
| | RHIZOPHORA | SELALA (a | RHIZOPHORA |
| | MUCRONATA. | seedless form). | MANGLE. |
+-----------------+------------------+------------------+------------------+
|Height of tree | | | |
| in feet | 12-20. | 20-40. | 9-12. |
| | | | |
|Colour of foliage|Pale green. |Dark green. |Intermediate |
| | | |shade. |
| | | | |
|Base of leaf |Tapering. |Sub-rounded. |Tapering. |
| | | | |
|Apex of leaf |Acute, and |Acute, and |Very obtuse, |
| |terminating |terminating in |with no |
| |in a twisted |a twisted point |twisted point. |
| |point a line |less than a line | |
| |(2·5 mm.) long. |(2·5 mm.) long. | |
| | | | |
|Leaf-stalk |Rounded, 1 - |Rather flattened |Length as in |
| (petiole) |1-2/10 inch, |horizontally, |Selala, but |
| |(25-30 mm.) long, |5-8/10th inch |flattening |
| |about as long as |(12-20mm.) long, |very marked. |
| |the peduncle. |shorter than the | |
| | |peduncle. | |
| | | | |
|Inflorescence |Branching |Branching usually |Usually branching |
| |(dichotomous) two |twice, but |only once |
| |or three times |sometimes three |(trichotomous) and|
| |with four to eight|times; first |bearing only three|
| |flowers. |branching |flowers; but |
| | |trichotomous, rest|sometimes |
| | |dichotomous; six |branching again |
| | |to twelve flowers;|(dichotomous) and |
| | |in one variety, |bearing then six |
| | |flowers as many as|flowers. |
| | |twenty-four. | |
| | | | |
|Peduncle (lowest}|Rounded. |Flattened above. |{Flattening more |
| flower-stalk) }| | |{marked than with |
| | | |{Selala. |
| | | | |
|Pedicels |As stout as the |More slender than |As in Selala. |
| |peduncle, and |the peduncle, and | |
| |rounded. |angular. | |
| | | | |
|Bracts and |Large, 1-1/2 line |Small, 2/3 line |Very small or |
| bracteoles |(4 mm.). |(2 mm.). |absent. |
| | | | |
|Calyx |Very pale yellow, |As in R. |Pale or bright |
| |or dirty white, |mucronata. |green, angular at |
| |rounded at base | |base in the bud, |
| |in the bud, lobes | |lobes 3-1/2 lines |
| |4-1/2 - 5 lines | |(8 mm.). |
| |(11-12 mm.). | | |
| | | | |
|Length of style |1-1/2 lines |1 line (2·5 mm.). |1/2 line |
| |(4 mm.). | |(1·5 mm.). |
| | | | |
|Fruit |Ovoid and usually |No fruits |Conical, somewhat |
| |symmetrical, with |produced. |curved, and thus |
| |large persistent | |not symmetrical; |
| |bracteoles at | |bracteoles at base|
| |base. | |very small or |
| |(Hypocotyl 16 | |absent. |
| |inches.) | |(Hypocotyl 9 or |
| | | |10 inches.) |
| | | | |
|Colour of trunk, | | | |
| rootstock and | | | |
| roots |Reddish. | — | — |
+-----------------+------------------+------------------+------------------+

+-----------------+-------------------------------------+
| | ECUADOR. |
| CHARACTERS. +------------------+------------------+
| |RHIZOPHORA MANGLE |RHIZOPHORA MANGLE |
| | (Mangle chico). | (Mangle grande). |
+-----------------+------------------+------------------+
|Height of tree in| 10-15. | 50-80 and more. |
| feet | | |
| | | |
|Colour of foliage|Pale green. |Dark green. |
| | | |
|Base of leaf |Tapering or |Tapering. |
| |sub-rounded. | |
| | | |
|Apex of leaf |Very obtuse, with |Very obtuse, with |
| |no twisted point. |no twisted point. |
| | | |
|Leaf-stalk |Flattened above |Flat above with a |
| (petiole) |and below, with no|median groove, |
| |median groove, |1 inch (25 mm.) |
| |1/2 inch (12 mm.) |long, two-thirds |
| |long; not half as |the length of the |
| |long as the |peduncle. |
| |peduncle. | |
| | | |
|Inflorescence |As described under|Branching at least|
| |R. mangle of Fiji.|three times, |
| | |sometimes four or |
| | |five times, |
| | |trichotomous or |
| | |dichotomous, |
| | |twelve to |
| | |forty-eight |
| | |flowers. |
| | | |
|Peduncle (lowest}| | |
| flower-stalk) }|}Sub-angular. |Rounded. |
| | | |
|Pedicels |More slender than |More slender than |
| |the peduncle, and |the peduncle, and |
| |rounded. |angular. |
| | | |
|Bracts and |Scarcely |Well developed, |
| bracteoles |developed, 1/2 |1 line (2·5 mm.). |
| |line (1 mm.). | |
| | | |
|Calyx |As with R. mangle |As with R. |
| |of Fiji. |mucronata and |
| | |Selala of Fiji, |
| | |but lobes 4 lines |
| | |(10 mm.). |
| | | |
|Length of style |Less than a line |1-1/2 lines |
| |(2·5 mm.). |(4 mm.). |
| | | |
|Fruit |As in R. mangle of|Conical, not |
| |Fiji. |symmetrical, and |
| |(Hypocotyl 9 or |somewhat curved; |
| |10 inches.) |large persistent |
| | |bracteoles at base|
| | |as in R. |
| | |mucronata. |
| | |(Hypocotyl 12 to |
| | |15 inches.) |
| | | |
|Colour of trunk, | | |
| rootstock and | | |
| roots | — | — |
+-----------------+------------------+------------------+

Although the ovaries of the Selala contain four ovules, which in size
and appearance do not differ from those of Rhizophora mangle and R.
mucronata, its pollen when compared with that of the other two forms
presents a remarkable difference. The pollen of these three mangroves
was examined in five localities far apart from each other, and in all
the same results were obtained. The pollen-grains of the Selala are much
smaller than those of the other two, and differ much from them in form.
They are irregularly oval in outline, and have a shrunken look beside
the spherical symmetrical grains of the two species with which they are
compared. They are from one-fourth to one-third the size of those of
Rhizophora mucronata, and from one-third to one-half the size of those
of Rhizophora mangle.

There is much to support the view that the Selala is a cross between the
other two species, its intermediate characters and its seedless
condition being especially indicative of such a derivation; but there
are several difficulties in accepting this explanation.

(1) The circumstance of the anthers bursting in the flower-bud would
considerably lessen the chances of cross-fertilisation; but this
objection is not insurmountable, since numerous insects, such as flies,
ants, and small coleoptera, visit the newly opened flowers, and they
might sometimes produce a result. When I made this suggestion to Prof.
Schimper he replied that insect-pollination was quite possible after the
expansion of the flowers.

(2) If, as seems highly probable, the pollen of Selala is impotent
and the ovules fertilisable, then its seedless condition implies not
only an incapacity for self-fertilisation, but also for
cross-fertilisation; and if Selala with its impotent pollen does not
admit of cross-fertilisation, this would still less be expected of
Rhizophora mucronata and R. mangle where the pollen is potent and
where fertilisation takes place in the bud. I endeavoured to
fertilise the Selala flowers with the pollen of the two other
species; but there were no results, the flowers falling off in a few
days. It may here be remarked that on one Selala tree I found a
solitary flower with an enlarged ovary, as if through fertilisation.

(3) It is not easy to explain the gregarious growth of the Selala if it
is a seedless hybrid. The colonies could not be renovated by mere
intercrossing, especially in places where, as on the north coast of
Vanua Levu, the dense belt of mangrove is for many miles composed in
mass of Selala trees, with a few trees of the Asiatic and American
Rhizophoras growing on the outskirts.

It is obvious that in order to clear the way for considering this
problem the means of renovating the Selala colonies should be inquired
into. In the first place, whilst seedlings occur in numbers under the
trees of the other two Rhizophoras they are never to be found under the
Selala trees. The mode of reproduction of the Selala is evidently
vegetative, and the question arises as to what mode of vegetative
reproduction occurs. The Selala trunks, as already observed, are often
inclined, the trunks being supported on trestle-like aërial roots. These
trunks send out branches which in their turn drop aërial roots; and when
the decay of the parent trunk takes place, the branches are able to live
independently. The primary branches in due time send out secondary
branches which also let fall aërial roots; and thus the process is
repeated indefinitely, the result being a maze of semi-prone trunks,
branches, and aërial roots. The first stage of the process ends with the
death of the parent trunk, and the primary branch, supported by its own
aërial roots, is often all that the observer can distinguish in the
centre of a colony. This is evidently the mode by which the Selala
colonies are renovated in their interior. One sometimes observes in the
midst of one of these colonies extensive bare mud-flats 100 to 500 yards
across from which apparently the trees have died off _en masse_. The
natives assert that when part of a Selala tract is cleared the trees
never grow again.

But _pari passu_ with this process of vegetative reproduction of the
Selala, by which the mass of the colony is preserved and renovated,
there is evidently some other process of reproduction in operation
amongst the trees of Rhizophora mangle and R. mucronata at the edge of
the colony, as a result of which Selala seedlings are produced. Whilst
no seedlings are to be observed striking into the mud under the Selala
trees, numbers occur, as before observed, under the trees of the other
two species. Those under the trees of R. mangle possess in nearly all
the cases the distinctive leaf-characters of that mangrove, and would be
recognised at once as belonging to that species. On the other hand,
those beneath the trees of R. mucronata are of two kinds, some of them
being readily recognised by their foliage as of the Selala type others,
again, being typical seedlings of R. mucronata. Only those seedlings, or
“keimlings” as we might term them, were noted that had dropped plumb
from the branches above.

Such were the results of my investigations on Vanua Levu. My field of
inquiry was then shifted to the Rewa delta, where, with the assistance
of the Daku natives, who, like most Fijians, display a keen interest in
matters relating to their plants, I spent a few days in investigating
the origin of the Selala trees that grow sporadically in that locality.
On pulling up some of the young trees we found that the original
radicular or hypocotyledonary portion of the keimling could be still
distinguished. My zealous native friends also pointed out to me that
though the leaves in form and colour were those of the Selala, the
rootstock was reddish like that of R. mucronata, and not white as with
R. mangle. The natives averred that the Selala trees are produced in the
first place from fruits of R. mucronata. When young, they said, they are
Tiri-tambuas (R. mucronata), but when old, Selalas. Yet although R.
mucronata may be now regarded as the source of the Selala trees, and my
Vanua Levu observations pointed unmistakably in this direction, it could
not be definitely settled whether this was the result of a cross with
the male element of R. mangle or whether the Tiri-tambua (R. mucronata),
in producing two types of seedlings, one fertile with the parent
characters and the other seedless of the Selala type, brought about the
same end. On the whole I am inclined to the view that the Asiatic
Rhizophora presents us in the dimorphism of its seedlings the true
explanation.

This inference is supported by the behaviour of Rhizophora mangle on the
coast of Ecuador, a subject which is discussed in Chapter XXXII, and I
have given the results of my observations on the Ecuadorian Rhizophoras
side by side with those on the Fijian trees in the table before given.
There are two very distinct forms of the American Rhizophora (R. mangle)
in the swamps of Ecuador. There is the low coast tree, the “Mangle
chico” of the Ecuadorians, ten to fifteen feet in average height, which
grows on the sea-front of the swamps and has all the general appearance
and the more conspicuous characters of the American Rhizophora in Fiji.
There is also a tall tree, 60, 80, or even 100 feet high, that forms the
great mass of the mangrove swamps. In its inflorescence, in the dark
green colour of its foliage, and in other characters, it comes near the
Fijian Selala; but it differs in fruiting abundantly. This is locally
termed the “Mangle grande,” and its true relation to the Fijian Selala
appears to be as follows. Whilst both as regards the flowers approach
the Asiatic Rhizophora (R. mucronata), the Fijian Selala resembles the
Asiatic tree also in its foliage, whilst the “Mangle grande” or the
Ecuadorian Selala more resembles the typical American tree (R. mangle)
in its leaves and also in its seedlings. Here in the Ecuadorian swamps
there can be no question of crossing, since both, according to Baron von
Eggers, belong to one species. Therefore I am inclined to the opinion
that whilst the Asiatic Rhizophora displays dimorphism in Fiji, the
American Rhizophora displays dimorphism in Ecuador. The reversion on the
part of the “Mangle grande” of Ecuador to some of the characters of the
Asiatic plant is remarkable, and points to the greater antiquity of the
Asiatic R. mucronata as compared with the American R. mangle.

This accords with the opinion expressed by Schimper in his work on the
Indo-Malayan strand flora that the American Rhizophora is either a
degenerated descendant of the Asiatic R. mucronata or a sister form
derived from a common ancestor. America, as we have seen, possesses only
one of the three species of Rhizophora, and this is the only
representative that it owns of the four Asiatic genera (Rhizophora,
Kandelia, Ceriops, Bruguiera) that constitute the tribe Rhizophoreæ. The
rule prevailing with current-dispersed plants that America is a
distributor and not a recipient evidently does not apply to the
Rhizophoreæ; and to explain their distribution we must go back to some
epoch very remote from the present. That Fiji derived its
representatives of Rhizophora mangle from America by the agency of the
currents I do not for a moment admit. The restriction of the species and
indeed also of the genus to the Western Pacific is very significant. It
is far more likely that, as I have pointed out in the case of Lindenia
(see page 396), the American Rhizophora was once widely distributed over
the tropics of the Old and New Worlds, and that it is now on the “down
grade” towards extinction. Its survival in the Western Pacific could
thus be explained without our being obliged to suppose that the
seedlings or keimlings have been carried uninjured across the Pacific
Ocean, an ocean voyage for which, as shown in a later page, they are not
well fitted.

_The Occasional Occurrence of more than one Seed in the Fruits of
Rhizophora mucronata and Rhizophora mangle (Polyembryony)._

The bilocular ovary contains four ovules, one of which only as a rule
becomes a seed. But it is incorrect to say that the fruits are always
one-seeded, since two or even three seeds are occasionally produced, and
they may all germinate. In November, 1897, I noted eight hundred fruits
of Rhizophora mangle germinating on the trees in one of the creeks of
the Rewa delta. Out of this number eight fruits had two germinating
seeds and one had three, the protruding radicles being in all stages of
growth. Just two years afterwards I counted eight hundred more fruits in
the same locality, and then observed seven with two germinating seeds
and none with three, the radicles protruding in all cases. On another
occasion at Wailevu in Savu-Savu Bay I counted four hundred, and none
had more than a single radicle protruding. The results appear to vary
with the locality, but in the Rewa creek the proportion of fruits in
which more than one seed germinated was fairly constant at dates two
years apart, namely, about one per cent. Occasionally, however, in
particular localities a greater proportion may be noticed. Thus near
Daku in the Rewa delta I found that the proportion was between two and
three per cent. for the same species (R. mangle), those with three
germinating seeds being about half per cent.

The case of more than one seed germinating in the fruits of Rhizophora
mucronata never came under my observation; but in one locality, where I
examined a considerable number of fruits near the stage of germination,
between ten and fifteen per cent. showed two seeds approaching maturity.

Warming thoroughly investigated the polyembryony of Rhizophora more than
twenty years ago, seemingly from materials brought to him from the West
Indies (Engler’s _Botanische Jahrbücher_, band iv., 1883). With the
usual German thoroughness he deals with the work of earlier observers,
and goes back to Piso in the middle of the 17th century. Of the four
ovules, he remarks, three usually abort, and only in rare cases are two
seeds developed. He quotes Baron von Eggers to the effect that only in
three per thousand cases was more than one seedling observed protruding
from a germinating fruit. These remarks evidently all apply to the
American species. I do not find any reference in my notes to
polyembryony in Ecuador, and evidently its occurrence is not so frequent
there as in Fiji.

It is frequently apparent in the cases where more than one seed
germinates in a fruit that on account of the difference in the length of
the protruding seedlings germination does not always begin at the same
time. Thus in Fiji the difference in the length varied between one and
three inches, an amount representing at least from ten to twenty days’
growth, as will be subsequently pointed out. Warming in one of his
figures gives a fruit where an interval of some months seems to be
indicated, since one of the seedlings has fallen out and the other is
protruding less than an inch. By cutting across a fruit containing two
seeds one may sometimes observe one seed quiescent and the other
beginning to germinate. The significance of this occasional interval
between the germination of seeds in the same fruit will be referred to
in a later page.

_The Seasons of Flowering and Fruiting of the Species of Rhizophora in
Fiji._

The Selala flowers all the year. With the two American and Asiatic
species there are considerable variations between different localities.
Generally speaking, they flower and fruit all the year through; but the
flowers are usually less abundant in the warm season from December to
February, and the germinating fruits which are to be observed on the
trees every month of the year are more numerous in that season.

_The History of the Reproductive Process in Rhizophora from the
Fertilisation of the Ovule to the Falling of the Plantlet or Seedling
from the Tree._

I devoted great attention to this subject in the instance of Rhizophora
mangle, being desirous of determining two points, in the first place as
to whether there was any period of rest between the maturation and
germination of the seed, and in the second place as to the period that
elapsed between the commencement of germination and the fall of the
seedling.

The principal change in the ovary for the first three or four weeks
after fertilisation is shown in its increased breadth. The increase in
height is but slight during this period; and in fact after thirty days
the ovary only added 2 millimetres to its original height of 3
millimetres. After this the growth of the fruit proceeds until the tip
of the radicle pierces its summit, the fruit being then about eleven
lines (2·8 cm.) long. _From the date of fertilisation to the time the
radicle pierces the top of the fruit a period of about fifteen weeks
elapses._ (The fruit, it should be here remarked, continues to grow in
length and breadth after the radicle has protruded, attaining a length
of thirteen or fourteen lines [3·5 cm.] when the seedling or “keimling”
is ready to fall.)

By referring to the table below it will be observed that there is no
period of rest in the growth of the fruit up to the date of the
protrusion of the radicle. It will now be shown that there is normally
no pause between the epoch of the maturation of the seed and the
beginning of germination, or, in other words, that from the time of the
fertilisation of the ovule to the onset of germination there is no
cessation in the process of growth of the embryo. That period of dormant
vitality which almost all seeds pass through forms no normal feature in
the life-history of this species of Rhizophora.

RHIZOPHORA MANGLE AND R. MUCRONATA.

+-------------------------------------------+
| RHIZOPHORA MANGLE. |
+-------------------------------------------+
| _Growth of fruit in height._ |
+---------------------+---------------------+
| Lines or tenths of | Number of days |
|an inch (millimetres | since |
| in brackets). | fertilisation. |
+---------------------+---------------------+
| | |
| 2 (5) | 30 |
| 3 (7·5) | 42 |
| 4 (10) | 50 |
| 5 (12·5) | 61 |
| 6 (15) | 67 |
| 7 (17·5) | 74 |
| 8 (20) | 83 |
| 9 (22·5) | 92 |
| 10 (25) | 100 |
| 11 (28) | 105 |
| | { Protrusion|
| | 107{ of the |
| | { hypocotyl |
| |
| _Growth of the protruding hypocotyl._ |
| |
| 10 (25) | 127 |
| 20 (50·5) | 141 |
| 30 (76) | 151 |
| 40 (101·5) | 160 |
| 50 (127) | 167 |
| 60 (152) | 175 |
| 70 (177·5) | 185 |
| 80 (203) | 202 |
| 90 (228) | 222 |
| | {Fall of |
| | 229{the |
| | {seedling |
+---------------------+---------------------+

_Explanation of the Table._

We have here shown the period between fertilisation and the fall of the
seedling from the tree.

This period divides itself into two parts, the first being concerned
with the continuous growth of the fruit and of the inclosed embryo until
the tip of the hypocotyl appears through the apex of the fruit, the
second being indicated by the growth of the protruding hypocotyl until
the fall of the seedling.

The height of the fruit is measured from the base of the calyx-lobes,
and the length of the hypocotyl at first from the apex of the fruit and
afterwards from the edge of the protruding neck of the cotyledonary
body. The height of the ovary at the time of fertilisation is about 3
millimetres; and from that time onward it is to be regarded as a fruit.

[_To face page 452._

FIGURES ILLUSTRATING THE DEVELOPMENT OF
THE SEED AND THE GERMINATING PROCESS
OF RHIZOPHORA AND BRUGUIERA

(Natural size. Drawn for convenience of description in the erect
position.)

1. Rhizophora mucronata Fruit 3-1/3 lines (8-9 mm.) high, six to
seven weeks after fertilisation. The
micropyle is but slightly dilated, and is
occupied by a small plug of endosperm.

2. Rhizophora mucronata Seed of fruit represented in Fig. 1.

3. Rhizophora mucronata Fruit 5 lines (12 mm.) high, eight to nine
weeks after fertilisation. Germination is
about to begin. A large plug of endosperm now
protrudes through the dilated micropyle, but
still covers the lengthening hypocotyl.

{Seed of fruit represented in Fig. 3. In
4.} Rhizophora mucronata {Fig. 4 the plug of endosperm is shown on the
5.} {upper end of the seed; whilst in Fig. 5 it
{has been removed, exposing the tip of the
{hypocotyl.

6. Rhizophora mucronata Fruit 7-1/2 lines (18 mm.) high, eleven to
twelve weeks after fertilisation.

7. Rhizophora mucronata Fruit, seventeen or eighteen weeks after
fertilisation.

8. Rhizophora mucronata Full-grown fruit with upper portion of
seedling just before detachment from the
tree. The long tapering plumule is here
inclosed in the cotyledonary body. The *
indicates the point of detachment of the
seedling.

9. Rhizophora mucronata The cotyledonary body of Fig. 8.

9A.} {Illustrating different stages in the
9B.}Rhizophora mucronata {development of the plumule and of the neck of
9C.} {the cotyledonary body resulting finally in
9D.} {the expulsion of the plumular end of the
{seedling from the fruit cavity as in Fig. 8.
{(See page 458.)

10. Rhizophora mucronata Fruit with two seeds.

FIGURES ILLUSTRATING THE DEVELOPMENT OF
THE SEED AND THE GERMINATING PROCESS
OF RHIZOPHORA AND BRUGUIERA—(_continued_)

(Natural size. Drawn for convenience of description in the erect
position.)

11. Rhizophora mangle Fruit, six weeks after fertilisation.

12. Rhizophora mangle Seed with plug of endosperm, as shown in Fig. 11.

13. Rhizophora mangle Fruit, eight weeks after fertilisation. The tip of
the hypocotyl is now piercing the plug.

14. Rhizophora mangle Embryo (enlarged) shown in Fig. 13.

15. Rhizophora mangle Fruit, ten weeks after fertilisation. The growing
hypocotyl has now pierced the plug.

16. Rhizophora mangle Embryo shown in Fig. 15.

17. Rhizophora mangle Fruit, nearly sixteen weeks after fertilisation.

18. Rhizophora mangle Full-grown fruit, just before the detachment of the
seedling from the tree. The long tapering plumule
is inclosed in the tube of the cotyledonary body.
The point of detachment of the seedling is
indicated by *.

19. Rhizophora mangle The cotyledonary body of Fig. 18.

20. Rhizophora mangle Fruit with two seedlings in different stages of
growth (given in the first plate).

21. Bruguiera Rheedii Fruit, about four weeks after fertilisation. (The
shaded portion is the calyx-tube or cup, in the
midst of which rises the style.)

22. Bruguiera Rheedii Germinating seed.

23. Bruguiera Rheedii Germinating fruit, about eight weeks after
fertilisation.

24. Bruguiera Rheedii Germinating fruit, about ten weeks after
fertilisation. Here the growing hypocotyl,
carrying the style with it, has pushed upwards the
lining membrane of the floor of the calyx-tube,
which has ruptured and forms a cap on its extremity.

25. Bruguiera Rheedii Germinating fruit, thirteen or fourteen weeks after
fertilisation.

26. Bruguiera Rheedii Fruit with full-grown seedling just before its
detachment from the tree.

[_To face page_ 453.

RHIZOPHORA MUCRONATA.

_Growth of the first seven inches of the hypocotyl after it protrudes
from the fruit._

10 lines (25 mm.) after 26 days
20 lines (50·5 mm.) after 41 days
30 lines (76 mm.) after 51 days
40 lines (101·5 mm.) after 61 days
50 lines (127 mm.) after 70 days
60 lines (152 mm.) after 78 days
70 lines (177·5 mm.) after 86 days

In my description of the germinating process of Rhizophora mangle from
this particular standpoint I adopt the general views of Prof. Schimper,
the observations being my own, the phraseology employed being his. It
would be out of place here to deal with the biological significance of a
process to which observers like Warming, Goebel, Karsten, Schimper and
Haberlandt have applied their greater talents as well as their greater
experience. I investigated the subject carefully from my own standpoint
of inquiry, and whilst the reader will find in my rough sketches of the
various stages of the process a little aid in following the argument, he
is referred for detailed treatment of the subject to the memoirs of the
above-named botanists as well as to those of yet more recent
investigators.

After fertilisation, according to Prof. Schimper (_Ind. Mal.
Strandflora_), the embryo-sac is filled with endosperm, which
subsequently protrudes and forms a plug completely closing the micropyle
(see my figures). As my observations showed, the seed during the first
eight weeks after fertilisation increases continuously in size, and the
plug of endosperm, at first inconspicuous, becomes of considerable size,
the seed attaining a length of seven millimetres. The embryo meanwhile
grows rapidly, and at the end of this period of eight weeks the
radicular tip or the point of the hypocotyl begins to protrude from the
micopyle, still covered by the plug of endosperm, the fruit being
between four and five lines (10-12 mm.) in length (figures 11-14). In
another week, when the fruit has grown another line in length, the tip
of the radicle is on the eve of piercing the plug, and this may be
termed _the commencement of germination, nine weeks after the act of
fertilisation_. The next stage, after an interval of one and a half
weeks, is illustrated in figure 15; and _after a period of about fifteen
weeks from the date of fertilisation the tip of the radicle pierces the
top of the fruit_. As shown in the figures, the fruit grows in length
throughout the process.

The question as to whether the matured seed passes through a stage of
quiescence before it germinates finds its answer in the statement that
only nine weeks elapse between fertilisation and germination. It may,
however, be urged that the maturation of the seed could be accomplished
in a few weeks, and that after this a period of dormant vitality might
follow. This objection can be at once disposed of and the whole matter
placed beyond reasonable doubt by making, as I did, a large number of
vertical sections of the fruit in all its stages. It will then be
perceived that there is a fairly constant relation in all stages of
growth between the seed and the fruit, whether maturating or
germinating. Since the growth of the fruit is continuous (see Table) up
to the time of the protrusion of the tip of the hypocotyl through its
coats, it follows that there can be no appreciable pause between the
completion of maturation and the commencement of germination of the
seed. In other words, both fruit and seed preserve the same relation
during the process, and the absence of any period of rest is to be
inferred from the uninterrupted growth of the fruit.

We will take, to illustrate this point, a fruit between four and five
lines long in the stage that immediately precedes germination (see
figure 11). The fruit proceeds with its growth, and the seed, we will
suppose, remains quiescent for a month. At the end of that time (see
Table) the fruit would be eight lines long, and the seed, of course,
would be unchanged. This condition of things never presented itself to
me. Fruits eight lines long were always far advanced in germination (see
figure 15). If the seed passes through an interval of rest before
germination, it must be of a very short duration and practically _nil_.

This absence of any period of rest between the final maturation of the
seed and its fertilisation had already been assumed by Prof. Schimper.
Writing to me on July 14, 1898, when my observations were in progress,
he says:—“I am ready to assume, according to my own experience, that
there is continuous development until the falling off of the embryo.
More accurate observations on the subject would be interesting, and
would not present any great difficulties.” At the end of the same month
he wrote the preface to his great work on Plant-Geography; and he
expresses himself decidedly on this point. Speaking of Rhizophora
mucronata (English edition, p. 396), he says that “the fruit ... soon
after the completion of its growth is pierced at its summit by the green
hypocotyl, as the embryo does not undergo any period of rest, but
continues to develop without interruption.”

Though the rest-period is normally non-existent with the seeds of
Rhizophora, it has already been observed that it is indicated in rare
cases and under exceptional conditions. Thus I have already remarked
that in Fiji about one per cent. of the germinating fruits of the
American species exhibit more than one seed. These seeds usually begin
to germinate about the same time, but in a few cases, say, one in ten, a
marked difference in the length of the protruding hypocotyls points to
the fact that one of the seeds began to germinate some weeks after the
other. We at times also meet with fruits which when cut across display
two seeds, of which only one is beginning to germinate. Such cases
indicative of a pause between the maturation of the seed and the
beginning of germination would be very rare. With Rhizophora mangle,
probably one in a thousand fruits would be a generous estimate.

In passing it may be remarked that the same stages occur with Rhizophora
mucronata in the development of the seed and in the subsequent
germinating process. When the fruit is three lines long the micropyle is
but slightly dilated (see figures 1 and 2). When it is four lines long
the endosperm begins to escape from the gaping micropyle and forms a
projecting plug. The growth of the embryo now becomes rapid, the
endosperm escapes in greater quantity, and by the time the fruit is five
lines long the tip of the radicle is on a level with the micropyle,
although still covered by the plug (see figures 4, 5). After this,
germination begins; and when the fruit is six lines in length the
radicle is in the act of penetrating the plug. Ultimately the tip of the
radicle pierces the top of the fruit when this last is nine or ten lines
long. As shown in the figures there is continuous growth of the fruit
during the maturation and germination of the seed, until, in fact, the
plantlet drops into the water. With reference to the stage when
germination begins, it should be remarked that the formation of the
large plug of endosperm outside the micropyle does not necessarily
indicate the beginning of germination. Germination is in progress only
when the hypocotyl or radicle begins to lengthen and is on the point of
piercing the plug of endosperm that fills up the gaping micropyle. This
is well shown in this species in the case of fruits with two seeds. Both
seeds may have large plugs of endosperm, and yet only one may show
indications of germination in the lengthening hypocotyl.

We must now return to the subject of the growth of the hanging seedling
of Rhizophora mangle. We have already remarked that, as shown in the
Table, about fifteen weeks (107 days) is the average time elapsing
between the fertilisation of the ovule and the protrusion of the tip of
the radicle through the top of the fruit. A further period of seventeen
and a half weeks (122 days) is occupied by the growth of the seedling on
the tree, at the end of which period it drops into the water or mud
according to the state of the tide. This gives a total period of nearly
thirty-three weeks (229 days) as the duration of the time between
fertilisation and the fall of the seedling. This may be divided, as has
been already implied, in the following manner:—

(1) Period between fertilisation and germination. 9 weeks.

(2) Period between the commencement of germination 6-1/2
and the protrusion of the tip of the radicle weeks.
through the top of the fruit.

(3) Period occupied by the growth of the hypocotyl 17-1/2
outside the fruit, and terminating in the fall weeks.
of the seedling from the tree.

——

Total 33 weeks.

This represents the average of numerous observations, the deviations
being from two to three weeks on either side. In the latter part of its
growth, the lower end of the hypocotyl becomes thickened or club-like,
and during the last week or ten days the increase in length is arrested
altogether.

My observations on the growth of the seedling on the tree of Rhizophora
mucronata were comparatively few; but, as shown in the Table on page 453
they give nearly the same rate of growth. Taking the average length
attained by the hypocotyl on the tree at sixteen inches, and employing
as well the data supplied by Rhizophora mangle, a period of 26-1/2 weeks
would elapse from the time the hypocotyl pierces the top of the fruit
until the plantlet falls from the tree. If we then add, as in the case
of the other species, 15-1/2 weeks for the preceding period between
fertilisation and the protrusion of the hypocotyl, we get a total of 42
weeks for the whole period from fertilisation to the fall of the
seedling. In the extreme cases where a length of almost two feet is
attained on the tree, the period would somewhat exceed twelve months;
and in those rare instances in other regions, when, according to
Schimper, the seedling is a metre in length, probably eighteen months
would be required. The period for Rhizophora mucronata is thus
considerably longer than for R. mangle, which is sufficiently indicated
by the difference in the average length of their hypocotyls on the tree
in Fiji, that for R. mucronata being sixteen inches, and that for R.
mangle nine or ten inches.

The only other observations that have come under my notice relating to
this subject are those made by Jacquin on Rhizophora mangle in the West
Indies in the middle of the eighteenth century. The results are
literally quoted by Warming; but I have referred to the original account
in the work of Jacquin, entitled _Selectarum Stirpium Americanarum
Historia_, Vindobonæ, 1763. According to this observer the seedling
falls from the tree in the twelfth month from the fecundation of the
flower. This happened in my observations on the same species in Fiji in
the eighth or ninth month. Jacquin states that the tip of the radicle
protrudes from the fruit in the third month, whilst my results give it
as taking place in the fourth month. The difference in the length of the
total period, it may be remarked, would be to a great extent determined
by the varying length acquired by the seedling before it drops from the
tree. In ordinary conditions it averages about ten or eleven inches, and
the hypocotyl itself attains a length of nine or ten inches on the tree,
both in Fiji and Ecuador; but in sheltered localities it may attain a
length half as long again. I have already pointed out in the case of the
fruits of Rhizophora mucronata that a year and more would be sometimes
required, and the same remark would apply to unusually long fruits of R.
mangle. Local conditions would often produce varying results, both in
the rate of growth of the hanging seedling and in the duration of the
period of its attachment to the tree; but it is probable that nine or
ten months would represent for the genus the average length of the
period between fertilisation of the ovule and the detachment of the
seedling from the parent tree.

_The mode of separation of the seedlings of Rhizophora mangle and
Rhizophora mucronata_

This is a process of expulsion almost akin to parturition, and is
brought about by the outward growth of the neck of the cotyledonary
body. There is much that is of great interest in this subject; and I may
add that Haberlandt, in a memoir published in the _Annales du Jardin
Botanique de Buitenzorg_ for 1894, gives the results of an elaborate
study of the viviparous process in this and other genera of mangroves.
The same analogy seems also to have presented itself to him, but only in
connection with the means employed in some of the genera, as with
Bruguiera, for conveying nourishment to the growing embryo. He remarks
that he was involuntarily reminded by these structures of the
chorion-tufts and lobes in the placenta of mammals, and that such
structures in the mammal are functionally nothing more than true
_haustoria_ as found in the viviparous mangroves.

When studying the germination of the American and Asiatic Rhizophoras in
Fiji, I observed that the neck of the cotyledonary body did not begin to
form, nor the inclosed plumular bud to show signs of differentiation,
until the hypocotyl had protruded about 4-1/2 inches with R. mangle, and
between 6 and 7 inches with R. mucronata. The neck of the cotyledonary
body then proceeds to grow in length, pushing before it the plumular end
of the embryo-seedling, which it surrounds as a sheath. This operation
continues until the hypocotyl has acquired a length of about seven
inches with R. mangle, and about nine inches with R. mucronata, when the
neck begins to protrude outside the fruit. The cotyledonary neck
proceeds with its growth, and by the time the seedling is ready to fall
from the tree it protrudes about an inch from the fruit-shell, having
carried the growing plumular bud with it. The plumular end of the
seedling has been now more or less expelled from the fruit-cavity, and
the connection between the suspended seedling and the fruit now alone
depends on a slight bond between the base of the plumule and the inner
margin of the cotyledonary neck, as indicated by a cross in the figures
given in the plate. The union is soon broken and the seedling falls.

Whether there is anything more than an analogy between the expulsion of
a Rhizophora seedling and the birth of a mammal seems most unlikely; but
the process is at all events a very remarkable one.

_The means of dispersal of the genus Rhizophora_

My experiments and observations were for the most part made on the
Asiatic and American species in Fiji; but I enjoyed the opportunity of
confirming some important points on the coast of Ecuador. We can only
look to the currents for the explanation of the capacity of the genus to
cross tracts of ocean; but, given this capacity, there is much that is
difficult to understand in the distribution of the genus and of a
species like Rhizophora mangle; and it is probable that we shall have to
look behind the means of dispersal to a distant age in the distribution
of shore-plants of the mangrove type.

When Schimper published his work on the Indo-Malayan strand flora in
1891, but little was known of the duration of the floating capacity of
Rhizophora seedlings (p. 166). In giving the results of my
investigations I am merely describing the agencies of dispersal at
present in operation. Such agencies have their limitations, and we may,
perhaps, be thus able to explain why Rhizophora is restricted in the
Pacific islands to the archipelagoes of the Western Pacific; but many
serious objections would at once present themselves if we regarded the
occurrence of the genus in America, as well as in Asia and Africa, as a
matter depending on capacities and means of dispersal.

The fruits of Rhizophora, as they display themselves before the
protrusion of the germinating seed, have no buoyancy, and the
germinating fruits until the hypocotyl has protruded for some inches (6
inches in the case of R. mangle) also sink in sea-water. With a further
increase in the length of the hypocotyl, the germinating fruit acquires
buoyancy; and when the seedling, usually 10 or 11 inches in length,
becomes detached from the fruit on the tree and falls into the sea, it
floats readily in 95 per cent. of the cases. Such seedlings occur very
commonly in the floating drift of the estuaries and out at sea both in
Fiji and in Ecuador.

Out of five seedlings of the Asiatic species, Rhizophora mucronata, that
had fallen naturally from the tree, three were afloat and healthy after
eighty-seven days’ immersion in sea-water. Out of twenty seedlings of
the American mangrove, Rhizophora mangle, sixteen floated after ninety
days and four were afloat and healthy after one hundred and twenty days,
the greater number sinking during the fourth month. These results
indicate considerable powers of buoyancy, and go to show that extensive
tracts of ocean could be traversed by the floating seedling.

It should, however, be observed that not all the full-sized seedlings
float. With Rhizophora mangle about 5 per cent. sink in sea-water and
from 20 to 50 per cent. sink in fresh-water; whilst with R. mucronata
the proportion of non-buoyant seedlings is rather greater. There would
thus appear to be a rather nice adjustment of the specific weight of the
seedlings to the density of sea-water. Generally speaking, they may be
seen floating vertically or steeply inclined in the fresh-water of
estuaries and horizontally in the sea. With the buoyant seedlings of
Rhizophora mucronata, as a rule, about 90 per cent. float horizontally
in sea-water, and about 70 per cent. float vertically or steeply
inclined in fresh-water. The same general rule applies to R. mangle,
whether in the rivers and seas of Fiji or in those of Ecuador. In those
cases where the seedling drops prematurely on account either of storms
and floods or of the depredations of a grub that frequently attacks the
fruit, this rule would not apply. One may frequently notice in Fiji
after heavy weather that seedlings detached prematurely, and often
carrying the fruit, are floating in numbers horizontally in the rivers.
In a few days, as a rule, the fruit-case becomes detached and sinks.

It may be remarked that the horizontal position is much better adapted
for the safety of the seedling in transport than the vertical position.
In the last case the plumule, which protrudes above the water, would be
unable, as indicated in my experiments, to withstand the scorching rays
of the sun in a smooth sea; whereas in the horizontal position, which
the seedlings assume in sea-water, the plumule is more or less
completely submerged, and the risk of withering in the sun is very much
less. The Rhizophora seedlings would certainly have little chance of
crossing in safety a large tract of sea, if they floated, as they do in
river-water, with the plumule exposed above the surface. It is not
unlikely that the comparatively restricted area occupied by Rhizophora
conjugata may be due to the attitude its seedlings assume when floating
in sea-water.

The stranded seedlings of Rhizophora readily establish themselves for a
while in very different situations; and it is by no means necessary that
they should be washed ashore on a muddy coast. When half-buried amongst
the heap of vegetable drift piled up on a sandy beach they are
frequently to be found striking into the sand and showing their first
leaves. Here they ultimately perish in the great majority of cases; but
when protected long enough to reach the moist sand four or five inches
below, they may give rise to a little mangrove colony. When caught in a
fissure in the bare reef-flats these plantlets are sometimes able to
establish themselves. Rhizophora seedlings would, however, require a
coast prepared by them by the work of ages before they could form
extensive swamps. It is, therefore, not surprising that Prof. Penzig
found no evidence of mangrove-settlements on the shores of Krakatoa
fourteen years after the eruption.

Yet suited as Rhizophora seedlings are for crossing tracts of sea, I
regard them as quite unfitted for being transported by the currents
unharmed across an ocean. The plumular bud is insufficiently protected
for such a long voyage of many months, and perhaps of years. Though the
horizontal position of the seedling would secure the plumule against
being scorched in the sun, it increases considerably the risk of injury
from direct impact.

As bearing on their capacity for dispersal in other fashions, it may be
remarked that Rhizophora seedlings can withstand long drying. Five which
had been kept dry for nine weeks, after having been found stranded on a
beach, were planted in the mud of a mangrove-swamp. In a fortnight two
of them were developing the first leaves and throwing out roots. As long
as they are protected by a covering of vegetable _débris_ and sand, the
stranded seedlings might retain their vitality for months.

BRUGUIERA RHEEDII (Blume)

This species is reduced in Hooker’s _Flora of British India_ to
Bruguiera gymnorhiza (Lam.), and thus viewed it has a very wide range in
the Old World, corresponding very much to that of Rhizophora mucronata,
namely, tropical East Africa, tropical East Asia to the Liukiu Islands,
the Indian Archipelago, New Guinea, tropical Australia, and Western
Polynesia, as in New Caledonia, Fiji, Tonga, and Samoa. There are four
or five species of the genus, but all are confined to the Eastern
Hemisphere, none occurring in America.

As with the species of Rhizophora, this plant is indebted for its
present dispersal to the floating seedling, which, however, often falls
from the tree whilst still attached to the fruit, but is generally freed
in a day or two. The seedlings float for a long time in sea-water. I
kept one of them afloat for 117 days, when it was quite sound and
healthy. They appear to be better fitted than the species of Rhizophora
for the “rough-and-tumble” of ocean transport, since the plumule is much
less prominent, projecting only one line (2·5 mm.) or less, whilst with
the two Fijian species of Rhizophora the plumule measures from seven to
twelve lines (18 to 30 mm.). In the latter part of the year they are to
be found in abundance in the floating drift of rivers, and there they
readily develop the first leaves and roots. They are also frequent in
the sea off the coasts, and they are stranded in large numbers on the
beaches, where they readily strike into the sand when partially buried
amongst the vegetable drift.

The empty flowers and the germinated fruits containing the cotyledons
are very common in floating drift. They look much alike, but the flowers
are much smaller and possess the long style, whilst the fruits contain
the cotyledons at the bottom of the seed-cavity.

As with Rhizophora, there is a rather curious adjustment of the buoyancy
of the seedling to the density of sea-water. About 75 per cent. of those
afloat in the fresh-water of rivers assume the vertical position, the
plumular end protruding between two and five lines (5 to 12 mm.) above
the surface, while the remainder float horizontally or nearly so. In
sea-water about 50 per cent. float either vertically or steeply
inclined, and the other half float horizontally.

With regard to the times of flowering and fruiting, it may be remarked
that the trees are mostly in flower during the hot months from November
to February, and that the fruiting is in active operation in the latter
half of March. The floating seedlings occur in abundance in the
river-drift at the end of the year, a circumstance which corresponds
with the fact that a period of six months passes between the
fertilisation of the ovule and the fall of the seedling into the water.

Fertilisation, or, more correctly speaking, the discharge of the pollen,
takes place after the opening of the flower, and not before, as in the
case of the species of Rhizophora. The flower-bud is at first erect, but
subsequently it begins to bend downwards, and ultimately it hangs more
or less vertically. The provision to secure fertilisation under these
circumstances is rather curious. Without some such contrivance as is
below described, the pollen would merely fall out of the flower. Each
petal has its sides rolled or folded inwards so as to completely inclose
two stamens. In the bud the folded petals are white and flexible, but as
the flower expands they redden and become dry and elastic, and are only
prevented from flying open with a spring by the interlocking of the
hairy tips of their lobes. Whilst the folded petals are becoming stiff
and elastic during the opening of the flower, the inclosed stamens are
at the same time preparing themselves for their function. The anthers
are dehiscing and the filaments are acquiring elasticity. All is now
ready, and a slight shake or a touch puts the mechanism into action. The
petals unfold themselves with a spring, and the stamens thus suddenly
exposed and released fly forward, and a little shower of pollen is
thrown towards the centre of the flower. This process is accomplished in
ordinary fine weather during the first twenty-four or thirty-six hours
after the expansion of the flower. When the opening occurs in the early
morning, half of the stamens will be found released in the evening and
the rest on the following day. During the next day or two the petals and
the stamens fall out of the flower. In wet weather, the petals never
acquire elasticity, and in consequence do not unfold. In this case
pollenisation is never effected, and the folded petals soon fall to the
ground, carrying the stamens within them. Cross-fertilisation would be
much more likely to occur with species of Bruguiera (if, as is probable,
the same process of pollenisation is usually followed) than with species
of Rhizophora, since the stamens are securely inclosed in the petals for
some hours after the expansion of the flower.

Nearly eight weeks pass between the date of fertilisation and the
commencement of germination. This is somewhat similar to the period
given for Rhizophora mangle, namely, nine weeks, and it obviously leaves
little or no time for any stage of quiescence or dormant vitality in the
case of the seed. The changes which the fruit undergoes in this interval
are a considerable increase in girth and a thickening of the calycine
walls, together with a contraction of the mouth of the tube. However, I
found no method sufficiently accurate for recording the rate of increase
of the fruit.

It is known that germination is in progress when the end of the
hypocotyl begins to lift up the lining membrane at the bottom of the
calycine tube (see Figs. 21 to 26). The floor of the tube begins to
bulge up, but since this cannot be well seen at first, a better index is
afforded in the elevation of the style which accompanies it. The top of
the style preserves previous to this time a constant level with regard
to the tips of the calycine teeth. But this does not indicate the actual
beginning of germination. As shown in Fig. 21, the seed lies about two
and a half lines (6 mm.) below the floor of the calycine tube, and the
tip of the hypocotyl has to penetrate the intervening tissues before it
can push up the lining membrane and raise the style. Judging from the
subsequent rate of growth, seven or eight days at least, and perhaps as
much as two weeks, are requisite for this purpose. It is not necessary
to give further details here, and it may be at once stated that the
average of numerous observations on the length of the interval between
fertilisation and the elevation of the style was sixty-four days, the
range being fifty-nine to sixty-nine. After deducting ten days for the
time occupied for the radicle in reaching the floor of the calycine tube
(see Figs. 22 and 23), we obtain, as already remarked, nearly eight
weeks as the time elapsing between fertilisation and germination.

The radicle or hypocotyl, therefore, in the first stage of germination
pierces the tissues above it and reaches the floor of the calycine tube.
It does not, however, pierce the lining membrane of the tube but pushes
it upward until it ruptures about 4 millimetres below the base of the
style which is carried up with it. Thus a kind of cap is formed, as
shown in Fig. 24, which does not fall off from the end of the hypocotyl
until it has protruded rather more than an inch. The hypocotyl attains a
length varying between 5 and 11 inches, the average being about 8
inches.

The whole period may be thus divided up:—

(1) Period between fertilisation and germination 7-1/2
weeks.

(2) Period between the beginning of germination and the 1-1/2
protrusion of the point of the hypocotyl at the floor of the weeks.
calycine tube

(3) Period occupied in the growth of the hypocotyl 8 inches 18 weeks.
outside the fruit and terminating in the fall of the
seedling

——

Total 27 weeks.

The total period of twenty-seven weeks between fertilisation and the
fall of the seedling is thus six weeks shorter than that estimated for
Rhizophora mangle. On comparing the two tables it will be seen that the
difference mainly lies in the length of the second period, namely, that
between the commencement of germination and the protrusion of the
hypocotyl from the fruit. With Rhizophora mangle the fruit grows
considerably in length during this period of the germinating process. On
the other hand with Bruguiera rheedii there is, during this period,
practically no increase in the length of the fruit, and the radicle has
only to penetrate the tissues, 2-1/2 lines in thickness, between the
seed and the floor of the calycine tube.

In the mode of separation of the seedling there are very marked
differences between this species of Bruguiera and the species of
Rhizophora. With Bruguiera rheedii the four small cotyledons, which are
united at the base, are, however, left behind at the bottom of the
seed-cavity, when the seedling is detached. But there is no expulsion of
the seedling, the connection being ultimately severed at the contracted
base of the cotyledons inside the fruit. When the seedling is full-sized
the nutritive supply begins to fail, and in consequence the pressure of
the sides of the fruit on the inclosed plumular end of the seedling
becomes slacker, the union with the cotyledons becomes weaker, and the
connection of the fruit with its peduncle at the basal joint becomes
slighter. Usually the fruit falls before the seedling is ready to drop
out, and the connection is severed after a few days’ flotation in the
water; but sometimes the union between the seedling and fruit is weaker
than that between the fruit and its peduncle, and in that case the
seedling falls and leaves the fruit containing the cotyledons on the
tree. The whole process of separation is much simpler than with species
of Rhizophora. Here it is mainly a matter of the failure of the
nutritive supply, whilst with Rhizophora it is almost a process of
parturition.

Haberlandt, in the memoir before quoted, describes quite a different
mode of detachment in the case of Bruguiera eriopetala. Here the
seedling falls normally whilst still attached to the fruit, and the
separation is subsequently effected by the expansion of the mouth of the
calyx-tube due to the swelling of the “endosperm-neck” from the entrance
of water.

_Summary_

(1) There are four typical mangroves of the Rhizophoraceæ in Fiji,
Bruguiera rheedii, Rhizophora mucronata (the Asiatic species),
Rhizophora mangle (the American species), and the Selala, a seedless
form intermediate between the two species of Rhizophora just named, but
nearest to the Asiatic species.

(2) It is shown that the sterility of the Selala is connected with the
impotent character of the pollen; and since the ovules appear capable of
fertilisation this is held to indicate that cross fertilisation has not
been in operation in producing the barren form.

(3) Good reasons are given for the belief that the Asiatic species of
Rhizophora is the parent of the Selala, not as the result of a cross
between the Asiatic and American species, but as connected with
dimorphism, the Asiatic species producing two kinds of offspring, one of
them with impotent pollen.

(4) In support of this view it is pointed out that there are two forms
of Rhizophora mangle in Ecuador, one of which comes near the Fijian
Selala, though producing seed. There could thus be no question of
crossing, since but one species occurs there.

(5) The Selala reproduces itself in a vegetative fashion when growing,
as it often does, in an inclined position. The parent trunk dies and the
primary branches supported by the aërial roots, remain alive and in
their turn give rise to secondary branches similarly supported.

(6) Although, as a rule, only one of the four ovules of Rhizophora
becomes a seed, occasionally a fruit contains more than one seed. With
R. mangle in Fiji about one per cent. of the germinating fruits
displayed more than one hypocotyl.

(7) As a result of a protracted series of observations in Fiji, it was
established that in the case of a seedling of average length of
Rhizophora mangle a period of thirty-three weeks elapsed between the
date of fertilisation of the ovule and the detachment of the seedling
from the tree. In the instance of R. mucronata it was placed at
forty-two weeks. A period of thirty-eight weeks, or nine to ten months,
is regarded as typical for the genus.

(8) It is established that normally there is no rest-period for the seed
in the case of Rhizophora, the seed at once beginning to germinate on
reaching maturity. In those exceptional instances, however, where there
is more than one seed, it is shown that in some cases the seeds do not
begin to germinate together, and that a rest-period of at least some
weeks can be at times postulated for one of the seeds.

(9) An analogy exists between the process of expulsion ending in the
detachment of the seedling of Rhizophora from the fruit and the process
of parturition.

(10) Experiments show that Rhizophora seedlings can float unharmed in
sea-water for a period of at least three or four months. Though
nine-tenths or more float in sea-water, as much as a fourth or a half
sink in fresh-water. As a rule they float vertically in fresh-water and
horizontally in sea-water, the horizontal position safe-guarding the
plumule against the risk of being withered up by the sun in a calm sea.

(11) It is shown that in the case of Bruguiera rheedii the seedlings
when detached from the tree can float unharmed in sea-water for months.
In their specific weight they display a similar fine adjustment to the
density of sea-water, as is above described in the case of Rhizophora.

(12) With this species of Bruguiera, fertilisation takes place not in
the unopened flower, as in Rhizophora, but after the flower’s expansion;
and a very singular mechanism is here described which secures the
completion of the process.

(13) A period of twenty-seven weeks elapses between the fecundation of
the ovule and the detachment of the seedling from the tree in the case
of Bruguiera rheedii; and it is shown that there is normally little or
no room for any rest-period, and that, as with Rhizophora, the seed on
reaching maturity begins to germinate.

(14) Though the seedlings of Rhizophora and Bruguiera could be
transported in safety a few hundred miles across the sea, it is held
that they could never cross the Pacific and reproduce the plant. That
the American species of Rhizophora has reached the Western Pacific from
the New World is not accepted. Rather is its present distribution
regarded as representing its original wide range over much of the
tropical zone.

CHAPTER XXXI

A CHAPTER ON VIVIPARY

The significance of vivipary.—The scale of germinative capacity.—A lost
habit with many inland plants.—The views of Goebel.—The shrinking in
the course of ages of tropical swamp areas.—The variation in the
structures concerned with vivipary.—Abnormal vivipary.—Summary.

IT was remarked in Chapter IX that the study of the germination of the
floating seed carried us to the borderland of vivipary; and we may now
observe that our study of the mangroves, Rhizophora and Bruguiera, in
the previous chapter, has brought us into contact with vivipary in its
most complete development in the tropical swamps of our age. There is a
great gap between the two extremes, represented by the occasional
germination of a seed in a capsule or in a berry on the plant, and by
the elaborate process of vivipary exemplified by Rhizophora; but most of
the intermediate stages can be illustrated by known examples of
vivipary. There is, however, no pretension to deal with this subject
here in anything but a cursory fashion; but it will, I venture to think,
add completeness to a work in which germination on and off the plant has
been such a frequent theme if I endeavour to connect together some of
the various sets of facts known to us concerning germination from the
standpoint of vivipary.

The principal argument here followed has been already outlined in
Chapter IX, where I have remarked that it is possible to construct a
scale of the germinative capacities of plants, presenting a continuous
series beginning with the mangroves, where germination takes place on
the tree, and ending with those numerous inland plants where seeds are
liberated in an immature condition. It is suggested that vivipary was
the rule under the uniform climatic conditions of early geological
periods, and that with the differentiation of climates that has marked
the emergence of the continents the viviparous habit has been lost over
much of the globe, the mangrove-swamps alone illustrating the climatic
conditions once prevailing. The rest-period of the seed is regarded as
an adaptation to climatic differentiation and to seasonal variation; and
even the seed-stage may be broadly regarded as the price paid for
adaptation on the part of the evolutionary or determining power that
lies behind plant-development. When discussing the germination of
Cæsalpinia in Chapter XVII, I have shown that the contraction and
induration of the seed-tests appear merely as an adaptation to climatic
differentiation and to seasonal variation, and that it would be quite
possible by exposing the maturing seed to very warm and moist conditions
to induce germination without any rest-period, as actually occurs with
Rhizophora. One would then dispense altogether with the final processes
of the contraction and induration of the seed-coats, as illustrated in
the Leguminosæ; and the rest-stage would appear as an adaptation to
secular differentiation of climate in the later epochs of the world’s
history.

The significance of occasional vivipary was long ago pointed out by
Goebel in his _Pflanzenbiologische Schilderungen_ (teil I., 117-134,
Marburg, 1889), when he observed that vivipary, as displayed in the
mangroves, and particularly in the Rhizophoreæ, represented the fullest
expression of a habit that is only occasionally exhibited by other
plants under exceptionally moist conditions. His view was that the seeds
of plants living in wet places are suited in a varying degree for rapid
germination, and that vivipary presents itself as the most complete
development of this capacity. If I regard the views of Goebel and of
Kerner aright, vivipary as normally developed in the mangrove is to be
traced in a descending scale to small beginnings, the principal
determining condition lying in the great difference that exists amongst
plants in the readiness of the seed to germinate. In the ascending scale
we would have first the detachment of the immature seed, where the
embryo is often in a rudimentary state, the ripening of the seed taking
place in the soil. Then would come those plants where the seeds on being
detached are quite mature and are ready to germinate as soon as they
fall to the ground. Then would follow the stage represented by those
plants where the seeds merely begin to germinate on the plant, such as
occurs more or less normally with some mangroves like Laguncularia, and
abnormally with a number of plants living in drier stations. After this
come those mangroves, where, as in Avicennia, germination is completed
on the tree or shrub, but the seedling at once liberates itself from the
parent. Last of all there is the stage of the typical mangroves,
Rhizophora and Bruguiera, where the seedling remains for months growing
on the tree and hangs from the branches.

Vivipary, as above stated, presents itself as a matter of small
beginnings. My own view, however, is that it is a matter of small
“endings”; and that if we were to commence the scale not with the
immature seed lying on the soil, but with the seedling suspended from
the branches of a Rhizophora tree, we should record the various epochs
in the history of vivipary throughout the plant-world. From this
standpoint the occasional cases of incomplete vivipary displayed outside
the mangrove-swamp represent a lost habit belonging to a primeval period
when the climatic conditions were uniform over most of the earth, an age
almost of eternal gloom, when the air was ever saturated with aqueous
vapour, and when the sun’s rays were screened off by a dense
cloud-covering that enveloped the globe, an age of which the existing
mangrove swamps alone afford an imperfect indication. Yet even now we
can say with Schimper that “dense and frequently repeated cloudiness
apparently represents the most essential climatic condition for the
occurrence of mangrove in the tropics” (_Plant Geography_, p. 409).

But, to return to the subject immediately under consideration, if my
view is correct we ought to find indications of the lost habit in the
anomalous structure of the seeds of some inland plants; and, indeed, it
is shown in Note 50 that this view can be taken of the singular
structure of the seeds of the Myrtaceous genera, Barringtonia and
Careya, and of the genera of some other orders, and can be extended by
implication to several other plants possessing similar seed-structures.

With regard to the subject generally, it may be remarked that although
normal vivipary is mainly restricted to the plants of a mangrove swamp,
by no means all mangrove plants are typically viviparous. This habit in
its most complex form is exhibited as a rule by plants with firm,
somewhat fleshy, usually one-seeded, indehiscent fruits, such as we find
with Rhizophora and Bruguiera; but plants with follicular fruits, such
as occur with Ægiceras, may also display it in a fashion nearly as
complex. Generally speaking, however, plants with hard, dry fruits, such
as are owned by Excæcaria, Heritiera, and Lumnitzera, are
non-viviparous, though to all appearances quite at home in a
mangrove-swamp. Others again, like Carapa, Laguncularia, and Nipa,
whilst displaying vivipary in a varying degree, in some cases as a
general rule, in others only occasionally, exhibit no special structures
connected with it. This point is well brought out by Schimper in his
work on the Indo-Malayan strand-flora (p. 43), and no further mention
need be made of it here.

The structures connected with vivipary vary greatly in their degree of
specialisation. At the one end of the scale we have highly complex
structures, such as are described in the preceding chapter. At the other
end we have those cases of occasional germination on the parent plant
where there is seemingly no special structure of any sort. That the
complex arrangements concerned with the vivipary of Rhizophora,
Bruguiera, Ægiceras, and Avicennia are adaptations is argued by
Haberlandt and Schimper, both of whom devoted much attention to the
study of these plants. This is seemingly indicated by the circumstance
that complex structures concerned with vivipary are found in plants so
divergent in their characters (the four genera above-named representing
three orders, Rhizophoreæ, Myrsinaceæ, and Verbenaceæ) that they only
possess their stations in common. It does not, however, follow that all
mangroves that exhibit a complex form of vivipary are of the same
antiquity. I should be inclined to regard those of the Rhizophoreæ as
the more primitive types, whilst it is possible that plants of other
orders, though ancient denizens of a mangrove-swamp, may be more recent
intruders into the mangrove-formation after the differentiation of a
dry-land flora.

Of particular interest in this connection are the cases of abnormal
vivipary, or of “precocious germination,” that have been recorded from
time to time respecting a number of plants not denizens of a mangrove
swamp, none of which would appear, according to Schimper’s views, to
present anything of the nature of an adaptation. Goebel mentions a
number of instances, such as that of wheat-grains germinating on the
stalk in a wet summer, and that of Dryobalanops camphora, the Borneo
camphor-tree, when during a prolonged wet season in Java the seed
germinates in the fruit on the parent tree. Amongst other examples he
cites the Cacti, Epilobium, Agrostemma, and Juncus, the last case coming
also under my observation in a wet season in England. One may here
notice the instance of Dracæna, of which Mr. Hemsley, in April, 1902,
exhibited at a meeting of the Linnean Society of London a specimen
showing the seeds germinating in the berries on the plant.

Several cases of this kind came under my notice in Fiji. Pulpy fruits
rather favour the precocious germination of seeds. Thus I sometimes
found the seeds germinating in the Mandarin orange and in the Papaw
fruit (Papaya) shortly after they had been gathered. But more
interesting examples were displayed in those instances where the seed
was found germinating on the plant. When the Convolvulaceæ grew in wet
situations, as on the borders of a mangrove swamp, the seeds were
sometimes observed germinating in the capsule. This came under my notice
with Ipomœa glaberrima (Boj.) and with I. peltata, more particularly in
wet weather. With some other plants, like Hibiscus diversifolius, that
grow in wet places, this at times occurs. A species of Croton, employed
as a support for the Vanilla plants in a plantation near Suva, displayed
seeds germinating on the plant. I was informed that the seeds of the
common cultivated Luffa (L. cylindrica) growing in a garden on Vanua
Levu sometimes germinated in the fruit still attached to the parent. It
is possible that the seeds of the parasitical genus, Myrmecodia, may
occasionally germinate on the plant, since I found them germinating
inside some of the small berries that had been lying forgotten within a
newspaper for a fortnight.

Perhaps the most curious case of abnormal vivipary observed by me in
Fiji was that concerned with the Coco-nut palm. Though not known to many
residents in the island, this habit was described to me by Mr. Matthew
Simpson, a planter on Vanua Levu, who told me that he had noticed nuts
germinating on the tree in unusually dry seasons. Coco-nut palms
displaying the nuts germinating on the tree came under my observation
near Bale-bale, Savu-Savu Bay. In these cases the mature fruit, instead
of falling, remains attached and dries on the stalk. In one case the
seedling was about eighteen inches high. This seems to be what takes
place normally according to Blume with Nipa fruticans, the swamp palm of
Indo-Malaya. Goebel quotes this author to the effect that the fruits are
not separated from the head before germination is so far advanced that
sea-water can no longer injure the seedling. The fruits, we are told,
may remain for years attached in a state of incomplete germination.

_Summary_

The scale of germinative capacity, that begins with the seedling hanging
from the branches of a mangrove like Rhizophora and ends with the
detached immature seeds of many inland plants that only germinate after
lying for some time in the soil, is regarded as supplying a record of
the various epochs in the history of vivipary throughout the
plant-world. In the occasional cases of incomplete vivipary occurring
among inland plants and in the singular structure presented by the seeds
of certain genera of the Myrtaceæ and other orders we perceive
indications of a lost viviparous habit belonging to a primeval period
when vivipary was the exception and not the rule, an age when the same
climatic conditions prevailed over much of the globe. At such a period
the sun’s rays were screened off by a dense cloud-covering that
enveloped the earth, and the atmosphere was ever charged with moisture.
With the differentiation of climate that has marked the emergence of the
continents during the secular drying of the earth, the viviparous habit
has been alone retained within the confines of the mangrove-swamp, where
the conditions once almost universal now survive; and as an adaptation
to the differentiation of climate and to the resulting seasonal
variation the rest-period of the seed has been developed.

CHAPTER XXXII

THE WEST COAST OF SOUTH AMERICA

MY acquaintance with the strand-flora of the west coast of South America
began at Corral, the port of Valdivia, in Southern Chile in lat. 40° S.,
and terminated at the mouth of the Guayaquil River, in Ecuador, about 2°
south of the equator. During the period December 23, 1903, to March 17,
1904, I examined the coast plants at sixteen localities in this region,
which covers 38 degrees of latitude and thus measures about 2,300 miles.
Travelling in a steamer to Callao that was trading on the coast I had
opportunities of staying for periods ranging from half a day to a couple
of days at a considerable number of places; and a week spent at
Valparaiso gave me a good opportunity of examining the beaches north and
south of it. At Lima I spent some weeks, and from that centre examined
the shore-plants at Callao, Ancon, and Chancay to the northward. North
of this I had not the same opportunities, until we passed the Peruvian
and Ecuadorian boundary; but from a visit to the shore at Paita, from
the general look of the country in places as we coasted along, and from
information derived from other sources, I was able to obtain a fair
general idea of the prevailing character of the beach plants. After my
previous experience to the southward, one could fairly gauge the
character of the beach-flora from the appearance of the land behind. In
the Gulf of Guayaquil and in the vicinity of the city of that name I
spent about three weeks in the investigation of the coast flora.

If it were not for the interposition of the great rainless deserts of
Northern Chile and for the scantily vegetated, scantily watered and
semi-sterile condition of almost the whole coast of Peru, the botanist
would be presented with a splendid opportunity of studying the
distribution of shore-plants along a meridian stretching through some
fifty degrees of latitude from Patagonia to Ecuador. As it is, drought
and sterility in one form and another reign over about half of this
great stretch of continental coast. This is reflected in the
beach-flora; and though the observer will often have his interest
attracted by the wonderful climatic anomalies arising from the presence
on the coast of the cold Humboldt current, to which the sea-border of
North Chile owes its desolation and the coast of Peru its
semi-sterility, yet for a long time he will feel as if Nature had hardly
dealt fairly with him.

[Illustration:

THE WEST COAST
OF
SOUTH AMERICA

John Bartholomew & Co., Edin^r.
]

Along the sea-border corresponding to the deserts of North Chile there
would seem to be practically no plants growing on the beaches, except
here and there where some stray plant from the saline districts inland
intrudes on the coast. Along the whole sea-border of Peru from Arica
north to Tumbez on the borders of Ecuador, the coast-districts, though
more or less rainless, receive the benefit of the drizzly garuas and
sea-fogs, and the sterility of the land immediately backing the beaches
is much less pronounced than with the sea-border corresponding to the
deserts of Northern Chile. This difference shows itself in a peculiar
type of littoral vegetation, a strand-flora that is very scanty but one
where on the beaches Sesuvium prevails. North of Tumbez the
mangrove-formation predominates along the sea-borders of Ecuador and
Colombia to Panama, excepting on a stretch of sterile coast extending
north from the Gulf of Guayaquil to the equator.

Though in one sense the botanical observer will be disappointed with the
littoral floras of the west coast of South America, in another sense
when he remarks the manner in which the coast-vegetation reflects the
abrupt changes in the prevailing climatic conditions he will be
fascinated by the interesting problems presented to him. We are
accustomed to connect a tropical coast with mangroves, coral-reefs, and
beaches of calcareous sand supporting a luxuriant littoral flora.
Climatic conditions banish all these from the tropical west coast of
South America until within four degrees of the equator, and then with
startling suddenness the dominion of the mangrove begins, the
neighbouring hills commence to be clothed with tropical jungle, and the
climate is completely changed. Mr. John Ball, who sailed along this
coast about twenty years ago, referring to this remarkable phenomenon on
the borders of Peru and Ecuador, remarks that no such abrupt and
complete change both in climate and vegetation is known elsewhere in the
world, and he adds that few parts of the American coast better deserve
careful examination (_Naturalist in South America_). This subject has
since been discussed at length by Dr. Wolff in his “Geografia y
Geologica del Ecuador,” and by Baron von Eggers in a paper to be
subsequently quoted, two very competent observers, but the latter
considers that the subject still requires a systematic investigation,
and suggests that an observing station should be established on this
coast by the combined meteorological societies of Europe. A sojourn of
more than a week in the swamps at Puerto Bolivar, a few miles from
Tumbez, enables me to appreciate the nature of the problem, and to throw
a little light on the line of investigation required.

But to return to the general subject of the littoral floras of the west
coast of South America, I may say that beginning with the island of
Chiloe in lat. 42° S., this coast may be divided into four zones.

(1) The Convolvulus soldanella zone of Southern Chile, which extends as
far north as Coquimbo about 30° S. lat.

(2) The Plantless or Desert zone stretching north to the vicinity of
Arica in lat. 18° 30ʹ, and corresponding to the coast of Northern Chile.

(3) The Sesuvium zone, extending north from Arica to the 4th parallel of
south latitude in the vicinity of Tumbez, a sea-border of semi-sterility
that comprises the entire coast of Peru.

(4) The Mangrove zone, stretching from Tumbez, on the frontiers of
Ecuador, to the equator and on to Central America, but interrupted at
first by a strip of sterility on the coast extending from the Gulf of
Guayaquil to the borders of Colombia, or, strictly speaking, to the
equator.

THE CONVOLVULUS SOLDANELLA ZONE (SOUTHERN CHILE).

This zone, which answers to the coast of Southern Chile, from Chiloe as
far north as Coquimbo, corresponds to watered and vegetated inland
regions, in which, however, the amount of rain and the degree of
fertility decreases from south to north, that is to say, as we approach
the desert regions. Here we find none of the dry beaches that prevail
for twenty-five degrees of latitude north of Coquimbo. When we scoop
with our hands to a depth of three or four inches in the sand we find it
relatively cool and more or less moist, as in an English beach. In a hot
summer’s day on a Valparaiso beach we should find that the temperature
of the sand at the surface (half-inch deep) was about 112° F., and at a
depth of four inches about 80°. This would be above the average for the
zone, which would be probably near the typical summer-temperature of an
English beach, namely, 102° at the surface and 77° four inches down.
This subject of beach temperature is discussed in Note 70.

Plants typical of the beaches of this zone, and evidently occurring over
the length of it, are Convolvulus soldanella, Nolana (paradoxa?);
Polygonum maritimum, Salsola Kali, and Selliera radicans. Nolana is a
Chilian and Peruvian genus. This beach plant, which is especially
abundant on the beaches near Coronel and at Bahia San Vincente, has the
creeping habit of its associate, the Convolvulus. However, it possesses
seeds, or rather seedvessels, of more limited buoyancy; and it is shown
in Note 71 that prolonged drying is needed for effective dispersal by
currents over great distances. This beach species of Nolana has narrowly
escaped being a widely-spread littoral plant; whereas it is now
restricted to the Chilian beach flora. Selliera radicans, a little
creeping Lobeliaceous plant, growing under the shade of tall clumps of
Juncus at the edge of the beach or in wet places where springs ooze out
in the sand, is a very interesting species that occurs also on the other
side of the Pacific in Australasia. Of the mode of dispersal of its
small seeds I know nothing, as the fruits were not ripe at the time of
my visit; but I would suggest that some resident botanist should
investigate this important point. I found it at Corral and at Coquimbo;
and Gay speaks of it as growing on wet coast places from Chiloe to
Coquimbo, a range of 12-1/2° of latitude.

It is probable that all the shore-plants of this zone extend south to
Chiloe in latitude 42° S.; and it is likely that some of them reach
towards the Straits of Magellan. I did not find any of them within the
Straits on the beaches in the vicinity of Punta Arenas, where, however,
I noticed the three plants recorded by Ball, namely, Armeria maritima,
var. andina; Senecio candidans, also found in the Falkland Islands; and
Plantago maritima; besides a Chenopodiaceous plant not in fruit. The
Plantago has no capacity for dispersal by currents, and probably none of
the other plants are thus dispersed. I formed the opinion when in the
Straits that the beach plants on the Pacific and Atlantic coasts of
Patagonia could have but little communication by the currents, and that
they are in this respect quite cut off from each other. A botanist who
investigates the strand-flora of Patagonia and Tierra del Fuego in
connection with the littoral plants of the opposite coasts ought, if he
has not already done so, to obtain some very interesting results from
the standpoint of plant-dispersal.

The northern limit of the plants of this zone near Coquimbo, in lat. 30°
S., is not determined by the change in climatic conditions that goes
normally with decrease in latitude, but by the vicinity of the great
deserts of Northern Chile, the aridity extending to the beaches.

Amongst the other plants occurring generally in the Convolvulus
soldanella zone of Southern Chile, species of Salicornia and Samolus are
to be observed in wet places. On the beaches near Valparaiso and in the
vicinity of Talcahuano there thrives a species of Franseria, a Composite
plant possessing prickly fruits well suited for conveyance in bird’s
plumage, but not adapted, as shown in Note 71, for dispersal by
currents. Mesembryanthemum is a typical beach-plant at Coquimbo, and an
intruder from the adjoining hill-slopes at Valparaiso. Raphanus,
seemingly R. maritimus, occurs in places, but apparently only as an
intruder from the cultivated districts behind the beaches. One or two
species of Euphorbia are not uncommon. A few small trees or bushes of
Acacia farnesiana grow typically on the beach at Coronel and in
neighbouring sandy tracts at Talcahuano, though the plant, as Gay
observes, has been introduced. Sophora tetraptera, found also in New
Zealand, and one of the most interesting plants of the Antarctic flora,
thrives as a small tree on the hill slopes overlooking the harbour of
Corral, becoming bushy where in places it intrudes on the beaches, and
fruiting there as freely as on the slopes above. It was by testing the
buoyancy of the seeds of this plant that I was led to the discovery of
its mode of dispersal by the currents (I am indebted to Mr. Holland for
the specific determination of the fruits sent by me to the Kew Museum).
Other shore-plants, of course, occur in this zone; but I have gone far
enough to illustrate the subject. Of the numerous occasional intruders
from the neighbouring inland districts, frequently Compositæ, I say
nothing. The results of my observations on the floating power of the
seeds and seed-vessels of some of the shore-plants of this zone are
given in Note 71.

Stranded seeds and fruits that belong to the proper beach-drift are not
easily found on the beaches of Southern Chile, as they are often buried
in rubbish. Those most characteristic are seeds of Convolvulus
soldanella and drupes of Nolana (paradoxa?), both typical beach-plants
of the zone. Portions of Salsola Kali bearing mature fruits, as
described in Note 17, are also frequent. Seeds of Sophora tetraptera
were found on the beach of Bahia San Vincente, whither they must have
been brought by the Humboldt Current from the south, as I did not
observe the tree in the vicinity. On this beach, as well as at
Valparaiso, the prickly fruits of Franseria were abundant in the drift,
doubtless derived from the plants growing on the same beaches. In
addition we get as frequent components of the beach-drift materials that
mark the white man’s presence over much of the globe. Corks are widely
distributed over the beaches of the world; but on no coast have I found
them more numerous than on the Valparaiso beaches. Here we find Medicago
fruits, the empty stones of the cherry, the plum, and the peach, empty
filberts and other materials, all of which I have gathered on the shores
of the Straits of Messina and on English beaches. Amongst this medley we
find also Casuarina cones and fruits of Eucalyptus. Then we find special
indications of the New World in the pea-nut (Arachis hypogæa) and in the
abundant seeds of a huge pumpkin (Cucurbita), which is a favourite food
with the Chilian indigenes. These seeds are cited as an example of
futile buoyancy in Chapter XIII.

THE PLANTLESS OR DESERT ZONE (NORTHERN CHILE).

This zone of the coast, which stretches north for some 700 miles from
Coquimbo to near Arica (30°-18°30ʹ S. lat.), corresponds to the great
desert region of North Chile. On the beaches of Antofagasta, Tocopilla,
and Iquique, which are situated in the midst of this zone, I found no
plants. This rainless sea border of barren mountains, presenting to the
eye of the traveller from the deck of a passing steamer nothing but rock
and sand, must be one of the most desolate coasts on our globe. It is
therefore not a matter for surprise that the beaches are of dry loose
sand in which the hand fails to find on scooping below the surface that
refreshing coolness which is the character of beaches in all latitudes
where the land is vegetated and a subsoil drainage seaward exists. Under
ordinary conditions the sensation of moisture in the sand a few inches
down is not produced by the mere proximity of the sea. On the
Antofagasta and Iquique beaches the temperature in the heat of the day
of the surface half inch ranged from 120° to 130° F., whilst four inches
down it was 95° to 100°, and no moisture was found by scooping five or
six inches down. On the Taltal beach, which lies towards the southern
end of the desert region, I noticed, besides a few plants of Suæda
fruticosa, two other species of the orders Santalaceæ and Nolanaceæ,
evidently intruders from the inland regions. Where the zone of extreme
aridity terminates at the north between Pisagua and Arica a few bushes
are to be seen on the hill-slopes behind the beaches.

Very little seed-drift came under my notice on the beaches of the desert
zone. Here and there I found a few Medicago pods and some seeds of the
large pumpkin above noticed, but that was all. This is due as a rule to
the seed-drift being masked by an enormous amount of rubbish, mostly
brought from the south by the Humboldt Current. My walk for five miles
along the beaches immediately north of Antofagasta gave me an experience
in the way of stranded drift such as I have never met with on the
beaches of any other region. All the dead bodies of the Chilian coast to
the southward seem to have been stranded in the bend of Moreno Bay, on
the shore of which Antofagasta lies; and the air was tainted with
decaying flesh, the past being mixed up with the present in a most
unrefreshing fashion. Besides carcases of sea-lions, six feet in length,
sharks, dog-fish, and fish of many sorts, some of them dried up, others
in a state of putrefaction, there were dead penguins, dead pelicans,
dead sea-birds of other kinds, the bodies of horses, cattle, dogs, &c.,
all preyed upon by the numerous vultures and skuas, and in some
localities by hungry-looking dogs of large size that took no notice of
me as they slunk along. The past was represented by great quantities of
bones that lay bleaching on the sand, with here and there a vertebra of
a whale, making in all quite a varied osteological collection. But this
was not all. Carcases of all sorts were drifting towards the beach. Here
a vulture, there a skua, there again a dog stood just beyond the
tide-wash looking keenly seaward; and by following the direction of
their gaze one could see that each had marked down a carcase slowly
drifting in. Now and then they would make a dash, scarcely waiting for
the new arrival to be washed up by the waves. But there was no
competition, since there was enough for all.

Under such conditions my investigation into the seed-drift was out of
the question; but I saw what would be considered by some as more
interesting, namely, the dead of many latitudes piled up on the beach by
the Humboldt Current.

THE SESUVIUM ZONE (THE PERUVIAN COAST).

This zone, which comprises the whole Peruvian sea-border from Arica in
18°30ʹ S. to the vicinity of Tumbez in about 3°30ʹ S., usually possesses
in its scanty littoral flora one or two species of Sesuvium, and in some
places Sesuvium alone occurs on the beach. The beaches here do not line
a region of almost complete aridity, as in the coast corresponding to
the great desert region of North Chile. Though here also scarcely any
rain falls, the sea-border receives the benefit of the “garuas” or
drizzling sea-fogs; whilst the region immediately behind the coast may
either be desert or semi-sterile during much of the year, or may be
scantily vegetated, or, as along the river-valleys, may display a
vegetation more fitting to the latitude. The general aspect, however, of
the coast of Peru is one of aridity; but there are probably few beaches
where a certain amount of subsoil drainage from the land sea-ward does
not exist. This is well exhibited at Ancon, north of Callao, where in
the most unlikely situations water is reached by digging wells; but in
spite of this the Sesuvium alone grows on the beach. The beaches
examined by me in the heat of the day in February, as at Mollendo and
Ancon, had much the same surface-temperature noticed in the preceding
month on the beaches of North Chile, namely, 120° to 130° F., and in one
place 135°; whilst at a depth of four inches the sand was rather cooler,
and instead of being between 95° and 100°, as on the Antofagasta and
Iquique beaches, it was here usually only about 90°. But it was only
occasionally that the sand felt at all moist at a depth of five or six
inches; and in this zone, therefore, only a few shore plants of a
peculiar type could be expected to find a station on the beaches,
excepting, of course, those localities where low marshy districts or
lagoons lie behind the beach.

The beach plants of the coast of Peru as observed by me though usually
scanty, presented two types according to the character of the district
bordering the beach. I make no mention here of those local plants, often
belonging to the Compositæ, that as at Callao and Arica descend the
valleys to the beaches, or to those numerous introduced plants that
accompany cultivation, such as we find at Arica. In those coast
localities, as at Arica, Callao, and Chancay, where salt-water pools or
brackish lagoons lie behind the beach, or where a stream or a river
empties into the sea, Sesuvium portulacastrum, Heliotropium
curassavicum, and a Salicornia are to be generally noticed, and, as at
Callao, Batis maritima may also abound. On the Chancay coast, about 30
miles north of Callao, there lies inside the shingle-beach a large
shallow lagoon of brackish water (spec. gr. 1·012) with extensive muddy
marginal flats, the temperature of the water at the edge being at
mid-day on Feb. 3rd, 90° F. In the water flourished Ruppia maritima,
which was also exposed in dead, dry, matted masses on the bordering
mud-flats. On these mud-flats grew Sesuvium portulacastrum, which near
the water’s edge was associated with a small species of Salicornia,
whilst further away from the water it was accompanied by Heliotropium
curassavicum.

But the most typical beach-flora of the Peruvian coast is such as we
find on the dry beaches skirting the base of sand-covered or barren
hill-slopes such as occur at Mollendo, Ancon, and Paita. As at Ancon,
sand-covered hills and plains may extend miles inland, displaying here
and there lines of shifting sand-mounds or “medanos.” On such beaches we
may often find only a solitary plant, a species of Sesuvium which seems
to differ only in its larger flowers, its much larger leaves (2 inches
long), and its stout stems, of the thickness of the little finger, from
the ordinary Sesuvium portulacastrum. This seems to be the only plant
that can make its home on such beaches. At Mollendo, where there are
signs of desiccated pools behind the beach which are occasionally filled
with sea-water, the vegetation was of an intermediate character and more
abundant; and here grew Sesuvium portulacastrum, a tall Salicornia, and
Suæda fruticosa; whilst the commonest plant was a prostrate Nolanaceous
species with a handsome purplish flower.

Excepting with the fruits of Batis maritima, and perhaps the buoyant
joints of Salicornia, scarcely any of the prevailing shore-plants of the
coast of Peru possess a capacity for dispersal by currents. In this zone
I rarely found any seed-drift on the beaches. Much rubbish, such as
roots of bamboos, however, may be brought down by the rivers; and where
the Humboldt Current strikes a bend in the coast we get a repetition, on
a smaller scale, of the scenes on the Antofagasta beaches. Ancon Bay,
for instance, receives much of the floating offal of the south.

THE MANGROVE ZONE (THE COASTS OF ECUADOR AND COLOMBIA)

We come now to the mangrove zone which comprises, with the remarkable
exception of a long stretch of arid sea-border to the north of the Gulf
of Guayaquil, the whole remaining western sea-border of South America,
namely, the Ecuadorian and Colombian coasts. My own acquaintance with
this region is limited to the estuary of the Guayas or the Guayaquil
River and to the southern shore of the Gulf of Guayaquil; but I am able
to avail myself of the researches of Baron von Eggers, which cover the
entire Ecuadorian coast; and with Ecuador, therefore, I will bring this
brief sketch of the littoral flora of one side of a large continent to a
conclusion.

The Ecuadorian coast, lying, as Baron von Eggers observes, between the
rainless and desert coasts of Peru and the “ewig grüne” coasts of
Colombia, may be regarded as a transition-area presenting very varied
and complicated conditions. With the cause of the remarkable contrasts
exhibited by the strand-flora, not only on the coast of Ecuador, but
along the whole west coast of South America through some forty-five
degrees of latitude from Patagonia to Colombia, I will presently deal.
Here it may be remarked in passing that the Humboldt Current has played
the determining part in producing the abnormal climatic conditions to
which these remarkable contrasts in the strand-flora of this coast of
the continent are mainly due.

The mangrove zone, marking a more or less abrupt transition from a
region of drought and semi-sterility to one of humidity and rank
tropical vegetation, begins about lat. 3° 30ʹ S., that is, in the
vicinity of Tumbez, or perhaps nearer the boundary-line between Ecuador
and Peru in lat. 3° 20ʹ (see Note 72). Occupying the southern shore of
the Gulf of Guayaquil it extends up the Guayas estuary to Guayaquil and
rather beyond. But when we follow the coast of Ecuador northward from
the island of Puna towards Santa Elena Point, we come upon one of the
most remarkable phenomena presented on the west coast of South America.
The dry region begins again and the mangroves disappear; and these
conditions continue through about 2-1/2 degrees of latitude until we
reach the equator, when the mangrove zone soon recommences, and, as I
infer, continues northward without a break to the coast of Central
America.

Dealing first with the mangrove districts of the south side of the Gulf
of Guayaquil and of the Guayas or Guayaquil estuary, we may observe that
probably in few localities of the globe have the forces of nature worked
more in unison to produce the conditions favouring the growth of the
mangrove. The reason why this particular locality has been thus favoured
will be discussed later on in this chapter. I may here observe that
Baron von Eggers was so struck with the exceptional features of the
mangrove-growth in this region that he was inclined to look for the
American centre of the genus Rhizophora, the prevailing mangrove, in the
estuary of the Guayas River.

I will not enter into a detailed description of the mangrove-formation
of this coast, which has indeed been given by the German botanist; but I
will merely refer to the leading features such as they presented
themselves to me. In the first place, reference will be made to the
sea-border of the province of Eloro, where I spent nine or ten days,
making Puerto Bolivar, the port of Machala, my headquarters—a locality
about thirty miles east of Tumbez. Except in the Guayas estuary I have
never seen such a magnificent growth of mangrove.

By following the line of light railway that runs about six kilometres
inland from Puerto Bolivar to Machala, the capital of the province, we
obtain a good section of the mangrove-belt, which may be thus described.
The mangrove-swamp proper extends about three kilometres inland. Whilst
the small variety of Rhizophora mangle (mangle chico) immediately fronts
the sea, Laguncularia grows on the islets close to the seaward margin of
the swamp. When we enter one of the numerous broad creeks that intersect
the border of the mangrove-belt we soon find ourselves in the true
mangrove forest, where prevail tall trees of Rhizophora mangle (mangle
grande) that rise to a height of 70 or 80 feet or more. Gloomy as the
depths of the swamp are, they acquire quite a funereal aspect, the
branches of the trees being draped with pendent Tillandsias. These long,
hair-like, tangled growths hang vertically from the branches of the
trees and may be 20 or 30 feet in length. In the rear of the zone of
tall mangroves we come upon a more open district of the swamp. The
forest proper gives place to a tract occupied by small trees of
Rhizophora, Laguncularia, and Avicennia, with here and there whole acres
occupied only by the shrubby Salicornia peruviana which attains the
height of a man.

[_To face page 484._

[Illustration:

Emery Walker sc.
Rough Plan of the Gulf of Guayaquil.
]

(The main stream of the Humboldt current, as indicated by the arrows,
turns off to the north-west at Cape Blanco; whilst a small branch
crosses the mouth of the Gulf of Guayaquil and flows along the Ecuador
coast north of Santa Elena Point.)

Here terminates the mangrove-swamp proper, and about three kilometres
from the sea it passes gradually into a region of extensive bare
mud-flats which are penetrated by salt-water creeks, two or three yards
across and a foot or two in depth, that are bordered by low and shrubby
Avicennias, the Salicornia bushes above noted, and dwarfed trees of
Rhizophora mangle only four or five feet high. These flats, which are
evidently only overflowed by the sea at the higher spring tides, were at
the time of my visits much sun-cracked and in some parts incrusted with
salt; but the mud was rather soft, and in places Sesuvium portulacastrum
and Batis maritima flourished in quantity on it. These mud-flats, about
two kilometres across, pass by degrees into the low-lying level district
known as the Machala plains, on which the capital of the province is
built. Here the soil is dryish, and, notwithstanding that it displays on
its surface when exposed to the sun a white saline efflorescence, a dry
jungle type of vegetation of the xerophilous character here thrives. I
noticed casually the Algaroba (Prosopis), a yellow-flowered Cordia,
cacti of the Opuntia and Cereus kinds, besides several small trees and
shrubs often thorny.

These Machala plains, on account of the fine saline incrustation above
mentioned, are of much interest, since at a distance of six kilometres
from the coast they thus display on their surface the effect of
sea-water infiltration, their level above the sea being only a few feet.
We have seen the three stages of this infiltration landward of
sea-water: first, the mangrove-swamps daily overflowed by the tide;
second, the mud-flats behind them which are only overflowed by the
fortnightly spring-tides; third, the vegetated plains behind all, which
are sufficiently raised to be above the reach of the tides, but which
are nevertheless soaked with sea-water that displays its presence in the
salt left by evaporation on the surface of the soil.

But another interesting point is here raised. At the back of the
mangrove-belt, in most parts of the world, we usually find a
particularly rank and luxuriant vegetation where the Scitamineæ often
take a leading part; whereas on the sea-border of this part of Ecuador
the mangrove-swamps pass gradually into arid saliniferous plains. With
this singular fact is to be associated the circumstance that we see here
in operation, some four or five miles from the coast, a process by which
great quantities of sea-salts are accumulating below the surface. This
may possibly be concerned with the origin of the great saline deposits
of Northern Chile. However this may be, there is some reason for
believing—and I understand that this is the opinion of Dr. Wolff, the
historio-geographer of Ecuador—that in the course of ages the tendency
will be towards an extension of the dry, sterile regions of Northern
Peru into Ecuador. This subject is referred to again in a later page of
this chapter.

Whilst in this neighbourhood I made the ascent for some fifteen miles of
the Santa Rosa River, which opens into the sea near Puerto Bolivar. It
is a tidal estuary that has no proportion in size to the small river
that enters it. In its lower third we passed at first between long
mangrove-islands formed almost entirely, as viewed from the boat, of the
tall Rhizophora trees draped with Tillandsias, and presenting really a
magnificent spectacle. In the middle third we were penetrating into the
rear of the mangrove-belt. The giant swamp-fern (Chrysodium aureum)
abounded, and here and there we passed by a patch entirely held by the
large shrubs of Salicornia peruviana. The tall Rhizophora trees were
replaced by the short variety, the “mangle chico,” which ceased
altogether about ten miles from the mouth of the estuary, but probably
only about five miles from the nearest part of the coast. The water at
the place where the Rhizophora trees ceased was evidently quite fresh
during nine out of the twelve hours, being only salt in the latter part
of the rising tide. Above the mangroves, in the upper third of the
ascent, Hibiscus tiliaceus, with Chrysodium aureum, flourished on the
banks. The shallows at the margins were occupied by a considerable
variety of semi-aquatic and other plants, such as Pontederia (two
species); one of the Alismaceæ, with the flower and fruit of Sagittaria
and the leaves of Alisma; Typha, Polygonum, and an Amaryllid like
Crinum. Plants of Pistia and Pontederia floated in the stream.

I have said enough to give a general idea of the composition of the
mangrove-belt of the Ecuador littoral, and will refer but briefly to the
mangroves and other river-side plants in the neighbourhood of the city
of Guayaquil, some forty miles up the Guayas estuary. As I have remarked
in Note 38, the water of the river off the city is usually quite fresh
except at high water; but the sea has much freer access to the channels
at the back of Guayaquil, where at high water the density was 1·014. In
these channels are displayed the typical mangrove formation, trees of
Rhizophora mangle bordering the water, whilst behind they are mingled
with Avicennia tomentosa and Laguncularia. On the banks of the main
river, where they are overflowed at high water, Anona paludosa was the
most frequent tree, being associated with the Rhizophora, Hibiscus
tiliaceus, and other trees. Above the city, Polygonum glabrum was
growing in dense masses at the river’s edge, whilst Pontederia and
Pistia flourished on the low muddy banks and floated in quantities in
the river.

Before quitting the subject of the mangrove-formation of Ecuador, I will
refer shortly to the two varieties of Rhizophora mangle that here occur.
Baron von Eggers received the impression that the common type of this
species, a low tree bordering the coast, did not exist in Ecuador, such
a type as he says is characteristic of the West Indies and of Central
America, and, I may add, also of Fiji. The species he regards as
acquiring a new facies in Ecuador, where it exists as tall forest-trees,
branchless for half their height, and exhibiting other divergent
characters. However, I found that the common type of the species occurs
normally on the coast in the vicinity of Puerto Bolivar, thirty miles
east of Tumbez, a district above described.

There are two distinct forms of Rhizophora mangle exhibited in the
mangrove-belt of the coasts around Puerto Bolivar. One of them, which
the indigenes name “mangle chico,” is a small tree, 10 to 15 feet high,
with useless timber, that immediately borders the sea, and, in fact,
largely forms the margin of the swamp, not only on its seaward side, but
also on the land side, where it passes into drier ground. The other, the
“mangle grande,” a tall tree reaching to 60 or 80 and sometimes perhaps
to 100 feet in height, composes the interior, and indeed the bulk, of
the mangrove-belt, and possesses a hard and durable timber much employed
in the district.

Distinct as these two types are, it is not difficult to find
intermediate forms, and, in truth, in some localities they prevail. But
the interesting point is that this peculiar Ecuadorian type of the
species, a type that attracted the attention of the eminent German
botanist, comes near the “Selala,” the mysterious seedless Rhizophora of
the Fijian swamps—a subject fully discussed in Chapter XXX., where I
have compared the Fijian and Ecuadorian Rhizophoras. Both the “Selala”
of Fiji and the “mangle grande” of Ecuador are intermediate between the
American Rhizophora mangle and the Asiatic R. mucronata, resembling the
last in their inflorescence, but in other points approaching the
American species. The “Selala,” however, comes nearer to the Asiatic
tree, whilst the “mangle grande” comes nearer to the American tree.
Unlike the Fijian tree, that of Ecuador is not sterile, but matures its
fruit; and it displays no evidence of the vegetative reproduction so
characteristic of the “Selala.”

Sandy beaches are not common on the mangrove-fronted shores of the south
side of the Gulf of Guayaquil. However, on the seaward side of the long
low mangrove island of Jambeli, on which the lighthouse is placed off
Puerto Bolivar, there is a long stretch of beach of whitish, mainly
non-calcareous, sand. The Coco palms behind the beach give the coast
quite the aspect of a Pacific island strand. Ipomœa pes capræ flourishes
on the sand nearest to the sea; and immediately behind, the beach is
more or less occupied by a Cyperus 2 to 3 feet high, and by Canavalia
obtusifolia. Further back grows a small Acacia tree, and behind it the
yellow-flowered Cordia tree of the district; and in the rear of all lie
extensive mud-flats, partly occupied by stunted bushes of Avicennia
tomentosa and by Sesuvium portulacastrum, which in their turn pass into
the mangrove-swamps.

On account of the enormous amount of drift of all descriptions that is
carried to the sea by the Guayas or Guayaquil River, floating vegetable
materials are abundant in the Gulf of Guayaquil, and are thrown up in
quantity on the coasts of Ecuador. One of the most interesting
spectacles at Guayaquil is presented by the floating river-drift. Huge
tree-trunks and floating islets, the last-named ranging from 3 or 4 to
30 or 40 feet or more across, were, at the time of my visit in February,
being carried to and fro unceasingly in front of the city by the tide,
gradually making their way down the river, and ultimately reaching the
open waters of the gulf. Floating plants of Pistia were in abundance;
and their fate when they reached the sea must have been tragical. The
islets were exceedingly interesting; they were evidently formed of
materials lifted up bodily from the shallows at the margin of the river,
and then carried off in the stream. They were mainly composed of two
species of Pontederia and of Polygonum glabrum in the position of
growth; the first often in flower. Pistia and a variety of smaller
plants nestled among them, such as Salvinia, portions of Azolla, Lemna,
&c.; and in one islet I noticed, oddly enough, the growing rhizome of a
sensitive plant (Mimosa pudica). A great quantity of floating seeds
collect amongst the roots and stems of the plants composing the islets,
and here I obtained much of the smaller seed-drift.

Most frequent in the floating drift of the river at Guayaquil were the
seeds of Anona paludosa, often in a germinating condition. The seeds are
liberated by the decay of the floating fruit, which was also common in
the drift. Amongst the larger materials were the seeds of Entada
scandens and of Mucuna; the empty seeds of the vegetable-ivory palm
(Phytelephas macrocarpa), the sound seed possessing no floating power;
the “stones” of Spondias lutea, L., as identified by Mr. Holland, of the
Kew Museum; the empty small nuts of several palms, including,
apparently, Oreodoxa, &c. Amongst miscellaneous materials were small
gourds, which are referred to in Note 47, and an occasional empty cacao
fruit. Smaller seeds were also abundant, and included those of Hibiscus
tiliaceus, Erythrina, Vigna, Ipomœa, and others. Carried into the river
from the neighbouring mangrove-creeks, where they abound, there were
floating seedlings of Rhizophora and Avicennia, fruits of Laguncularia
often germinating, and the seeded joints of Salicornia peruviana.

There was of course, in addition, much that was strange in the floating
drift of the Guayas River, which received its contributions not only
from the river-side vegetation and the neighbouring mangrove-swamps, but
also from the interior mountain ranges culminating in Chimborazo, the
slopes of which are drained by its tributaries. I had several
opportunities of meeting the drift of the Guayas River in the open
waters of the Gulf of Guayaquil. Much of it is carried along the south
side of the gulf; and I picked up at sea, ten to twenty miles from the
mouth of the estuary, many of the things above enumerated, such as
Erythrina and Mucuna seeds, seeds of Hibiscus tiliaceus, the empty
vegetable-ivory seeds, the seedlings of Rhizophora and Avicennia, and
the germinating fruits of Laguncularia and Salicornia peruviana. Much of
these materials mingled with local drift is stranded on the long beach
of Jambeli Island, thirty miles from the mouth of the estuary. Here,
besides the seeds of Canavalia obtusifolia and Ipomœa pes capræ derived
from the locality, I found the seedlings of Rhizophora and Avicennia,
and the fruits of Laguncularia and Salicornia peruviana, that might have
been in part derived from the adjacent swamps, as well as much of the
drift of the Guayas River, such as the seeds of Anona paludosa, Entada
scandens, Erythrina, and Mucuna, the small gourds, the same small
palm-nuts, the empty seeds of Phytelephas, the “stones” of Spondias
lutea, and much other material previously familiar to me, but nowhere a
sign of the floating Pistias and of the flowering Pontederia islets of
that estuary.

_The Stretch of Dry Coast from the Vicinity of Puna Island to the
Equator._— This remarkable piece of sea-border, covering nearly three
degrees of latitude, and in its aridity and general character recalling
the sterile sea-coast of Peru, is placed between the humid
mangrove-fronted coast of the Guayas estuary and the similarly humid and
mangrove-fronted coasts of Northern Ecuador and Colombia. The mangrove
seems to be almost absent from this stretch of dry coast. Mr. F. P.
Walker, of the Santa Elena Cable Station, tells me that some time ago a
little mangrove-growth existed near the Point, but that it has
disappeared; and Baron von Eggers implies the absence of mangroves from
the whole coast. The first-named speaks of the dry character of the
coast district from Santa Elena Point to within half a degree of the
equator; and the last-named, in his description of the coast, mentions
cacti and thorny plants as typical of the vegetation. Since this region
represents a typical locality where the direct influence of the Humboldt
current on the climate of almost the whole west coast of South America
can be put to the proof, I will refer to its peculiar climatic
conditions below in my discussion of the general question, and will here
content myself with saying that on this dry portion of the coast of
Ecuador we have reproduced, but in a less pronounced degree, the
climatic conditions of the coast of Peru.

_The Humboldt or Peruvian Current and the Climate of the West Coast of
South America._—The question we will now briefly consider is one that is
concerned with the determining causes of the singular distribution of
coast-plants on the west coast of South America. The reader will have
already seen that the matter is an affair of climate; but it is an
affair of climate in which (although it affects forty or more degrees of
latitude), latitude, in a general sense, scarcely counts. All the
naturalists, from Humboldt onward, who have sojourned in this region of
the globe have displayed a deep interest in this subject; and I suppose
there can be no region of the globe where there are so many climatic
anomalies as interesting to the meteorologist. Here, for instance, might
be obtained materials for solving the irritating mystery of a London
fog; and if the suggestion of Baron von Eggers, before alluded to, is
carried out, and a station is established by the Meteorological
Societies of Europe and America at some suitable locality like Santa
Elena on the coast of Ecuador, we might obtain, among other results,
another line of investigating the causes of the fogs of our metropolis,
a subject about which Captain Carpenter has recently made an important
preliminary inquiry.

I will assume that my readers are already acquainted with the nature of
the problem to be discussed relating to the climate of the west coast of
South America, and that they are familiar with the view generally held
that the aridity of this extensive coast region, stretching from the
thirtieth parallel of south latitude to the equator, arises from the
loss by the trade-winds of all their moisture in the interior of the
continent before reaching the western countries of Chile and Peru. Mr.
Ball, in his book on South America, opposed this view, though from
reasons only partially valid, since he instanced the Ecuador coast as
being, contrary to the theory, a wet coast, whereas we know that a large
stretch of it is arid and not unlike Peru. The parting of the ways in
this discussion lies in the answer to the query, Why should the
south-east trade carry so much moisture to the east side of South
America, whilst the south-west winds, that are equally prevalent on the
west coast of the continent, are drying winds which convert the
sea-border into a desert, as in Northern Chile, or into a region of
semi-sterility, as in the instance of Peru? Other things being equal, we
should expect both sea-borders of the continent in these latitudes to be
well watered. In the answer to the question why the south-east trade
should be a wet wind and the south-west wind a dry one lies a fatal
objection to the prevailing view.

When Professor Davis, in his article on North America in the
_Encyclopædia Britannica_ (vol. 25), observes in connection with the
arid coast regions on the west side of the continent that the southerly
flow of the winds along the Pacific coast gives them a drying quality,
thus causing the extension to the coast in South California and in North
Mexico of the arid regions of the interior, he seems to imply that these
winds acquire their drying capacity in flowing from cooler to warmer
latitudes. On this view all trade-winds should be drying winds, whereas
the reverse would appear to be the case.

There is some condition, present on one coast of the South American
continent and absent on the other, which determines why a southerly
wind, blowing landward, is in the one case moist and in the other dry.
According to my own view the winds of the arid coast regions of western
North America cross the cool waters of the Californian current, and thus
acquire their drying quality on striking a sea-border more highly heated
than the winds themselves. On the tropical west coast of South America
the winds also become drying winds by passing over the cold waters of
the Peruvian or Humboldt current, where mists are in consequence of
frequent occurrence; and on striking the more highly heated land-surface
at the sea-border the moving air does not part with any more moisture
until an altitude of some thousands of feet above the sea is reached,
when the cloud-belt forms. On the mountains bordering the coast of the
Antofagasta province, in January, the clouds gathered at an elevation of
4,000 to 5,000 feet. Perhaps the best way to contrast the east and west
coasts of tropical South America in this respect would be to say that
whilst the wind blows landward in both regions, the land is the
condenser on the east side, and the sea, owing to the interposition of
the cold Humboldt current, is the condenser on the west side.

During a fortnight spent in February at Ancon, about twenty miles north
of Callao, I noticed that with the prevailing cool south-westerly wind
the coast was clear, but it was misty at sea. On the few days when there
were warm westerly and north-westerly breezes, the weather was thick at
sea; and if this condition was pronounced, the whole coast was enveloped
in mist; but more usually the coast-line was fairly clear except at the
promontories, along the sides of which clouds blown in from the sea
rolled in lines inland, not generally attaining an elevation over 300 or
400 feet, but sometimes reaching 900 or 1,000 feet, and gradually
disappearing a mile or two inside the coast-line. These sea-born clouds
thus vanished as they traversed the more highly heated land-surface; and
the air-current continuing its inland course mounted the slopes of the
adjacent mountain ranges of the Andes, some three or four miles from the
coast, until at an altitude of some 5,000 or 6,000 feet condensation
again occurred and the cloud-belt was formed at those cooler levels.

From the summit of a range rising to a height of about 2,500 feet to the
north-west of Lima I had presented to me a splendid spectacle, on
February 12th, in the formation of the coast-belt of clouds. The
forenoon was clear, but about 2 p.m. the sea-born clouds began to roll
inland, concealing the lower two-thirds of the island of San Lorenzo,
which has an elevation of almost 1,400 feet, and completely covering up
Callao and the low country bordering the sea, but extending only a mile
or two from the coast-line. The dense cloud that covered Callao
appeared, as I looked down upon it from my mountain-peak, like a billowy
field of snow sparkling in the sun, with the summit of San Lorenzo
standing out like some bare alpine summit from amidst the snows. Yet
beneath that dazzling covering Callao lay all in gloom; whilst only six
miles up the broad valley of the Rimac the city of Lima stood in a blaze
of sunlight, its domes and towers reflecting back the light as I looked
at the strange contrast it presented with the buried city of the coast.
The mystery of a London fog seemed to lie unfolded at my feet, ready for
the man who can read the signs aright.

That the mere presence of a cold current on a coast with the winds
blowing off the land (as in the case of the Labrador current, which
extends down the Atlantic coast of North America to Cape Hatteras and
beyond) produces no sterilising effect on the vegetation of the
sea-border of a continent is well brought out in the beautifully
executed maps in Prof. Russell’s recent work on North America. The
essential condition for producing sterility on the sea-border of a
continent is not only that the waters of a cold current should wash its
coasts, but that the regular winds should blow landward across its cool
surface. These are what we find on the west coast of South America.

Not with the hope of adding anything new to our knowledge of the
climatology of this region, but with the purpose of becoming personally
acquainted with the problem involved, I paid considerable attention to
this subject during the three months passed on the west coast of South
America between Port Valdivia and Guayaquil. It was not until I had
dropped my thermometer into the cool water of the Humboldt current and
had watched the formation of the fogs on the sterile coast of Peru that
the real nature of the problem presented itself. From the pages of a
work like Tschudi’s _Travels in Peru_ one acquires an excellent idea of
the extraordinary climatic conditions of this region, and the same may
be said of the narratives of Darwin and other travellers; but it is
necessary to be brought into personal contact with these conditions
before one can appreciate their significance.

As is well known, says Baron von Eggers, the Humboldt current explains
the anomalous climate of the coast of Peru, and one may add of North
Chile and Ecuador. The current, which represents the extension
northwards of the west wind-drift of the Roaring Forties (see Dickson in
_Encycl. Brit._, xxxi. 404; and Admiralty Current Charts of the
Pacific), begins on the coast between the 33rd and 40th parallels of
south latitude, according to the season. North of Valdivia, as we
approach Valparaiso, in lat. 33°S., the effect of its presence is at
once seen in the increasing dryness of the climate and in the alteration
in the character of the vegetation. It has, however, been shown that the
current needs the co-operation of the prevailing southerly and westerly
winds as they blow landward over its cool waters. On the coast of Peru
these moist winds often generate fog and mist as they cross the current.
They reach the coast as drying winds, having a temperature much cooler
than the lower coast regions; and the air-currents do not precipitate
any moisture on the land until an elevation of 4,000 to 6,000 feet is
attained where the cloud-belt is formed.

In order to establish this theory it is, however, necessary to show that
when the Humboldt current leaves the coast normal conditions of humidity
occur, to which the vegetation responds, and that when the current
strikes the coast again the conditions of aridity reappear. In its
course northward the current divides off Cape Blanco, the principal mass
of its waters making towards the Galapagos Group, whilst the remainder,
after crossing the Gulf of Guayaquil, flow along the coast of Ecuador
between Santa Elena and the equator. Now, it is along this stretch of
the Ecuador coast that the conditions of aridity reappear and that the
climate of the Peruvian sea-border is in a modified form reproduced. In
the interior of the Gulf of Guayaquil, on the other hand, where the
sea-border is no longer subjected to the influence of the cold waters of
the Humboldt current, the genius of the tropics, repressed through so
many degrees of latitude, bursts its bonds, and presents us with a
spectacle of littoral vegetation that, so far as mangrove-growth is
concerned, is probably unrivalled on our globe.

This contrast is well shown in the mean annual temperatures on the
opposite sides of the Gulf of Guayaquil. Baron von Eggers, quoting Dr.
Wolff, states that whilst the mean for the year at Puna is about 75° F.,
and at Santa Elena about 73°, on the south side of the gulf at Balao it
is several degrees warmer and is evidently not under 80°. The mean
temperature for the second week of March during my sojourn at Puerto
Bolivar, which is near the beginning of the mangrove region on the south
side of the gulf, was 79°, the mean daily range being 74° to 83·5°. This
stretch of dry coast reaching north from Puna to the equator is
evidently regarded by Baron von Eggers and others who have studied the
climatology of Ecuador as the critical area required to confirm the
theory connecting the aridity of the west coast of South America with
the Humboldt current. Here the sea for the greater part of the year has
a temperature (according to the British Admiralty chart of
surface-temperatures) of 70° to 75°; the mean temperature of the air is
73° to 75°; the rainy season, instead of covering a period of six months
and over, as in the humid regions north and south of this coast, has a
duration of only two or three months; the prevailing wind is south-west;
whilst the direct influence of the cool waters of the current is shown
in the general cloudiness that prevails during the last half of the year
and in the drizzling mists that are frequent from June to October.
Reference has already been made to the manner in which the vegetation on
this dry coast of Ecuador responds to the arid conditions, as, for
instance, in the absence of mangroves and in the prevailing character of
the plants of the sea-border, cacti, thorny plants, and such like. For
my information on this exceedingly interesting tract of coast, which is
the test-ground of the Humboldt current theory, I am indebted to the
papers of Baron von Eggers (see end of this volume) and to Mr. F. P.
Walker, of the Central and South American Telegraph Company’s Station at
Santa Elena, who very kindly communicated with me by letter. Some
additional remarks are given in Note 73, and my own observations on the
temperature of the Humboldt current from Antofagasta northward are
summarised in Note 74.

Before quitting the Ecuador coast a word may be said relating to the
prediction of Villavicencio that the climate of this sea-border will
assimilate itself to that of the rainless coasts of Peru. This is, I
believe, also the opinion expressed by Dr. Wolff in his _Geografia y
Geologia del Ecuador_ (Leipzig, 1892); and it is referred to by Mr.
Webster in his article on Ecuador in the seventh volume of the
_Encyclopædia Britannica_. There is a prevailing impression amongst the
more observing residents that I met in the Ecuadorian province of Eloro,
on the Peruvian border, that the country is drying up. A few pages back
I have described how in the Machala district of this province the
mangrove-belt passes landward into an arid region suggestive of the
sea-border of Peru. This transition is startling to one who expects to
find behind the mangrove-belt, as he would find in most parts of the
world, a humid region where Nature revels in the rank luxuriance of
plant-growth. This is, however, not always the case, since on the lee or
dry sides of the large islands of Fiji the mangrove-belt is backed by
extensive arid plains, for an explanation of which, as I have shown in
Note 22, we have to appeal rather to the hygrometer than to the
rain-gauge. This is true also of Ecuador; but whilst the reason is
intelligible enough in Fiji, it only carries us a step farther back in
the case of the Machala plains in Ecuador. These plains are continuous
with similar districts across the Peruvian border where they reach the
coast; and if the reader will refer again to my description of the
section of the mangrove-belt and the plains in its rear from Puerto
Bolivar to Machala, he will incline to the view that the desiccation of
the sea-border of Ecuador is now in progress.

Evidence of a more direct nature could doubtless be supplied by those
who have long resided on the coast of Ecuador, and in illustration I
will give an extract from one of Mr. Walker’s letters dated May, 1904,
from Santa Elena.—“The rainfall here might for the last ten years be put
down at two showers per year. It is said that the last good rainy season
was in 1891. The inhabitants say that formerly it always rained enough
to make the grass grow every year, but during the eleven years I have
been here there appears to be a marked falling off of the rainfall.”

It has been only possible to touch the fringe of this interesting
question here; but from the standpoint of the study of the littoral
flora of the west coast of South America it is of some importance.
Immediately behind the epoch of the present marine molluscan fauna of
this coast there lies an age when, as we learn from Philippi, the shells
of Chile were more akin to those of the Atlantic and Mediterranean
faunas than to those now found on the Chilian coasts. The transition is
a sudden one; and amongst other explanations of this strange
transformation Suess suggests the sealing up of a communication through
the Panama isthmus by volcanic eruptions and the appearance of the
Humboldt current (_Das Antlitz der Erde_, French edit. by Margerie, ii.
825). May it not be, my readers may ask, that the west coast of South
America is still in the age of progressive sterility; and that before
this age began Peru possessed a normal tropical strand-flora? It has
been remarked in Chapter VIII. that the same species of mangroves occur
on both the Atlantic and Pacific coasts of America, and that at all
events their present distribution belongs to an age when the Gulf of
Mexico was in communication with the Pacific Ocean. May we not, again,
suppose that in that age the mangroves extended far south on the coast
of Peru, just as they do now on the coast of Brazil?

Coral reefs are stated not to exist in tropical latitudes on the west
coast of South America in our own day; but we might almost expect that
at the close of the Tertiary period, and perhaps before the appearance
of the Humboldt current, they existed with the mangroves on the coast of
Peru. As bearing on the subject of a change of climate on that coast in
times geologically not remote, I may allude to the circumstance, which
is discussed more in detail in Note 75, that I found, sometimes in fair
quantity, blocks of massive coral, long since dead, much pierced by
boring shells, and in places undergoing a chemical change, at Arica
(lat. 18° 25ʹ S.), at Callao (12° 3ʹ S.), and at Ancon (11° 45ʹ S.) on
the coast of Peru.

These masses, which varied from a few inches to two or three feet in
size, gave me the impression of having been torn off the bottom, in some
cases in recent times, in others perhaps centuries ago, by the huge
sea-waves that from time to time overwhelm this coast. At Ancon, where
they were sufficiently abundant to be used for bordering the flower beds
in the hotel garden, they were most numerous in the vicinity of a rocky
spur of andesite that protruded from the beach between the tide levels
and was more or less covered at high water. A few paces inland from the
beach some of these coral masses, evidently stranded long ago, were
undergoing that queer process of disintegration which everything
calcareous seems to undergo on the beaches and plains of this almost
rainless coast. Like the bones of the Incas lying bleaching on the
neighbouring plains, like the sea-shells and bones of bird and beast
cast up long ago on the beach, they were falling to powder where they
lay, and the coral fragment lay often in the midst of its own _débris_.
The blocks on the beach proper were for the most part still hard and
compact, and the same may be said of those observed on the beaches of
Callao and Arica.

The corals were quite different from those with which I was familiar in
the reefs of the Pacific islands, and, bearing in mind the known
distribution of coral reefs, I was a little dubious about them.
Accordingly I sent some specimens to the British Museum, and Mr. Jeffrey
Bell has kindly informed me that they seem to be decayed and much
injured perforated examples of Porites. When powdered they effervesce in
an acid, but the bulk of the material remains undissolved.

No more eloquent testimony could be afforded of the rainless climate
than these corals crumbling on the Ancon plains when washed a few paces
inland from the beach. They could be noticed in all stages of
disintegration from the block surrounded by a little line of
disintegrated material, representing the initial products of its own
decay, to the crumbling mass, almost friable in the fingers, that was
lying in the midst of its own dust and loose polyp-tubes, and finally to
the little mound of _débris_ that alone remained. Mr. Darwin, in his
_Journal of Researches_ (chap. xvi.), refers to a similar process of
decay in the elevated shell-beds of San Lorenzo, off the coast of
Callao. On the higher terraces a layer of saline powder, consisting of
sulphates and muriates of lime and soda but with very little carbonate
of lime, was the sole indication of the shell-beds. Dry climatic
conditions at the sea-border evidently favour, as he observes, the early
decay of exposed calcareous remains.

_The Shore-plants and Stranded Seed-drift of the Panama Isthmus._

I spent two days at Panama and two days at Colon in examining the
neighbouring beaches and estuaries of the Pacific and Atlantic coasts of
the isthmus. On the Panama side the mangrove-belt was formed on the
seaward border of “mangle chico” (the small prevailing type of
Rhizophora mangle), Laguncularia, and Avicennia; whilst behind it passed
into extensive swampy tracts occupied by the Swamp Fern (Chrysodium
aureum), Hibiscus tiliaceus, and other plants. On the Colon or Atlantic
side the mangrove-belt had precisely the same composition and presented
the same species, Rhizophora and Avicennia usually forming the outposts
on the reef-flat, whilst Laguncularia was abundant in the rear. In the
estuary of the Rio Chagres, Rhizophora and Laguncularia were abundant
near the mouth, and Chrysodium aureum and Hibiscus tiliaceus by the
waterside higher up. Dr. Seemann, in his volume on the botany of the
voyage of H.M.S. _Herald_, observes that the species of Laguncularia
common on both the Atlantic and Pacific coasts of the Panama isthmus is
L. racemosa. This species differs in the form of its fruit from the
Ecuador tree. Laguncularia racemosa, Rhizophora mangle, and I may add
Anona paludosa and Conocarpus erecta, are all plants of the
mangrove-formation that occur not only on the Pacific and Atlantic
coasts of America but also on the west coast of Africa. It is likely, I
may add, that the “mangle grande,” the Ecuadorian type of Rhizophora
mangle, exists in the Panama isthmus, since in the higher part of the
estuary of the Chagres I found trees approaching it in characters.

Amongst the plants growing on the Panama beaches I noticed Canavalia
obtusifolia, Hibiscus tiliaceus, and Ipomœa pes capræ, all of which
occur also on the Atlantic side of the isthmus. The Manchineel
(Hippomane mancinella), found also on the Atlantic side of the
continent, grows on the Panama beaches. Its fruits, which look like
crab-apples, lose their outer fleshy covering when drying on the sand.
Not being familiar with this poisonous tree, I allowed some of the milky
sap of the fruits to touch the skin, and suffered great pain for five or
six hours. The fruit possesses an inner coat of air-bearing cork-like
tissue; and the stone, if I may so term it, thus acquires great floating
power. I kept some afloat in sea-water for five weeks, and no doubt they
will float for months.

The seed-drift to be observed stranded on the beaches and floating in
the estuaries on both sides of the isthmus is, generally speaking, the
same—a circumstance of great importance in plant-distribution, since we
can here see rivers bringing down the same seeds from the same “divide”
to the shores of the Pacific and Atlantic oceans. In the case of a plant
like Entada scandens, which grows in the interior, this is a matter of
much interest, as it thus possesses here a centre of dispersal from
which its seeds can be carried by the currents eastward to the West
African coast and westward across the Pacific to Malaya and (given time)
around the shores of the Indian Ocean to the East African coast. In
describing the possible routes of dispersion from this centre I have
described the distribution of the species.

I am indebted to Mr. Holland, of the Kew Museum, for the identification
of some of the drift-seeds and fruits collected by me on the isthmus,
those identified by him being followed by the letter H. On the beaches
and floating in the estuaries on both sides of the isthmus I found
Rhizophora seedlings; seeds of Entada scandens and Mucuna urens
(medic.), H.; seedvessels of Spondias lutea (Linn.), H.; Prioria
copaifera (Griseb.), H., with decayed seed; and the empty nuts, 1-1/2 to
2 inches in size, of more than one species of Astrocaryum, H. Although
in the case of the two last-named genera the seedvessels were useless
for dispersal, being evidently brought down from the interior by the
rivers, they serve to illustrate the important principle that the rivers
bring down the same seed-drift on both the Atlantic and Pacific coasts
of Central America. Mr. Hemsley includes amongst the seed-drift stranded
on the coast of Jamaica the seedvessels of Spondias (probably S. lutea)
and of Astrocaryum (_Bot. Chall. Exped._, iv. 299, 304).

Those of Spondias lutea were found by me floating in the Guayaquil River
and stranded on the beaches of Ecuador and of the Pacific and Atlantic
coasts of the Panama isthmus. This is the Hog-plum, which in tropical
America and the West Indies is both wild and cultivated. Its buoyant
“stone” has a covering of cork-like air-bearing tissue. This is a
remarkable case of non-adaptation in the matter of buoyancy. The
seedvessels cut across contained sound seeds; and they are provided with
the essential qualities of “long floaters.”

_Summary._

(1) The strand-district of the west coast of South America is divided
into four zones:—

(_a_) The Convolvulus soldanella zone of Southern Chile.

(_b_) The Desert or Plantless zone of Northern Chile.

(_c_) The Sesuvium zone of Peru.

(_d_) The Mangrove zone of Ecuador and Colombia.

(2) The mangroves do not extend south of Ecuador or, more strictly,
south of Tumbez (3° 30ʹ S.).

(3) The absence of mangroves on the tropical coasts of Chile and Peru is
attributed to the Humboldt current, which has so influenced the climate
that it has converted the sea-border of North Chile into a desert and
that of Peru into a region of semi-sterility.

(4) It is considered that this has been effected through the prevailing
winds acquiring drying qualities on crossing the cold waters of the
current in tropical latitudes.

(5) To establish this it is shown that when the Humboldt current leaves
the coast at Cape Blanco mangroves thrive in the Gulf of Guayaquil, and
that when it strikes the coast again near Santa Elena Point and courses
along that seaboard to the equator we find the Peruvian conditions of
semi-sterility reproduced.

(6) The probability that the arid climate of Peru is in our own time
extending northward into Ecuador is pointed out; and from the presence
of old coral blocks on the Peruvian beaches it is considered likely that
when these corals throve the mangroves extended far down the coast of
Peru.

(7) It is shown from the presence of the same species of mangroves on
the Pacific and Atlantic coasts of America that there must have been,
not long ago, a communication between these two oceans across Central
America; but it is at the same time observed that this could not be
inferred from shore-plants with buoyant seeds that, like Entada
scandens, occur inland, since, although they occur on both sides of the
continent, their distribution can be explained by the transport of their
seeds by rivers to the Atlantic and Pacific coasts, such as we see in
operation on the Panama isthmus in our own day.

(8) Stress is laid on the great development of mangroves in the Gulf of
Guayaquil and in the Guayas estuary; and it is pointed out that there
are in this locality two varieties of Rhizophora mangle, a large and a
small variety, the first approaching in some of its characters the
Asiatic species, R. mucronata, and being akin also to the seedless
intermediate form of Fiji.

(9) Amongst other matters dealt with in this chapter are the floating
seed-drift of the Guayas or Guayaquil River and the shore plants and
stranded seed-drift of the Panama isthmus. From the locality last named
we learn that rivers bring down from the interior to the Atlantic and
Pacific coasts much the same seed-drift, and that from this centre
littoral plants with buoyant seeds can be distributed over the whole
tropical zone.

CHAPTER XXXIII

SEED-DISPERSAL AND GEOLOGICAL TIME

The shifting of the source of the Polynesian plants from the New to the
Old World.—The floral history of Polynesia stated in terms of
geological time.—The suspension of the agencies of dispersal in later
periods.—Parallel differentiation in the course of ages of climate,
bird, and plant.—New Zealand.—Insects and bats as agents in
plant-dispersal.—The effective agency of sea-birds in other
regions.—The observations of Ekstam.—The Spitzbergen controversy.—The
efficacy of ducks as distributors of aquatic plants.—Summary.

IN the matter of the dispersal of seeds by birds in the tropical
Pacific, there are at least two questions which my readers must have
frequently put to themselves. The one would be concerned with the
shifting of the source of the Polynesian plants from America to the Old
World, which occurred probably near the close of the Tertiary period.
The other would be connected with the suspension of the work of
dispersal over a large portion of Polynesia, which has become more and
more pronounced as we approach our own day.

_Suggested Cause of the Shifting of the Source of the Polynesian Plants
from the New to the Old World._—In previous chapters we have discussed
the various epochs in the plant-stocking of these islands. There was
first the age of Coniferæ, in which the islands of the Western Pacific
were only concerned, an age prior to the appearance of the volcanic
groups of the Tahitian and Hawaiian regions, and placed in the Secondary
period. Then followed, in the Tertiary period, after the birth of Hawaii
and Tahiti, and when the island-groups of the Western Pacific were
mainly submerged, the general dispersion from America of the Compositæ,
Lobeliaceæ, and other orders, now represented only by genera peculiar to
the Hawaiian and Tahitian islands. Last of all, towards the close of the
Tertiary period, when the island-groups of the Western Pacific had
re-emerged, a general dispersion of Old World plants, mainly Malayan,
took place over all the present archipelagoes of the tropical Pacific.

Since the climate of Hawaii must have, to a great extent, shared in the
vicissitudes of the continental climates of the northern hemisphere
before, during, and after the Glacial epoch, it is assumed that in the
Ice Age no tropical plants reached the group, and that only the plants
now represented in its mountain-flora could have then reached there. The
area of active dispersion of tropical plants was pushed far south.
During the Ice period, Indo-Malayan plants doubtless crowded into the
equatorial region of the Western Pacific; but, cramped and confined
within this limited area, they were cut off by a climatic barrier from
the cool latitudes of Hawaii. As the cold ages passed away, migratory
birds, confined during that period to the southern hemisphere, would
extend their ranges north, sometimes reaching Hawaii, and transporting
to it the seeds of New Zealand and Antarctic genera, now represented by
endemic species on its mountain-slopes. The Indo-Malayan plants, with
the increasing warmth in the climate of the northern hemisphere, would
overrun the Pacific, set free from their prison in the south-west
portion of that ocean. Dispersal, we might imagine, would be at first
very active over the whole ocean.

My point is, then, that whilst the Malayan era of the plant-stocking
began after the Ice Age in the northern hemisphere, the dispersion of
the New Zealand and Antarctic genera over the Pacific took place during
that period; whilst, as before noticed, the dispersion of the Compositæ,
Lobeliaceæ, and other orders, represented now in Hawaii by endemic
genera, would be pre-Glacial and well back in the Tertiary epoch. I
would, therefore, suggest the following scheme, in illustration of the
floral history of the tropical Pacific.

(1) The Age of Conifers of the Western Pacific during the Mesozoic
period, and before the appearance of the Hawaiian and Tahitian
archipelagoes.

(2) The Age of Compositæ and Lobeliaceæ, and of other genera. This is an
era of American plants, and it is referred to the Tertiary period. In it
only the newly-formed Hawaiian and Tahitian groups shared, the islands
of the Western Pacific being largely submerged.

(3) The Age of Malayan plants, regarded as mainly post-Glacial, and
subsequent, therefore, to the re-emergence of the Western Pacific
islands at the close of the Tertiary period.

Dispersion then was general over the Pacific. The distribution of the
New Zealand and Antarctic genera, plants that take a subordinate part in
the floras of the Pacific islands, is regarded as having occurred during
the glaciation of the northern hemisphere.

_On the Suspension of the Agencies of Dispersal in the Tropical
Pacific._—If the remark of Drake del Castillo that genera possessing
only non-endemic species in the Pacific islands owe their presence in
this region to existing agencies of dispersal looks something like a
truism, we must remember that, assuming Nature to be uniform in her
methods, it involves not merely the original co-operation of the same
agencies with genera that own only peculiar species, but also the
subsequent suspension of the work of these agencies.

The nature of the connection between freedom of dispersal and specific
differentiation is well brought out by Beccari in contrasting the
species of Ficus and the palms of Borneo; whilst out of fifty-five
species of Ficus collected by him in that island, 30 per cent. were
apparently peculiar, 85 per cent. of the 130 Borneo palms had not been
found elsewhere. In the English edition of his _Nelle Foreste di Borneo_
he says that “the explanation lies in the fact of the facile
dissemination of the various species of Ficus through the agency of
birds, an explanation which applies to all trees which produce edible
fruits specially relished by animals.” He shows, also, that the same
principle applies within the limits of the genus Ficus, since of those
Bornean species known to him as belonging to the section Urostigma,
which possesses fruits most preferred by birds (pigeons, hornbills,
&c.), nearly all (fourteen out of sixteen) are found elsewhere; whilst
of ten species belonging to the section Covellia, where the fruits are
more or less hidden and inconspicuous, and with difficulty discovered by
birds that would effectively distribute the species, four, at the most,
are found elsewhere. “Such facts,” he goes on to say, “show that in
tropical countries the various kinds of Ficus are, to a large extent,
biologically connected with birds, which, perhaps, on their part, also
owe some of their peculiarities in the shape of the bill or in the
plumage to the nature and coloration of the fruits which form their
food.”

Whilst Dr. Beccari as a botanist lays especial stress on the biological
connection in Malaya between the plant, as illustrated by the genus
Ficus, and the bird, Mr. Perkins, as a zoologist, is similarly emphatic
on the biological connection in Hawaii between the bird, as illustrated
by the peculiar family of the Drepanididæ, and the plant. The plants
here are the arborescent Lobeliaceæ and the Freycinetias. To the flowers
of the arborescent Lobeliaceæ the nectar-feeding Drepanids are
particularly partial; and the development of the extreme forms of these
birds, as Mr. Perkins observes in the _Fauna Hawaiiensis_, “is not
comprehensible without a knowledge of the island flora.” Not only does
he point to the modifications in the form of the bill of the bird in
connection with the tubular form of the flowers; but in at least one
species of these arborescent Lobeliaceæ he shows that it is dependent on
the Drepanid for its fertilisation, and he inclines to the view that
changes such as that of lengthening of the bill may have taken place
side by side with the increasing length of the tubular flowers. In
connection with the Freycinetias of Hawaii, Mr. Perkins regards the bill
of the Ou, a finch-like Drepanid of the genus Psittacirostra, as
“entirely formed and adapted for the purpose of picking out the
component parts” of the fruiting inflorescence.

That in an isolated island-group birds and plants often “differentiate”
together is a fact well known in distribution. In Hawaii, for instance,
as I learn from Mr. Perkins, quite 45 per cent. of the birds are
peculiar; whilst according to Dr. Hillebrand 80 per cent. of the
flowering plants are confined to the group. Then, again, in the
Galapagos Islands, half of the plants and two-thirds of the birds are
confined to that archipelago. At the other end of the series we have the
Azores, with about a tenth of its plants peculiar, and about 4 per cent.
of its birds peculiar to the islands, and Iceland with no endemic plants
and, as far as I can gather, few peculiar birds.

Accepting Mr. Charles Dixon’s view (_The Migration of Birds_, 1897) that
specific differentiation does not occur along lines of migration, we
must assume that the differentiation of the avifauna of an isolated
group like Hawaii began with the breaking off of its regular
communication through birds with the outside world. I do not consider
that in the past these Pacific archipelagoes received their birds in any
haphazard fashion, as, for instance, in the guise of stragglers that had
lost their way. From the circumstance pointed out to me by Mr. Perkins
that 25 of the 67 genera of Hawaiian birds are peculiar, we must
postulate a high antiquity for the bird fauna dating far back into the
Tertiary period. Mr. Perkins, who kindly supplied me with his general
views of the nature of the Hawaiian fauna, tells me that it is
“positively oceanic-insular and could be continental only on the
supposition that everything continental had been at some time destroyed
and that the group had been subsequently re-stocked as would any oceanic
island.”

The view naturally presents itself that in past ages birds in the
Pacific were much more uniform in their characters, and the agencies of
dispersal far more active in their operations and far more general in
their range than in more recent times, “It may be accepted (says Mr.
Dixon) as an axiom of geographical distribution that all existing
species are surviving relics of more ancient forms or ancestral types,
whose dispersal in a remoter past was more continuous, and whose
affinities and characteristics were therefore more homogeneous.” I
assume that in past ages the differentiation of birds has largely been
favoured by differentiation of climate acting through the limitation of
their ranges. To these changes, plants, so often biologically connected
with birds, have largely responded.

There is, of course, no difficulty in imputing to birds the capacity of
reaching Hawaii in the mid-Pacific, and there are many regular migrants
now (sea-birds, waders, ducks, &c.). The only difficulty is in the
estimation of the time occupied in the trans-oceanic journey. According
to Gätke the journey, which is 1,500 to 2,000 miles, ought to be
accomplished within the limit of fifteen hours, which he regards as “the
longest spell during which a bird is able to remain on the wing without
taking sustenance of any kind.” As he considers that a bird might cover
the 1,600 miles between Newfoundland and Ireland in nine hours
(_Heligoland as an Ornithological Observatory_, p. 140), the Hawaiian
traverse would offer to him no difficulties. It has frequently occurred
to me in this connection that in ancient times, when the volcanoes of
the mid-Pacific were in full activity, their light at night-time would
have often given a direction to the migrating bird, and that they might
have sometimes determined the line of migration across the Pacific.

It has not been possible to discuss here the capacity of pigeons to
cross an ocean, a subject bearing directly on the floras of all the
Pacific groups (excepting Hawaii, which possesses no indigenous Columbæ)
and as concerning these islands generally presenting no difficulty. Dr.
H. de Varigny, who amongst his other studies has long displayed an
active interest in plant-dispersal, has directed my attention to two
very important papers on the flight of pigeons in the _Revue
Scientifique_, one by M. A. Thauziès (April 30, 1898) and the other by
M. M. Dusolier (Nov. 28, 1903). That land birds, as well as swimmers and
waders, cross the Atlantic is well known, and in this connection the
reader might profitably consult Prof. Heilprin’s _Geographical
Distribution of Animals_ (vol. 58, _Internat. Sci. Ser._ 1887).

Much might be said of these matters, but it would be out of place here;
and I will content myself with stating the view above indicated that the
suspension of the agencies of seed-dispersal over the Pacific is
probably connected with a general principle affecting the whole
plant-world. The tendency in the course of ages has been towards the
differentiation of climate, bird, and plant, the range of the bird being
largely controlled by the climate, and the range of the plant being
mainly dependent on the range of the bird.

It is evident that in some cases the plants themselves may make the
endemism of a flora more pronounced by creating their own difficulties
and by standing in the way of their own dispersal to outside regions. It
has been shown that some of the endemic Hawaiian genera (see Note 68)
have deteriorated in their capacity for dispersal by birds; and similar
remarks are made with reference to the genera Sicyos (page 365) and
Eugenia (page 350). Genera with stone fruits like Elæocarpus possess in
the different species stones of various sizes, some of them suitable in
point of size for conveyance in a bird’s body over an ocean, others so
large that one could only predicate for them a limited capacity for
distribution by birds over a few hundred miles of sea. One, for
instance, could safely assume that species of Elæocarpus, with stones an
inch and over in size, that occur in Fiji and Hawaii, are not suited for
distribution over an ocean now; whilst other species found in New
Zealand and Rarotonga have stones less than half this size, which are
quite fitted for distribution by birds over broad tracts of ocean (page
337).

This brings us to discuss the relative difficulties presented from the
dispersal-standpoint by the forest floras of Hawaii, Fiji, and New
Zealand. It is with the forest floras that nearly all the difficulties
of distribution lie; and I hope I shall not be considered presumptuous,
or at all events too heterodox, in expressing the opinion that judging
from the details given in Kirk’s _Forest Flora of New Zealand_ those
islands present no greater difficulties for the student of
plant-distribution, if we exclude the Coniferæ, than either Fiji or
Hawaii. Indeed, even with the Conifers included, New Zealand presents
fewer problems than Fiji, and Hawaii has its own special difficulties
connected with the inland species of the Leguminosæ. There is, on the
other hand, no special New Zealand difficulty. It possesses the Conifers
in common with Fiji; and it shares with Fiji and Hawaii genera like
Elæocarpus and Sideroxylon, that take a foremost place amongst the trees
of the Pacific forest flora presenting puzzling points to the student of
distribution. The existence of Elæocarpus in New Zealand admits of a
simpler explanation than the occurrence of the same genus in Hawaii.
Pandanus in Fiji is a more difficult genus from the standpoint of
dispersal than Corynocarpus in New Zealand, and in fact, than any of the
non-coniferous genera of forest trees in that region.

Whilst it is likely that birds of the genus Porphyrio have, up to almost
recent times, been active in distributing the seeds of New Zealand
plants outside the region (see p. 296), it would seem that the
fruit-pigeons, as represented by a solitary peculiar species of
Carpophaga, have long since ceased to be active in this direction. It is
true that Sir W. Buller gives a long list of trees, including
Corynocarpus, Elæocarpus, Litsea, Olea, Podocarpus, and many others, the
fruits of which are appreciated by this pigeon; but since the bird is
confined to this region its efforts in plant-dispersal possess only a
local interest. Mr. G. M. Thomson, indeed, has expressed the opinion
(_Trans. and Proc. N.Z. Instit._ vol. 33) that in recent times not a
single plant has been added through the agency of birds to the New
Zealand flora. Besides the regular migratory birds two cuckoos only
reach the region, the one from Australia and the other from Polynesia;
whilst Australian birds which had managed to survive the long flight
across the ocean have been met with only at times on the west coast of
the North Island. From the standpoint adopted in this work we should
have expected that, with the exception of current-dispersed plants, New
Zealand would be out of touch with the world outside. Varied only by
occasional inrushes of plants, its history, dating back to the Mesozoic
age, has been one of insular isolation.

When, however, we apply the principles of plant-dispersal in the
Pacific, deducible from the study of the Hawaiian flora, we learn that
the stocking of New Zealand with its plants could have been carried out
(with the exception of the Coniferæ and a few other genera like Fagus
that are in a geological sense ancient denizens in this region) by the
agencies that stocked Hawaii with its flora. New Zealand genera like
Elæocarpus, Sideroxylon, Sophora, etc., that are represented in the
forests of Hawaii, could not be taken to illustrate any former
continental connection. If we, so-to-speak, put the New Zealand forest
flora in the Hawaiian sieve, all will pass through with the exception of
Fagus, the genera of the Coniferæ, and plants of similar history in high
southern latitudes. This residuum belongs more to the palæobotanist than
to the student of means of dispersal.

I should be inclined to think that the tropical genera of the New
Zealand flora, more especially of the forest flora, reached that region
during the glaciation of the northern hemisphere, when the Indo-Malayan
plants were, so-to-speak, cornered in the Western Pacific. Yet it must
be noted that these are, as a rule, genera that either display an
indifference to the varying thermal conditions of different latitudes or
are known to at times extend their range beyond the tropics. Thus
Elæocarpus and Freycinetia are equally at home in the temperate
rain-forests of New Zealand and in the tropical rain-forests of
Polynesia and Malaya; whilst widely-spread tropical genera like
Pittosporum and Peperomia, that occur in New Zealand, exhibit their
power of adaptation to varying climates by extending outside the tropics
in other regions and by their vertical range in the Hawaiian mountains,
where they are found alike at low elevations a few hundred feet above
the sea and at altitudes of 6,000 or 7,000 feet.

All these plants, however, are in a relative sense recent intruders.
When the student of dispersal looks at the long list of the conifers of
the New Zealand forest flora and reflects that he knows but little of
their means of dispersal, and that if his acquaintance were far greater
it would not avail him much, he has no choice but to take his place
behind the earlier investigators of the flora, and to see in these trees
evidence in favour of a remote continental period, probably referable to
the mesozoic age.

_A Discussion of some Means of Dispersal._—Not many authors seem to have
discussed the possibility of insects as agents of seed-dispersal in the
Pacific. They appear to me quite suited for transporting the spores of
ferns and lycopods as well as the minute seeds of plants like the
orchids and the begonias. Darwin, who allowed few possible means of
dispersal to escape his notice, procured the germination, as my readers
will remember, of grass seeds found in the dung of Natal locusts. When
on the barren summit of Mauna Loa, I noticed that the recently dead
bodies of some butterflies, that had been carried up the slopes from the
forests below by the “southerly updraught,” were already attacked by
bugs, parasites that must have been transported from the lower regions
by some of the numerous larger insects that are blown up the slopes.

In Note 61 the occurrence of the wind-blown insects on the summit of
Mauna Loa is described. That insects can be transported into the upper
regions of the atmosphere by ascending air-currents was long ago
remarked by Humboldt, and the subject has been discussed with his usual
acumen by Whymper (_Travels amongst the Great Andes of the Equator_).
Carried along in the higher air-currents these insects might finally be
deposited at places far distant from their home. One reads occasionally
extraordinary accounts of a rain of insects. A very circumstantial
account was given to me when I was on Keeling Atoll of a shower of
dragon-flies that fell on the islands, their remains being found in
quantities in the lagoon. Dragon-flies, it is known, are often found at
sea far from land, and one species has been observed nearly all over the
world, including the Pacific islands. In this connection it is
interesting to recall Mr. M’Lachlan’s remark in his article on the
dragon-fly in the seventh volume of the _Encyclopædia Britannica_ that
some of the earliest fossil forms seem to have been washed ashore after
having been drowned at sea.

Another creature that has been often ignored as a possible agent in
seed-dispersal is the bat. Bats are found all over the world, including
the oceanic groups, and one can scarcely doubt that they must have often
transported seeds, at all events in their hair. They are found at times
high up in mountainous regions, and Sir H. Johnston, in his recent work
on the Uganda Protectorate, refers to the occurrence of bats at an
altitude of 13,000 feet. The large frugivorous bats (Pteropidæ) are
known to be very destructive feeders; but I doubt whether they swallow
the seeds. Dr. Warburg, as is remarked in Chapter XXV, says that they
feed on the flowers of Freycinetias, and I have already observed that
they visit the flowers of Geissois ternata in Fiji (p. 394). In this
fashion Dr. Warburg regards them as agents in pollenisation; and it
seems to me that if, as appears likely, they are attracted by trees with
large, brightly-coloured flowers, they would often aid in the dispersal
of the minute seeds of trees like Metrosideros.

Until recently sea-birds, and some particular birds of passage, have
been generally considered as only fitted for dispersing seeds in their
plumage. That they can also transport seeds inside their bodies is shown
below. Dr. R. Brown in his book entitled _Our Earth and its Story_,
1888, gives a general account of plant-dispersal with numerous
references to the Literature on the subject. On the direct route between
Scotland and Cape Farewell in Greenland snow-buntings (Plectrophanes
nivalis) and other birds of passage frequently alight, as we are told,
on ships when hundreds of miles from land. Dr. Brown says that when
taking this voyage he examined dozens of these birds. Only in one case,
however, did he find any seeds, namely, in the case of a snow-bunting
which carried, attached to its plumage, an achene of, perhaps, a
Ranunculus, and in its gizzard a seed like that of a Suæda. My discovery
of a small, hard seed in the gizzard of a Cape-pigeon (Daption capensis)
550 miles east of Tristan da Cunha has been referred to by Mr. Hemsley
in his introduction to the _Botany of the “Challenger” Expedition_ (p.
45). On p. 188 I have mentioned the probable dispersal of the seeds of
Cæsalpinia by frigate-birds and boobies; and in Note 59 reference is
made to some large seeds found in the crop of the Fulmar petrel.

Gulls, when they nest at the coast, where the sea-thrift (Armeria
vulgaris) and the sea-campion (Silene maritima) thrive, or inland
amongst the heather-covered slopes, must often carry the seeds of these
plants from place to place in their plumage (see Notes 15 and 16); but,
as shown below, they can also disperse plants with fleshy fruits which
at times form their food. Gulls, geese, and arctic grouse take an
important part in the dispersal of seeds in the cold latitudes of the
northern hemisphere; and few things are more suggestive in this way to
the student of distribution than the data supplied by Ekstam, Hesselman,
Sernander, and others for the region including Spitzbergen, Nova Zembla,
and Arctic Norway. The history of the discussion relating to the flora
and fauna of Spitzbergen reproduces in its main features the various
stages in the controversy that has been waged in connection with the
Pacific islands.

When Ekstam published, in 1895, the results of his observations on the
plants of Nova Zembla, he observed that he possessed no data to show
whether swimming and wading birds fed on berries; and he attached all
importance to dispersal by winds. On subsequently visiting Spitzbergen
he must have been at first inclined, therefore, to the opinion of
Nathorst, who, having found only a solitary species of bird (a
snow-sparrow) in that region, naturally concluded that birds had been of
no importance as agents in the plant-stocking. However, Ekstam’s
opportunities were greater, and he tells us that in the craws of six
specimens of Lagopus hyperboreus shot in Spitzbergen in August he found
represented almost 25 per cent. of the usual phanerogamic flora of that
region, in the form of fruits, seeds, bulbils, flower-buds, leaf-buds,
&c. This observer now also realised the importance of gulls and geese in
dispersing certain types of plants in those latitudes. Species of Larus,
he says, consume greedily all kinds of berries, and especially those of
Empetrum nigrum, the stones of which were found uninjured in their
droppings by Professor Lagerheim in Arctic Norway. Geese, as we are also
informed, are hearty plant-eaters in Spitzbergen; and Ekstam found in
their droppings the fruits of Oxyria digyna as well as an abundance of
uninjured bulbils of Polygonum viviparum, some of which proved to be
capable of growth (See Ekstam in _Tromso Museums Aarshefter_, vols. 18
and 20, 1895-7).

The result of Ekstam’s observations in Spitzbergen was to lead him to
attach a very considerable importance in plant-dispersal to the agency
of birds; and when in explanation of the Scandinavian elements in the
Spitzbergen flora he had to choose between a former land connection and
the agency of birds, he preferred the bird.

I have gone into some detail in this matter because the Spitzbergen
controversy in some respects might have equally centred around New
Zealand or some of the large continental islands of the tropical
Pacific. There is at first the endeavour in the absence of precise
knowledge to disregard the bird and to look for a land connection. With
the increase in our acquaintance with the efficacy of bird-agency in
seed distribution there is the abandonment of such a view. In both
localities, however, there are the same counter-indications of the
insect faunas, and the same considerations are raised by the absence or
presence of larger animals in the regions concerned. The principal
difference lies in the frozen sea, and yet, strangely enough, it does
not seem to affect the problem much. It would indeed appear that the
questions raised by the floras and faunas of the Pacific islands are not
peculiarly Pacific in their character; and it is probable that the
difficulties here presented are repeated in one form or other in the
case of large islands over all the globe.

_On the efficacy of Ducks and other Waterfowl in the Distribution of
Aquatic Plants._—It is highly probable that aquatic plants, like the
beach plants distributed by the currents and the ferns and lycopods
distributed mainly by the winds, have changed much less in the course of
ages than the plants of the inland forest. This in all three cases is
chiefly due to the uninterrupted freedom of communication by means of
the dispersing agency.

Wild ducks and their kind are active agents in the distribution of the
seeds of aquatic plants; but it is curious that the early experiments of
Caspary went far to discredit them in this respect. As quoted by Dr.
Schenck in his _Die Biologie der Wassergewächse_, 1886, he fed tame
ducks with the seeds of water-lilies and found that in a short time they
thoroughly digested the seeds. Those familiar with the seeds of our
British species of Nuphar and Nymphæa will not be surprised at such a
result; but, unfortunately, the inference drawn from this experiment has
been by some extended to aquatic plants in general. Since the seeds or
seed-vessels of some aquatic or semi-aquatic plants of the genera
Potamogeton, Sparganium, &c., appeared to me to be quite fitted for
conveyance without injury in a duck’s body, I made several years ago a
number of observations on this subject, the results of which were
published in _Science Gossip_ for September, 1894.

Out of 13 wild ducks obtained in the London market and stated to have
been sent from Norfolk and Holland, eleven contained in their stomach
and intestines 828 seeds, which I thus classed:—

295 seeds of Sparganium in 8 birds
41 seeds of Potamogeton in 3 birds
270 seeds of Cyperaceæ in 5 birds
222 not identified

In the case of four birds the germinating capacity of the seeds was
tested, and in three cases very successfully. The seeds of Potamogeton,
Sparganium, and of the Cyperaceæ germinated readily in water, but few of
them failing, the process beginning in a few days or a few weeks. At
that time I was conducting an extended series of observations on the
postponement of germination of the seeds of aquatic plants, the results
of which were published in the _Proceedings of the Royal Physical
Society of Edinburgh_ for 1897. It was there shown that the seeds of
these plants often postpone their germination to the second and even to
the third spring. It thus happened that, whilst seeds obtained from the
stomach and intestines of the wild duck germinated in a few days or
weeks, I had to wait often a year and more for such a result with seeds
in their ordinary condition. This was well brought out in another
experiment made on a domestic duck, which I have described on page 369.
That wild ducks are to be regarded in the light of “flying germinators”
is thus very evident.

_Summary._

(1) In explanation of the shifting of the source of the Polynesian
plants from the New to the Old World, it is suggested that during the
glaciation of the northern hemisphere the Indo-Malayan plants entering
this region were “cornered” in the tropical Western Pacific, and were
only set free after the cold period had passed away, when they overran
Polynesia.

(2) Whilst the age of the Conifers is placed in the Mesozoic period,
that of the Compositæ is accredited to the Tertiary period, and the era
of Malayan immigration followed the glacial epoch.

(3) The suspension to a great extent of the agencies of plant dispersal
in the Pacific in later times is connected with a general principle
affecting the whole plant-world. With the secular drying up of the globe
the differentiation of climate, bird, and plant have gone on together,
the range of the bird being mainly controlled by the climate and the
range of the plant being largely dependent on the bird.

(4) Accepting Hawaii as entirely insular in its history, it is pointed
out that the principles deducible from the study of its flora can be
applied to the forest-flora of New Zealand, with the exception of the
Conifers and some genera that are ancient denizens of Antarctic
latitudes, and indicate a remote continental age dating back to the
Mesozoic period. It is suggested that the Indo-Malayan element in its
flora arrived there during the glaciation of the northern hemisphere.

(5) Insects and bats have probably been effective agents in
seed-dispersal in the Pacific, and it is shown that sea-birds carry
seeds in their stomach and intestines as well as in their feathers.

(6) It is shown that birds of the grouse family, gulls, and geese are
active seed-dispersers in cold northern latitudes, and that the
discussion of their influence in stocking Spitzbergen with its plants
reproduces many of the points of the controversy concerning the floras
of the continental islands of the South Pacific.

(7) The results of experiments and observations are cited to establish
the efficacy of ducks in distributing the seeds of aquatic plants, the
seeds ejected in their droppings germinating in a few days or weeks,
whilst those remaining in the pond or river often do not germinate for a
year or more.

CHAPTER XXXIV

GENERAL ARGUMENT AND CONCLUSION

THE problems concerned in the study of the floras of the Pacific islands
from the standpoint of dispersal are here approached through the buoyant
quality of the seed and fruit; and it is shown when dividing the plants
into two groups, those with buoyant and those with non-buoyant seeds or
fruits, that there has been at work through the ages a great sorting
process, by which the plants belonging to the group first named have
been mostly gathered at the coast. Its operation may be also observed
within the limits of a genus, where the species possessing seeds or
fruits that float is stationed at the coast, whilst the species with
seeds or fruits that sink makes its home inland.

When the principle here involved is applied to the British flora, it
presents itself as part of a much wider principle, by which plants
endowed with buoyant seeds and fruits have been stationed at the
water-side, whether on a river-bank, or beside a lake or pond, or on a
sea-beach. The broader principle proves in its turn to belong to a far
larger scheme, in which the fitness or unfitness of a plant to live in a
physiologically dry station appears as the primary determining quality,
the xerophyte (the plant of the dry station), provided with buoyant seed
or fruit, finding its way to the coast, and the hygrophyte (the plant
growing under more moist conditions), that is similarly endowed,
establishing itself by the side of the river, or the lake, or the pond.

When dealing with the general character and composition of the
strand-plants of the tropical Pacific, it is shown that in Fiji the
beach-plants often assert their primary xerophilous habit or fitness for
occupying any dry station by extending into the inland plains on the dry
sides of the islands. The Fijian shore-plants are divided into three
formations, those of the beach, those of the mangrove-swamp, and those
of intermediate stations on the borders of the swamps. The great
majority of the Fijian shore-plants are dispersed by the currents. The
Tahitian Islands, which are representative of Eastern Polynesia, lack
the mangroves and most of the plants that grow at the margin of a
mangrove-swamp; and their strand-flora is mainly composed of plants of
the beach, such as are dispersed by the currents far and wide in
tropical regions. The Hawaiian strand-flora is very meagre in its
character, lacking not only the plants of the mangrove and intermediate
formations, but almost all the large-fruited beach-trees of the South
Pacific. Since Hawaii possesses but few current-dispersed shore-plants
that are not found in the New World, reasons are given for the inference
that such shore-plants were originally brought by the currents from
America, and not from the South Pacific.

We are led on various grounds to the conclusion that tropical
shore-plants distributed by currents belong to two great regions, the
American including the west coast of Africa, and the Asiatic, or Old
World Region, which includes the African east coast. It is held that
America is so placed with regard to the currents, that it is a
distributor, and not a recipient of tropical shore-plants dispersed by
that agency. From this it follows that all cosmopolitan tropical
beach-plants that are dispersed by the currents have their homes in
America.

The results of observation and experiment are given to show that there
is no direct relation between the specific weight of seeds and fruits
and the density of sea-water. Yet, although the floating or sinking of a
seed or fruit is but an accidental attribute, it has had indirectly a
far-reaching influence not only on plant-distribution, but on
plant-development. In accordance with this want of relation between the
specific weight of seeds and fruits and the density of sea-water, the
great variety of structures concerned with buoyancy are regarded in the
main, after a detailed examination of their character, as not arising
from adaptation. Rather, it is urged, is buoyancy connected with
structures that now serve a purpose for which they were not originally
developed. Nature, it is held, has never concerned herself directly with
providing means of dispersal of any sort.

In the discussion of the relation between the littoral and inland
Pacific floras, it is shown, as a result of the examination of those
genera possessing both shore and inland species, that they have been on
the whole developed on independent lines. Two special difficulties in
explaining the modes of dispersal of plants of the Pacific islands here
come into prominence. There is the Hawaiian difficulty, where with
genera containing both shore and inland species only the last are found
in Hawaii; and although the shore-plants are known to be dispersed by
the current, the inland plants display little or no capacity for this or
any other mode of dispersal. Here belong the Leguminous genera
Canavalia, Erythrina, Mezoneuron, and Sophora, and the Apocynaceous
genus Ochrosia; and it is assumed that the inland Hawaiian species are
derived from a current-dispersed shore-plant that has since disappeared
from the group. The Fijian difficulty is displayed in those genera where
both coast and inland species occur in the islands, but no known
existing means of dispersal across an ocean can be postulated for the
inland plants, though the shore species are distributed by the currents.
Of such genera Pandanus is the best example, and it is pointed out that
this genus presents the same difficulty in the Mascarene Islands, in
which case the agency of the extinct Columbæ is invoked.

As illustrating the methods of observation and experiment employed by
the author, the Leguminous shore-plants Afzelia bijuga, Cæsalpinia
bonducella, and Entada scandens are discussed at length; and in the
chapters on the enigmas of the Leguminosæ in the Pacific it is pointed
out that the behaviour of the plants of this order is a source of much
perplexity, and that they conform to no single rule of dispersal.

Coming to the inland plants of this region, the Fijian, Tahitian, and
Hawaiian groups are taken as the chief centres of distribution in the
Pacific. After discussing the relative sizes, the altitudes, and the
climates of these three archipelagoes, it is shown that Hawaii, on
account of the far greater altitude of the islands, is characterised by
a special mountain flora, and that it is comparable with Fiji, and to a
great extent also with Tahiti, only as regarding the plants of the
levels below 4,000 or 5,000 feet.

The first era of the plant-stocking is designated the Age of Ferns, and
it is observed that, whilst in Hawaii nearly half of the ferns and
lycopods are peculiar to that group, very few new species have been
developed in the Fijian and Tahitian regions.

The next era in the floral history of these islands is represented in
the first era of the flowering plants. This is indicated by the endemic
genera, which are particularly numerous in Hawaii, relatively scanty in
Fiji, and very few in Tahiti. On account of their preponderance, the era
is designated the Age of Compositæ and Lobeliaceæ. The genera of these
two orders, though mainly characteristic of Hawaii, are also to be found
in the Tahitian region, but they are absent from the Fijian area.
Chiefly American in their affinities, their dispersion over the Pacific
took place during the Tertiary submergence of the archipelagoes of the
Western Pacific, in which are included the groups of the Fijian area
(Fiji, Samoa, Tonga). These early forms of Compositæ and Lobeliaceæ are
often arborescent in habit; and it is observed that Tree-Lobelias also
occur high up the slopes of lofty mountains in tropical regions, as in
Equatorial Africa, under conditions similar to those prevailing on the
slopes of the Hawaiian mountains, where the Tree-Lobelias, termed by Dr.
Hillebrand “the pride of our flora,” abound.

The other Hawaiian endemic genera, marking the first chapter in the
history of the flowering plants, arrange themselves in two groups, one
chiefly American in general affinities, and containing highly
differentiated Caryophyllaceæ, Labiatæ, &c.; the other largely Malayan,
and indicating the close of the first era of the flowering plants, when
the main source of the plants was shifted from America to the Old World.
The Fijian endemic genera, which are few in number, miscellaneous in
appearance, and disconnected in character, are regarded as having
probably acquired their endemic reputation through their failure at
their sources in the regions to the west.

The second era of the flowering plants is indicated by the non-endemic
genera. Here we are concerned on the one hand with a mountainous flora
mainly Hawaiian, in which genera from the New Zealand and Antarctic
floras take a conspicuous part, and on the other with a low-level flora
chiefly derived from Indo-Malaya, and including the plants of the lower
slopes of Hawaii below 4,000 and 5,000 feet, and the floras in mass of
Fiji and Tahiti.

On account of their lower altitude, the extensive mountain flora of
Hawaii is but scantily developed in Tahiti, and is represented by a mere
remnant in Fiji and Samoa. Two-thirds of the Hawaiian non-endemic
mountain genera contain only species restricted to the group, and,
although amongst these disconnected genera, Acæna, Gunnera, Coprosma,
Lagenophora, &c., of the New Zealand and Antarctic floras take a
prominent part, a large proportion of the genera like Ranunculus, Rubus,
Artemisia, Vaccinium, and Plantago represent generally the flora of the
north temperate zone on the summits of tropical mountains. The Tahitian
mountain flora, scanty as it is when judged by the non-endemic genera,
displays much kinship with the Hawaiian mountain flora; but this kinship
is mainly confined to genera from high southern latitudes, such as
Coprosma, Cyathodes, Astelia, &c. In the possession on its mountain
slopes of the three genera of the Coniferæ, Dammara, Podocarpus, and
Dacrydium, the Fijian region is distinguished from that of Tahiti and
Hawaii; and it is assumed that they mark the site of a continental area
in the Mesozoic period, when the Tahitian and Hawaiian groups did not
exist.

The era of the non-endemic genera, in so far as it is concerned with the
low-level flora of Hawaii and the floras in mass of the areas of Fiji,
Samoa and East Polynesia, is termed Malayan, because many of the genera
are thence derived. Here we are dealing with all the oceanic groups of
the tropical Pacific, and not with a portion of them, as in the case of
the Age of Coniferæ, in the Secondary period, that was limited to the
Western Pacific, or in the case of the Age of Compositæ and Lobeliaceæ
that was restricted during the Tertiary epoch to the Hawaiian and
Tahitian regions. The first part of this era, as is indicated by the
endemic species, is an age of complete isolation in Hawaii, and of
partial isolation in the groups of the southern region. Amongst the
genera typical of this period are Pittosporum, Gardenia, Psychotria,
Cyrtandra, and Freycinetia. A later period in this era of the general
dispersal of Malayan plants over the Pacific is one where the extremely
variable or polymorphous species plays a conspicuous part, as
represented in such genera as Alphitonia, Dodonæa, Metrosideros,
Pisonia, and Wikstrœmia, the general principle being that each genus is
at first represented by a widely ranging very variable species, which
ultimately ceases to wander and settles down, and becomes the parent of
different sets of species in the several groups.

The facts of distribution in this age of general dispersion are just
such as we might look for in the case of a general dispersal over the
oceanic groups of the Pacific, with the altitudes of the islands playing
a determining part. But it should be remarked that the greater number of
the genera that have entered the Pacific from the Old World have not
advanced eastward of the Fijian region, half of the Fijian genera not
occurring in the Hawaiian and Tahitian regions. The explanation of this
is to be found, not in any lack of capacities for dispersal, but in a
want of opportunities. The story of plant-distribution in the Pacific is
bound up with the successive stages of decreasing activity in the
dispersing agencies. The area of active dispersion, as illustrated by
the non-endemic genera, at first comprised the whole of the tropical
Pacific. It was afterwards restricted to the South Pacific, and finally
to the Western Pacific only. The birds that carried seeds all over this
ocean became more and more restricted in their ranges, probably on
account of increasing diversity of climatic conditions. The plants of
necessity responded to the ever narrowing conditions of bird-life in
this ocean, and the differentiation of the plant and the bird have taken
place together.

During the stages of decreasing activity in the dispersing agencies, the
widely-ranging highly variable species continued to be an important
factor in the development of new species in the different groups. The
_rôle_ of the polymorphous species has always been a conspicuous one in
the Pacific.

Yet, as in the case of the Cyrtandras, it is shown that the display of
great formative power within a genus is not a peculiarity of an insular
flora; that the isolation of an oceanic archipelago does not exclusively
induce “endemism,” but only intensifies it; that the development of new
species may be nearly as active on a mountain in a continent as on an
island in mid-ocean; and that this is equally true of a land genus, like
Embelia, exposed to an infinite variety of conditions, and of an aquatic
genus, like Naias, where the conditions of existence are relatively
uniform all the world over.

In framing a scheme by which the eras of the floral history of the
Pacific are brought into correlation with those of geological time, the
age of the Coniferæ is placed in the Secondary period, that of the
Compositæ and Lobeliaceæ in the Tertiary period, whilst the era of
Malayan immigration is regarded as mainly post-glacial. The age of the
Coniferæ is concerned only with the Western Pacific, since the Hawaiian
and Tahitian islands had not then been formed. The age of the Compositæ
and Lobeliaceæ is concerned only with Hawaii and Tahiti, since the
islands of the Western Pacific were then more or less submerged. That of
the Malayan plants affects the whole Pacific as at present displayed to
us.

In the chapter on the viviparous mangroves of Fiji it is shown that both
the Asiatic and the American species of Rhizophora (R. mucronata and R.
mangle) exist in that group, and that there is in addition a seedless
form, the Selala, which, although intermediate in character between the
two other species, comes nearest to the Asiatic plant. Reasons are given
for the belief that the Selala is derived from the Asiatic species (R.
mucronata), not as the result of a cross but as connected with its
dimorphism; and in support of this it is pointed out that on the Ecuador
coast of South America, where only the American species exists, a
dimorphism is also displayed, one of the forms approaching in several of
its characters the Fijian Selala, though fruiting abundantly and bearing
the impress of a closer connection with the typical American species
than with the Asiatic plant. The view that Rhizophora mangle reached the
Western Pacific from America is rejected, and it is considered that this
species was originally as widely diffused in the Old World as in
America, and that it now survives only in a few places in the tropics of
the Old World. The results of detailed observations on the modes of
dispersal and on the germinating process both with Rhizophora and
Bruguiera are given; and the absence, as a general rule, of any period
of rest between the fecundation of the ovule and the germination of the
seed is established.

A special chapter is devoted to the significance of vivipary, and it is
considered that a record of the history of vivipary on the globe is
afforded in the scale of germinative capacity that begins with the
seedling hanging from a mangrove and ends with the seed that is detached
in an immature condition from an inland plant. It is suggested that with
the drying up of the planet in the course of ages the viviparous habit,
which was once nearly universal, has been for the most part lost except
in the mangrove swamp, which to some extent represents an age when the
earth was enveloped in cloud and mist and the atmosphere was saturated
with aqueous vapour. The lost habit is at times revived in the abnormal
vivipary of some inland plants, and traces of it are seen in the
abnormal structure of the seeds of some genera of the Myrtaceæ, like
Barringtonia, and in the seeds of genera of other orders. With the
desiccation of the planet and the emergence of the continents there has
been continual differentiation of climate resulting in seasonal
variation and in the development of the rest-period of the seed.

With the secular drying of the globe and the consequent differentiation
of climate is to be connected the suspension to a great extent of the
agency of birds as plant-dispersers in later ages, not only in the
Pacific Islands but over all the tropics. The changes of climate, bird,
and plant have gone on together, the range of the bird being controlled
by the climate, and the distribution of the plant being largely
dependent on the bird.

The history of climate, the history of the continents and of the oceans,
the history of life itself, but only in the sense below defined, all
belong to that of a desiccating world, or rather of a planet once
sunless and enveloped in mist and cloud, that through the ages has been
drying up. Life’s types were few and the sea prevailed, and one climate
reigned over the globe. With the diminution of the aqueous envelopes the
continents began to emerge, climates began to individualise, and
organisms commenced to differentiate, and thus the process has run on
through the past, ever from the general to the special both in the
organic and in the inorganic world.

The same story of a world drying up is told by the marine remains left
stranded far up some mountain slope, or by the bird akin to no other of
its kind that Time has stranded on some island in mid-Pacific. The bird
generalised in type that once ranged the globe is now represented over
its original range by a hundred different groups of descendants,
confined each to its own locality. Climate, once so uniform, now so
diversified, has by restricting the range of the bird favoured the
process of differentiation, and the plant dependent on the bird for its
distribution has in its turn responded to these changes.

The _rôle_ of the polymorphous species belongs alike to the plant and to
the bird. A species that covers the range of a genus varies at first in
every region and ultimately gives birth to new species in some parts of
its range. Then the wide-ranging species disappears and the original
area is divided up into a number of smaller areas each with its own
group of species. Each smaller area breaks up again, and forms, yet more
specialised, are produced; and thus the process of subdivision of range
and of differentiation of form goes along until each island in an
archipelago owns its bird and each hill and valley has its separate
plants. This is not the path that Evolution takes, since beyond lies
extinction whether of plant or of bird. Such is the upshot of the
process of differentiation exhibited in the development of species and
genera in the Pacific Islands, or, indeed, in any oceanic groups. It can
never do more than produce a Dodo or a Kiwi, or amongst the plants a
Tree-Lobelia.

Evolution here and elsewhere is a thing apart from species and genera,
which are but eddies on the surface of its stream. It is a scheme of
life introduced into a much conditioned world, and adaptation in endless
forms is the price it has had to pay. The whole story of life on this
earth is a story of a sacrifice, of an end to be won, but of a price to
be paid. Immortality is in the scheme, but death is the price of
adaptation. The same theme runs through our conceptions of the spiritual
life. There is the same duality, evolution adapting its scheme to the
exigencies of the physical world, the good principle ever in conflict
with the evil, and at times compelled to adapt itself to attain its
ends. There is the tale of adaptation in the one case and of sacrifice
in the other, and success is reached in both.

APPENDIX

LIST OF NOTES

Note 1. On the number of known species of Fijian flowering plants.

Note 2. The littoral plants of Fiji.

Note 3. Results of long flotation experiments on the seeds or
seedvessels of tropical littoral plants.

Note 4. Table illustrating the degree of buoyancy of the seeds and
fruits of inland Fijian plants.

Note 5. The inland Fijian plants possessing buoyant seeds or fruits.

Note 6. Table showing the degree of buoyancy of the seeds and fruits of
some inland Hawaiian plants.

Note 7. Some inland Hawaiian plants possessing buoyant seeds or fruits.

Note 8. The pyrenes of Morinda.

Note 9. The buoyancy of the fruits of Calophyllum.

Note 10. The buoyancy experiments on British plants.

Note 11. The effect of sea-water immersion on the germinating capacity
of seeds and seed-vessels.

Note 12. The buoyancy of the fruits of Galium aparine.

Note 13. The buoyancy of the seeds of Convolvulus sepium.

Note 14. Other long flotation experiments.

Note 15. The occurrence inland of Silene maritima.

Note 16. The buoyancy of the seeds or fruits of the British beach-plants
that also occur inland.

Note 17. The buoyancy of the seeds or fruits of the British littoral
plants that frequent salt-marshes and muddy shores.

Note 18. The buoyancy of the seeds or fruits of the British littoral
plants that are confined to the beach.

Note 19. On germination in sea-water.

Note 20. On the maximum heights reached by some shore plants in their
extension inland in Vanua Levu, Fiji.

Note 21. On the dwarfing of shore plants when extending inland in the
“talasinga” plains in Vanua Levu.

Note 22. The “talasinga” plains of Vanua Levu.

Note 23. Schimper’s grouping of the Indo-Malayan strand flora.

Note 24. Grouping of some of the characteristic plants of the strand
flora of Fiji.

Note 25. The strand flora of the Tahitian region.

Note 26. The Fijian shore plants not found in Tahiti.

Note 27. The intruders into the beach flora from the inland plants of
Tahiti.

Note 28. The littoral plants of the Hawaiian islands.

Note 29. Botanical notes on the coast plants of the Hawaiian islands.

Note 30. The beach drift of the Hawaiian islands.

Note 31. The inland extension of the shore plants of the Hawaiian
islands.

Note 32. The Fijian species of Premna.

Note 33. De Candolle’s list of plants dispersed exclusively by currents.

Note 34. The littoral plants of the eastern-most Polynesian islands.

Note 35. Distribution of the littoral plants with buoyant seeds or
fruits that occur in the Fijian, Tongan, Samoan, Tahitian, and
Hawaiian Groups.

Note 36. Hawaiian plants with buoyant seeds or fruits known to be
dispersed by the currents either exclusively or with the assistance of
frugivorous birds.

Note 37. On vivipary in the fruits of Barringtonia racemosa and Carapa
obovata.

Note 38. On the temperature and density of the surface water of the
estuaries of the Rewa River in Fiji and of the Guayaquil River in
Ecuador.

Note 39. On the Pacific species of Strongylodon.

Note 40. Precautions in testing seed-buoyancy.

Note 41. The buoyancy of the seeds of Convolvulus soldanella in
fresh-water and sea-water compared.

Note 42. On secular changes in sea-density.

Note 43. On the mucosity of small seeds and seed-like fruits when wet.

Note 44. Upon the effects of inland extension on the buoyancy of the
seeds or fruits of littoral plants.

Note 45. Tabulated results of the classification, according to
Schimper’s application of the Natural Selection Theory, of the buoyant
seeds and fruits of tropical littoral plants.

Note 46. On the modes of dispersal of the genus Brackenridgea.

Note 47. On the transport of gourds by currents.

Note 48. On the useless dispersal by currents of the fruits of the Oak
and of other species of Quercus, as well as of the Hazel (Corylus).

Note 49. On the distribution of Ipomœa pes capræ, Convolvulus
soldanella, and Convolvulus sepium.

Note 50. On the structure of the seeds and fruits of Barringtonia.

Note 51. On a common inland species of Scævola in Vanua Levu, Fiji.

Note 52. On the capacity for dispersal by currents of Colubrina
oppositifolia.

Note 53. On the genus Erythrina.

Note 54. On the genus Canavalia.

Note 55. The inland extension of Scævola kœnigii.

Note 56. On the capacity for dispersal by currents of Sophora tomentosa,
S. chrysophylla, and S. tetraptera.

Note 57. On the species of Ochrosia.

Note 58. On Pandanus.

Note 59. Seeds in petrels.

Note 60. Schimper on the halophilous character of littoral Leguminosæ
and of shore plants generally.

Note 61. Meteorological observations on the summit of Mauna Loa.

Note 62. On the relative proportion of vascular cryptogams in Fiji.

Note 63. On the table of vascular cryptogams of Tahiti, Hawaii, and
Fiji.

Note 64. On the distribution of the Tahitian ferns and lycopods.

Note 65. Distribution of some of the mountain ferns of Hawaii that are
not found either in Fiji or in Tahiti.

Note 66. Endemic genera of ferns in Hawaii.

Note 67. On the dispersal of Compositæ by birds.

Note 68. On some of the Hawaiian endemic genera excluding those of the
Compositæ and Lobeliaceæ.

Note 69. On the germination of Cuscuta.

Note 70. On beach-temperature.

Note 71. On the buoyancy of the seeds or seed-vessels of some Chilian
shore plants.

Note 72. On the southern limit of the mangrove formation in Ecuador.

Note 73. Additional note on the temperature of the dry coast of Ecuador,
between the island of Puna and the equator.

Note 74. Observations on the temperature of the Humboldt current from
Antofagasta northward between January and March, 1904.

Note 75. On the stranded massive corals of the genus Porites (?) found
on the coast of North Chile and Peru at Arica, Callao, and Ancon.

Note 76. Stranded pumice on English and Scandinavian beaches.

Note 77. On the mode of dispersal of Kleinhovia hospita.

Note 78. On the “Sea”, an unidentified wild fruit tree in Fiji.

Note 79. On willow-leaved river-side plants.

Note 80. Mr. Perkins on the Hawaiian Lobeliaceæ.

Note 81. On the vertical range of some of the most typical and most
conspicuous of the plants in the forests on the Hamakua slopes of
Mauna Kea, Hawaii.

Note 82. Aboriginal weeds.

Note 90. On the buoyancy of the seeds of Euphorbia amygdaloides and E.
segetalis.

Note 91. Mr. E. Kay Robinson on Aster tripolium.

NOTE 1 (page 13)

ON THE NUMBER OF KNOWN SPECIES OF FIJIAN FLOWERING PLANTS

Rather over 600 species of flowering plants are included in Seemann’s
_Flora Vitiensis_, excluding the weeds and the plants introduced by man.
Horne’s collections would probably add another 300 species; and many
more remain to be discovered.

NOTE 2 (page 13)

THE LITTORAL PLANTS OF FIJI

In the following table are incorporated the results of an extensive
series of observations and experiments on the buoyancy of the seeds and
fruits of the shore plants made by the author during his sojourn of two
years in Fiji, and based not only on prolonged buoyancy-tests, but also
on systematic examination of the stranded and floating seed-drift, both
of sea and river. The details would occupy many chapters: and it is only
possible here to give the bare results. Since Professor Schimper went
over much the same ground in the Malayan region, one enjoys in many
cases the great advantage of his authority; but a fair proportion of the
results are new; and, besides, there are a number of plants included,
the buoyancy of whose seeds or fruits has long been well established. In
all cases the seed or fruit is taken as it presents itself for dispersal
by the currents. Many of the plants are discussed with some detail in
various parts of this book, as indicated in the reference column of the
table.

Since the Gramineæ and the Cyperaceæ contain very few species suited for
direct transport by the currents over wide areas of sea, this list may
be regarded as containing nearly all the littoral flowering plants
possessing seeds or seed-vessels with any buoyancy of importance.

Nearly all the Tahitian strictly littoral plants are represented in
Fiji, and the few that have not been found there yet, such as Sesbania
grandiflora, Heliotropium anomalum, &c., may exist, as in the
first-named species, in the neighbouring Tongan group, and may probably
even exist in Fiji. Two other Tahitian littoral plants, that are widely
spread in the Pacific, namely, Suriana maritima and Sesuvium
Portulacastrum, are found in Tonga, and are included in my list of
Fijian shore plants, though not yet recorded from that group, where,
however, they will, without a doubt, be found by some future observer.

TABLE SHOWING THE BUOYANCY OF THE SEEDS OR FRUITS OF THE LITTORAL PLANTS
OF FIJI, EXCLUDING THE GRASSES AND, WITH ONE EXCEPTION, THE SEDGES

The letters placed before the plant name indicate that the species is
also found in Hawaii (H), in Tahiti (T), and in the Marquesas (M). The
Marquesan locality is only given where the plant is not in Tahiti.

The abbreviations in the reference column are as follows:

S=Schimper; G=Guppy; P=Earlier authorities and particularly the list
given by Hemsley in the Introduction to the _Botany of the Challenger
Expedition_.

Column headings:

A. Buoyancy of seeds or fruits.
B. Float for months.
C. Sink at once or in a week or two.
D. Authorities.
E. Pages of further reference. See also Index.

+--------------------------------------+--------------+-------+-------+--------+
| | | A | | |
| Species. | Family. +---+---+ D | E |
| | | B | C | | |
+--------------------------------------+--------------+---+---+-------+--------+
|HT Calophyllum inophyllum |Guttiferæ. | + |...|S.G.P. | 18 |
| | | | | | |
|HT Hibiscus tiliaceus |Malvaceæ. | + |...|S.G.P. | 21 |
| | | | | | |
| Hibiscus diversifolius (Jacq.) |Malvaceæ. | + |...|G. | 21 |
| | | | | | |
|HT Thespesia populnea |Malvaceæ. | + |...|S.G.P. | Note 3 |
| | | | | | |
| H Gossypium tomentosum (Nutt.) |Malvaceæ. |...| + |G. | |
| | | | | | |
| Heritiera littoralis |Sterculiaceæ. | + |...|S.G.P. | 45, 48 |
| | | | | | |
| T Kleinhovia hospita |Sterculiaceæ. | + |...|G. | 21 |
| | | | | | |
| T Triumfetta rhomboidea |Tiliaceæ. |...| +?| | |
| | | | | | |
| T Triumfetta procumbens |Tiliaceæ. |...| + |G. | 45 |
| | | | | | |
| T Suriana maritima |Simarubeæ. | + |...|S.G. | |
| | | | | | |
| Carapa moluccensis |Meliaceæ. | + |...|S.G.P. | 45 |
| | | | | | |
| Carapa obovata |Meliaceæ. | + |...|S.G.P. | 45 |
| | | | | | |
| T Ximenia americana |Olacineæ. | + |...|S.G. | 113 |
| | | | | | |
| Smythea pacifica (Seem.) |Rhamneæ. | + |...|G.P. | 106 |
| | | | | | |
|HT Colubrina asiatica |Rhamneæ. | + |...|G. | 137 |
| | | | | | |
|HT Dodonæa viscosa |Sapindaceæ. | + |...|S.G. | |
| | | | | | |
|HT Tephrosia piscatoria |Papilionaceæ. |...| + |G. | 45 |
| | | | | | |
| M Desmodium umbellatum |Papilionaceæ. |...| + |G. | |
| | | | | | |
|HT Dioclea violacea |Papilionaceæ. | + |...|G.P. | 82 |
| | | | | | |
| T Canavalia obtusifolia |Papilionaceæ. | + |...|S.G.P. |Note 54 |
| | | | | | |
| T Canavalia sericea |Papilionaceæ. | + |...|G. |Note 54 |
| | | | | | |
| T Canavalia ensiformis, var. turgida.|Papilionaceæ. | + |...|S.G.P.?|Note 54 |
| | | | | | |
|HT Mucuna gigantea |Papilionaceæ. | + |...|S.G.P. | 81 |
| | | | | | |
| T Erythrina indica |Papilionaceæ. | + |...|S.G.P. | |
| | | | | | |
|HT Strongylodon lucidum |Papilionaceæ. | + |...|G. | 82 |
| | | | | | |
|HT Vigna lutea |Papilionaceæ. | + |...|S.G. | 139 |
| | | | | | |
| Dalbergia monosperma |Papilionaceæ. | + |...|S.G. | 106 |
| | | | | | |
| Derris uliginosa |Papilionaceæ. | + |...|S.G.P. | 111 |
| | | | | | |
| Pongamia glabra |Papilionaceæ. | + |...|S.G.P. | |
| | | | | | |
| T Sophora tomentosa |Papilionaceæ. | + |...|S.G. |Note 56 |
| | | | | | |
| T Inocarpus edulis |Papilionaceæ. | +?|...|G.P. | |
| | | | | | |
|HT Cæsalpinia Bonducella |Cæsalpinieæ. | + |...|S.G.P. | 193 |
| | | | | | |
| T Cæsalpinia Bonduc |Cæsalpinieæ. | + |...|G.P. | 193 |
| | | | | | |
| Afzelia bijuga |Cæsalpinieæ. | + |...|G. | 173 |
| | | | | | |
| Cynometra sp. |Cæsalpinieæ. | +?|...|S.G. | |
| | | | | | |
| Entada scandens |Mimoseæ. | + |...|G.P. | 181 |
| | | | | | |
| Acacia laurifolia |Mimoseæ. |...| + |G. | 164 |
| | | | | | |
| T Leucæna Forsteri |Mimoseæ. |...| + |G. | |
| | | | | | |
| T Serianthes myriadenia |Mimoseæ. |...| + |G. | 424 |
| | | | | | |
| Parinarium laurinum |Rosaceæ. | + |...|G.P. | |
| | | | | | |
| Eugenia Richii |Myrtaceæ. |...| + |G. | |
| | | | | | |
| T Barringtonia speciosa |Myrtaceæ. | + |...|S.G.P. | |
| | | | | | |
| Barringtonia racemosa |Myrtaceæ. | + |...|G. | |
| | | | | | |
| Rhizophora mucronata |Rhizophoreæ. | + |...|S.G.P. | |
| | | | | | |
| Rhizophora mangle |Rhizophoreæ. | + |...|S.G.P. | |
| | | | | | |
| Bruguiera Rheedii |Rhizophoreæ. | + |...|G.P. | |
| | | | | | |
|HT Terminalia Katappa |Combretaceæ. | + |...|S.G.P. | |
| | | | | | |
| M Terminalia littoralis |Combretaceæ. | + |...|S.G.P. | |
| | | | | | |
| Lumnitzera coccinea |Combretaceæ. | + |...|S.G.P. | |
| | | | | | |
| T Gyrocarpus Jacquini |Combretaceæ. | + |...|G. | 423 |
| | | | | | |
| T Pemphis acidula |Lythraceæ. | + |...|S.G. | |
| | | | | | |
| T Luffa insularum (Gray) |Cucurbitaceæ. | + |...|G. | 426 |
| | | | | | |
|HT Sesuvium Portulacastrum |Ficoideæ. |...| + |G. | |
| | | | | | |
|HT Morinda citrifolia |Rubiaceæ. | + |...|S.G.P. | |
| | | | | | |
| T Guettarda speciosa |Rubiaceæ. | + |...|S.G.P. | |
| | | | | | |
| T Wedelia biflora |Compositæ. | + |...|G. | |
| | | | | | |
|HT Scævola Koenigii |Goodeniaceæ. | + |...|S.G.P. | |
+--------------------------------------+--------------+---+---+-------+--------+

TABLE SHOWING THE BUOYANCY OF THE SEEDS OR FRUITS OF THE
LITTORAL PLANTS OF FIJI, EXCLUDING THE GRASSES, AND WITH
ONE EXCEPTION, THE SEDGES (_continued_)

Column headings:

A. Buoyancy of seeds or fruits.
B. Float for months.
C. Sink at once or in a week or two.
D. Authorities.
E. Pages of further reference. See also Index.

+--------------------------------------+--------------+-------+-------+--------+
| | | A | | |
| Species. | Family +---+---+ D | E |
| | | B | C | | |
+--------------------------------------+--------------+---+---+-------+--------+
| T Cerbera Odollam |Apocynaceæ. | + |...|S.G.P. | |
| | | | | | |
| T Ochrosia parviflora |Apocynaceæ. | + |...|G.P. | |
| | | | | | |
|HT Cordia subcordata |Boraginaceæ. | + |...|S.G.P. | |
| | | | | | |
| T Tournefortia argentea |Boraginaceæ. | + |...|S.G.P. | |
| | | | | | |
|HT Ipomœa pes capræ |Convolvulaceæ.| + |...|S.G.P. | |
| | | | | | |
| H Ipomœa glaberrima (Boj.) |Convolvulaceæ.| + |...|G. | |
| | | | | | |
| Aniseia uniflora |Convolvulaceæ.| + |...|G. | |
| | | | | | |
| T Premna tahitensis |Verbenaceæ. | + |...|G. |Note 32 |
| | | | | | |
| Clerodendron inerme |Verbenaceæ. | + |...|S.G. | |
| | | | | | |
|HM Vitex trifolia |Verbenaceæ. | + |...|G. | |
| | | | | | |
|HT Cassytha filiformis |Lauraceæ. | + |...|G. | |
| | | | | | |
| T Hernandia peltata |Lauraceæ. | + |...|S.G. | |
| | | | | | |
|HT Wikstrœmia fœtida |Thymelæaceæ. |...| + |G. | |
| | | | | | |
| Drymispermum Burnettianum |Thymelæaceæ. |...| + |G. | |
| | | | | | |
| T Euphorbia Atoto |Euphorbiaceæ. |...| + |S.P.G. | |
| | | | | | |
| Excæcaria Agallocha |Euphorbiaceæ. | + |...|S.G. | |
| | | | | | |
| T Casuarina equisetifolia |Casuarineæ. |...| + |G. | |
| | | | | | |
|HT Tacca pinnatifida |Taccaceæ. | + |...|G. | 19 |
| | | | | | |
|HT Cocos nucifera |Palmeæ. | + |...|P. | |
| | | | | | |
|HT Pandanus odoratissimus |Pandaneæ. | + |...|S.G.P. | |
| | | | | | |
| Crinum asiaticum |Amaryllideæ. |...| + |G.P. | |
| | | | | | |
| Scirpodendron costatum |Cyperaceæ. | + |...|G. | 407 |
| | | | | | |
| Cycas circinalis |Cycadeæ. | + |...|G.P. | |
+--------------------------------------+--------------+---+---+-------+--------+

NOTE 3 (page 13)

RESULTS OF LONG FLOTATION EXPERIMENTS ON THE SEEDS OR SEED-VESSELS OF
TROPICAL LITTORAL PLANTS

At various times during the past twenty years I have made lengthened
experiments in England on the buoyancy in sea-water of the seeds or
seed-vessels of beach plants collected by me in the Solomon Islands, the
Fijis, Hawaii, Keeling Atoll, &c. In all the species enumerated below,
the floating powers were retained after twelve months’ immersion, the
seed-contents being to all appearance unharmed. In six species I
succeeded in getting the seeds to germinate after the experiment; and
there can be no doubt that the number of successful results would have
been largely increased, if I had not been obliged to resort to very
primitive methods in conducting the experiments. Some of the results are
referred to in a note to my paper on the flora of Keeling Atoll, dated
about 1889; and if I remember aright, Mr. Hemsley mentioned those
relating to Thespesia populnea and Ipomœa grandiflora in the _Annals of
Botany_, not long after. The others have not been previously published.
In one instance (Cæsalpinia bonducella) the flotation experiment was
prolonged to two and a half years, the seeds floating buoyantly and
being apparently quite sound at the end of the experiment.

As demonstrating that tropical seeds can be transported unharmed by
currents through cold latitudes, it should be noted that all these
experiments were conducted in England. In the cases of the Keeling Atoll
seeds the experiment was carried on through a very severe winter, the
vessel of sea-water being exposed to a degree of cold that kept
fresh-water frozen for three weeks on the same table. This did not
prevent the subsequent germination of the seeds of Thespesia populnea
and Ipomœa grandiflora. The same thing was established in a more natural
way by Lindman, who planted seeds of Entada scandens and Mucuna urens,
that had been stranded on the Norwegian coast, and found that they
retained their germinating capacity (see Sernander, p. 7).

The following are the seeds or seed-vessels that remained afloat after a
year’s flotation in sea-water, those that subsequently germinated being
preceded by G. In the other cases the germinating capacity was not
tested; but they were always sound in appearance when cut across at the
close of the experiment.

G Thespesia populnea (Malvaceæ)
Dioclea (violacea?) (Papilionaceæ)
G Mucuna gigantea, D C (Papilionaceæ)
G Mucuna urens, D C (Papilionaceæ)
Mucuna, sp. (Papilionaceæ)
Mucuna, sp. (Papilionaceæ)
G Strongylodon lucidum, Seem. (Papilionaceæ)
Sophora tomentosa, (Papilionaceæ)
G Cæsalpinia bonducella (Cæsalpinieæ)
Entada scandens (Mimoseæ)
Morinda citrifolia (Rubiaceæ)
Scævola Koenigii (Goodeniaceæ)
Cordia subcordata (Boragineæ)
Tournefortia argentea (Boragineæ)
G Ipomœa grandiflora, Lam. (Convolvulaceæ)
Tacca pinnatifida (Taccaceæ)

NOTE 4 (page 13)

TABLE ILLUSTRATING THE DEGREE OF BUOYANCY OF THE SEEDS AND FRUITS OF
INLAND FIJIAN PLANTS

(Unless otherwise indicated, the seeds or fruits sink at once or in a
day or two)

Abrus precatorius.
Acacia Richii.
Ageratum conyzoides.
Alphitonia excelsa.
Alpinia sp.
Alyxia (scandens?).
Artocarpus incisa.
Artocarpus integrifolia.
Barringtonia edulis (1 month)
Barringtonia sp.
Bauhinia sp.
Bischoffia javanica.
Cæsalpinia sp.
Calophyllum spectabile (2-4 weeks).
Calophyllum Burmanni (4-10 days).
Cananga odorata.
Canarium sp.
Canarium sp.
Canna indica.
Citrus aurantium (3-4 weeks).
Citrus decumana (1 month).
Citrus limonum (5 weeks).
Citrus vulgaris, R. (6-7 weeks).
Coix lachryma (2-7 days).
Commersonia platyphylla.
Cordyline sepiaria.
Couthovia corynocarpa (a few days).
Cucumis acidus (a few days).
Cucurbita sp. (several months).
Cupania sp.
Dammara vitiensis (7-10 days).
Dioscorea sativa (a few days).
Dioscorea sp.
Dracontomelon sylvestre.
Dracontomelon sp.
Elæocarpus sp.
Elæocarpus sp. (a few days).
Eranthemum sp.
Eugenia malaccensis (2-4 weeks).
Eugenia effusa? (4-7 days).
Eugenia confertiflora? (10-12 days).
Eugenia rariflora (a few days).
Eugenia corynocarpa (a few days).
Eugenia rivularis (a week).
Fagræa Berteriana (a few days).
Ficus Harveyi (7-10 days).
Ficus scabra (7-10 days).
Ficus sp. (7-10 days).
Gardenia vitiensis (4-5 weeks).
Geissois ternata.
Geophila reniformis.
Gnetum gnemon.
Grewia sp.
Guettarda sp. (a few weeks).
Hibiscus Abelmoschus (months).
Hibiscus seculentus.
Hydrocotyle asiatica (months).
Ipomœa batatas.
Ipomœa insularis (_nil_ or months).
Ipomœa peltata (weeks or months).
Ipomœa turpethum (_nil_ or weeks or months).
Ipomœa sp. (7-10 days).
Lindenia vitiensis (weeks or months).
Maba sp. (7-10 days).
Macaranga sp. (1-2 weeks).
Melastoma denticulatum.
Micromelum minutum.
Momordica Charantia (a few days).
Morinda Forsteri.
Mussænda frondosa.
Myristica sp. (3-7 days)
Myristica sp. (3-7 days)
Myrmecodia sp.
Nelitris vitiensis (a few days).
Nephelium pinnatum (a few days).
Ophiorrhiza leptantha.
Phyllanthus sp.
Phyllanthus sp.
Piper Macgillivrayi.
Pittosporum sp.
Pleiosmilax vitiensis.
Portulaca (lutea?).
Portulaca quadrifida.
Premna serratifolia.
Pritchardia pacifica.
Psychotria sp.
Psychotria sp.
Psychotria sp.
Psychotria sp.
Psychotria sp.
Ptychosperma sp.
Rhaphidophora vitiensis.
Sapota sp. (a few days)
Sapota sp. (a few days)
Scævola floribunda.
Spondias dulcis (a month).
Sterculia sp. (seeds _nil_, fruits months).
Stylocoryne sambucina (2 or 3 days).
Tabernæmontana (orientalis?) (a few days).
Tacca maculata (_nil_ or a few days).
Trichospermum Richii (a few days).
Urena lobata.
Veitchia Joannis.
Veitchia sp.

NOTE 5 (page 14).

THE INLAND FIJIAN PLANTS POSSESSING BUOYANT SEEDS OR FRUITS

They come under the following heads:

(a) Plants of the stream-border or the pond-side or of the inland swamp,
_e.g._, Lindenia vitiensis and Hydrocotyle asiatica. The extension of
the principle by which plants with buoyant seeds or fruits are located,
not only at the sea-side but at the water-side generally, is here
involved, as explained in Chapter III.

(b) Plants following the rule deduced by Schimper for Terminalia, that
when a genus comprises several species possessing buoyant fruits, only
those having fruits with the greatest floating power are found at the
coast, the least buoyant plants occurring inland; examples, Calophyllum
and Guettarda.

(c) Plants that like Ipomœa behave irregularly in respect to
seed-buoyancy, a difference in behaviour often associated with varying
stations both at the coast and inland.

(d) Plants with dehiscent buoyant capsular fruits, like Sterculia, where
dehiscence takes place on the tree and the seeds have no buoyancy.
Although the unopened fruit may float a long time, it does not in that
condition come under the influence of the currents.

(e) Plants like Citrus Decumana, Gardenia, sp., &c., that, although
apparently exceptions to the principle, do not offer much opposition to
it, since the first is most at home at the river-side and the second
often displays a decided inclination for a station at the coast.

(f) Genuine exceptions to the principle, such as Hibiscus Abelmoschus
(see page 21).

NOTE 6 (page 15)

TABLE SHOWING THE DEGREE OF BUOYANCY OF THE SEEDS AND FRUITS OF SOME
INLAND HAWAIIAN PLANTS

(Unless otherwise stated, the seeds or fruits sink at once or in a day
or two)

Acacia Koa.
Aleurites moluccana (1-2 weeks).
Alyxia olivæformis.
Argemone mexicana.
Argyreia tiliæfolia (_nil_ or months).
Bidens pilosa.
Campylotheca sp.
Canavalia galeata.
Capparis sandwicensis.
Cassia Gaudichaudii.
Cassia occidentalis.
Cheirodendron Gaudichaudii.
Colubrina oppositifolia (weeks).
Commelina nudiflora.
Coprosma ernodeoides.
Coprosma sp.
Coprosma sp.
Cyathodes Tameiameiæ (a few days).
Cyrtandra sp. (a few days).
Cyrtandra sp. (a few days).
Cyrtandra sp. (a few days).
Dianella odorata (a few days).
Dracæna aurea.
Eclipta alba (months).
Erythrina monosperma.
Gossypium tomentosum (a week).
Gossypium barbadense (a few days).
Gossypium sp. cultiv. (a few days).
Hibiscus Youngianus (weeks).
Hydrocotyle verticillata (weeks).
Ipomœa bona nox (_nil_ or months).
Ipomœa insularis.
Ipomœa pentaphylla.
Ipomœa reptans.
Ipomœa tuberculata.
Jacquemontia sandwicensis.
Jussiæa villosa (a few days).
Lobeliaceæ (Clermontia).
Maba sandwicensis.
Metrosideros polymorpha.
Mezoneuron kauaiense (pod, a week).
Mucuna urens (months).
Myoporum sandwicense.
Olea sandwicensis, see page 364.
Phyllostegia grandiflora.
Phyllostegia mollis.
Plectronia odorata.
Pritchardia Gaudichaudii (5 or 6 weeks).
Ricinus communis (7-10 days).
Rubus Macraei.
Scævola Chamissoniana.
Scævola Gaudichaudii.
Sida fallax.
Sisyrinchium acre.
Solanum aculeatissimum.
Sophora chrysophylla (pod, 1-2 weeks).
Viola Chamissoniana.
Waltheria americana.

NOTE 7 (page 15)

SOME INLAND HAWAIIAN PLANTS POSSESSING BUOYANT SEEDS OR FRUITS

Three of these, Eclipta alba, Hibiscus Youngianus, and Hydrocotyle
verticillata, frequent wet places, and come under the principle that
water-side plants generally have buoyant seeds or fruits. The buoyancy
of the seeds of Argyreia tiliæfolia and of Ipomœa bona nox varies with
station and may be explained as under Ipomœa in Note 5. The floating
power of the fruits of Colubrina oppositifolia may be akin to that of
inland species of Terminalia as indicated in Note 5, since another
species of the genus C. asiatica, which is a coast plant, has very
buoyant seeds. Mucuna urens was no doubt originally, as it now is in
tropical America, a littoral plant. The buoyant fruits of Pritchardia
Gaudichaudii offer a genuine exception to the principle (see page 330).

NOTE 8 (pages 18, 112)

THE PYRENES OF MORINDA

The pyrenes of the two Malayan inland species of Morinda (M. umbellata
and M. longiflora) examined by Professor Schimper do not possess the
bladder-like cavity to which those of M. citrifolia owe their floating
power, and it is to be inferred from his remarks (p. 183) that they have
little or no buoyancy. The pyrenes of a Fijian inland species, near M.
Grayi, had no floating power as tested by me, and they lacked the
bladder-like cavity.

NOTE 9 (page 18)

THE BUOYANCY OF THE FRUITS OF CALOPHYLLUM

Professor Schimper found that whilst the fruits of Calophyllum
inophyllum, the shore tree, remained afloat after 126 days, those of C.
amœnum, an inland species, sank in from three to fourteen days, both
possessing similar buoyant structures, but to a less degree in the case
of the inland species. This genus presents a parallel case to Terminalia
referred to on page 17; but the general discussion of the subject will
be found in Chapter XIII. According to the above authority C. Calaba, a
West Indian coast tree, has buoyant fruits. The same is also true of the
fruits of a large inland tree in the Solomon Islands experimented on by
me (_Solomon Islands_, p. 305). It would thus appear that the fruits of
the genus are as a rule buoyant, and that, as in Terminalia, the least
buoyant fruits belong to the inland species. Professor Schimper also
shows (p. 182) that the diminished floating power of the fruits of the
inland species is associated with diminution in thickness of the buoyant
seed-shell which is most developed in the buoyant fruits of the strand
species.

NOTE 10 (page 24)

THE BUOYANCY EXPERIMENTS ON BRITISH PLANTS

The experiments in all cases were made to test the floating power of the
seed or fruit in the condition in which it is detached from the plant.
It usually makes very little difference whether sea-water or fresh water
is employed, since in my numerous experiments there were but few
exceptions to the general rule that seeds or seed-vessels that sink in
fresh water sink also in sea-water. This subject is discussed in Chapter
X. However, it may be here observed that the chief effect of the
increased density of sea-water is merely to increase the proportion of
buoyant seeds or fruits in any particular species.

It is necessary in such experiments to imitate Nature as much as
possible. The seed or fruit, as the case may be, must be experimented
upon in the condition in which it falls from the plant, or in the
condition in which it would be ultimately found in river and pond drift.
The seed or fruit should be thoroughly wetted, and air-bubbles removed.

Prolonged drying has but a slight effect on the great majority of seeds
and seed-vessels experimented on, and this is just as true of tropical
plants. Those that sink at once in the mature and fresh condition rarely
float more than a day or two even after drying for a year. The usual
effect is to increase the floating capacity of seeds and fruits already
buoyant, and not to develop the capacity.

The results given in the table refer only to sound seeds. In fresh-water
experiments, in nearly all cases, the seeds ultimately germinate in the
water, and this is the usual cause of the close of the experiment. In an
ordinary collection of floating seed drift from a pond or river,
germination will go on for years at each successive spring, the
postponement of germination being a very striking feature with a fair
proportion of seeds in river and pond-drift. This subject is dealt with
in detail in my paper published in the _Proceedings_ for 1897 of the
Royal Physical Society of Edinburgh.

THE TABLE OF RESULTS OF OBSERVATIONS AND EXPERIMENTS ON THE
BUOYANCY OF THE SEEDS OR SEED-VESSELS OF MORE THAN 300
BRITISH FLOWERING PLANTS

EXPLANATION OF TABLE.—The capacity of floating for months is thus
indicated, ++; of floating for 1 to 4 weeks, +; and where sinking occurs
at once or within a week there is no entry. When buoyancy continued in
my experiments after 6 and 12 months, it is indicated by Roman numerals
(VI and XII). A=an aquatic plant; M=a beach plant; R=a river-side or
pond-side plant; var.=variable in floating power.

Total of the original list: 320 species belonging to 192 genera and 65
families. Of these, about 260 were tested by the author, the data for
the remaining species being mainly derived from the writings of Thuret,
Kolpin-Ravn, and Sernander, with a few from those of Darwin and Martins.

NOTE.—Whilst this work has been going through the press, the author has
added thirteen species, seven genera, and two families to the list above
given; but the general inferences are not affected by the additions. The
corrected total would, therefore, be 333 species, 199 genera, and 67
families.

_On the effect of drying on the buoyancy of seeds and seed-vessels_

It has been already observed that this is as a rule but slight, and that
in the great majority of cases the effect of prolonged drying for many
months, or even for years, is at the most to give a seed or fruit
originally non-buoyant a floating power of a few days’ duration. This is
a subject to which I have paid especial attention in my experiments,
since, of course, much depends on it in the way of dispersal by
currents. It is obvious that a seed or fruit possessing impermeable
coverings at the time of its separation from the parent can scarcely be
compared with one where the coverings only attain their water-proof
capacity by drying. Most gardeners know that seeds which dry easily take
up moisture easily, and the principle applies in a varying degree to the
great majority of seeds and fruits.

Darwin was inclined to attach importance to adventitious buoyancy
acquired by drying; and in the _Origin of Species_ he refers to
instances offered by the fruits of the Hazel (Corylus), the Asparagus,
and Heliosciadium. In Note 48 I have referred to the cases of the Oak
and the Hazel; and, indeed, we have only to examine the beach-drift in
various parts of the world, and to look at their respective stations, to
learn that this is not an effective mode of dispersal. Buoyancy of seed
or fruit is only one of many other qualities that is concerned with
distribution by currents. Nature does not act in this way in
seed-distribution, and there can be little doubt that the author of the
_Origin of Species_ would have been the first to abandon this view, if
his researches had been continued. It should be especially noted that
plants of the sea-beach, where the floating power happens to be _nil_,
or limited only to a week or two, would have derived great advantage
from the drying of their seeds or fruits if it was really effective in
aiding dispersal by currents. However, with plants like Cakile maritima,
Eryngium maritimum, Glaucium luteum, &c., the effect of drying is very
small.

NOTE 11 (page 25)

THE EFFECT OF SEA-WATER IMMERSION ON THE GERMINATING CAPACITY OF SEEDS
AND SEED-VESSELS

Berkeley, Darwin, Martins, and others, long ago established the capacity
of seeds to germinate after prolonged immersion in sea-water. The reader
will find a _resumé_ of their results in the appendix to Mr Hemsley’s
volume on the _Botany of the Challenger Expedition_. The subject is well
illustrated in the original papers of those authors, and in my later
papers on the flora of Keeling Atoll, and on the seed-drift of the
Thames.

I may here remark that the earlier observers often pay more attention to
the retention of the germinating capacity after sea-water immersion than
to the degree of buoyancy. For this reason I have not been able to make
great use of the buoyancy results of Martins, since he frequently does
not distinguish between temporary and long-sustained buoyancy, an
objection also pointed out by Thuret and Hemsley.

NOTE 12 (page 27)

THE BUOYANCY OF THE FRUITS OF GALIUM APARINE

Norman and Sernander (see p. 172) attribute considerable buoyancy to
these fruits on account of the hollow cavity in each. I used to find
them in England in floating river-drift in autumn; and Norman observed
them on the Scandinavian beaches. They do not, however, float long, as
the cavity is open; and in two sets of my experiments they sank within a
few days.

NOTE 13 (page 29)

THE BUOYANCY OF THE SEEDS OF CONVOLVULUS SEPIUM

This plant seeded freely in 1893 in the Lower Thames Valley, as at
Molesey. I kept some of the seeds afloat for thirty-three months, of
which the first nine months were spent in sea-water and the rest in
fresh-water. One seed, at the end of the period, germinated healthily in
the fresh-water.

NOTE 14 (page 26)

OTHER LONG FLOTATION EXPERIMENTS

Whilst keeping my collections of Thames seed-drift in water from year to
year, I obtained a number of records of long “flotations.” Thus in
several cases, as with Bidens cernua and different species of Carex,
germination of the floating fruit took place in the water after a period
of two years. The same is also true of the seeds of Iris pseudacorus and
of the drupes of Sparganium ramosum. The last-named remained afloat in
the vessels, with the seed still sound, after four years; and the fruits
of Carex paludosa germinated afloat after three years in water. Many
drift fruits and seeds did not germinate freely in the vessels until the
second spring, that is, after a lapse of eighteen months; and in those
cases where the experiments were still further prolonged, a few
germinated in the vessels in the third and sometimes even in the fourth
year.

NOTE 15 (pages 33, 280)

THE OCCURRENCE INLAND OF SILENE MARITIMA

Prof. Schimper appeared to be in doubt as to the inclusion of this
littoral plant amongst those found in elevated mountain districts.
However, an interesting note on the occurrence of this plant on the
summit of one of the inland Norwegian mountains is given by Sernander
(p. 405), and is referred to by me on page 280 of this work.

NOTE 16 (page 34)

THE BUOYANCY OF THE SEEDS OR FRUITS OF THE BRITISH BEACH-PLANTS THAT
ALSO OCCUR INLAND

My experiments in the case of Armeria vulgaris, Artemisia, Cochlearia
officinalis, Plantago, the maritime forms of Spergularia rubra with and
without winged seeds, and Silene maritima disclose little or no floating
capacity even after prolonged drying. Thuret obtained similar results
for the Spergularia. It is unlikely that other plants of the group
possess any floating power worth speaking of. As indicated in Note 71,
the fruits of Raphanus maritimus float only for 7 to 10 days.

Nature disperses the fruits of Armeria vulgaris inclosed in the
persistent calyx; but in this condition they float only for 2 to 4 days
in sea-water, and the buoyancy of the capsule and seed is still more
limited. They are sufficiently light to be blown some distance by strong
winds, and the stiff hairs would cause them to adhere to a bird’s
plumage in the case of gulls nesting where the plants grow.

Reference to Matricaria inodora is made under Note 18.

NOTE 17 (page 35)

THE BUOYANCY OF THE SEEDS OR FRUITS OF THE GROUP OF BRITISH LITTORAL
PLANTS THAT FREQUENT SALT MARSHES AND MUDDY SHORES

_Aster tripolium._ The achenes, with or without the pappus, sink in
fresh and salt water in a day or two even after a year’s drying.

_Glaux maritima_ }
_Plantago maritima_ } The small seeds, or the seed-like nucules as
_Samolus valerandi_ } in Suæda, have but little floating power
_Suæda fruticosa_ } even after prolonged drying.
_Suæda maritima_ }

_Salicornia herbacea._ Would be dispersed probably by floating portions
of the plant, which, however, soon break down and the liberated seeds
sink. The floating seedling thrives in sea-water and could be carried
great distances (see Note 19).

_Salsola kali._ I experimented on this plant, both on the coast of
Devonshire and in Chile, with the same results in both localities
whether in the fresh state or after drying for weeks. The fruit sinks,
but when the plant dries the fruit is often detached inclosed in the
perianth, and floats in that condition in sea-water for a few days.
Portions of the plant of various sizes bearing mature fruits all sank
within ten days. It would seem at first sight, from the observations of
Prof. Martins, that the fruits float for several weeks; but his
experiments were mainly directed to testing the powers of germination
after sea-water _immersion_; and it is often not at all clear whether
_flotation_ is implied or even to be correctly inferred. There is a
slight suspicion of germination on the plant. Sea-birds doubtless aid in
the dispersion of this plant; the dry crisp portions of the plant
carrying fruits catch readily in one’s clothes on account of the
prickly-pointed leaves.

_Scirpus maritimus._ The fresh fruits float a few weeks in sea-water in
most cases, but 10 per cent. remain afloat after two months. After
drying for some months 30 per cent. remain floating after two months’
immersion.

} The fruits float a few days or a week. Drying
_Triglochin maritimum_ } somewhat increases the buoyancy. Sir W.
_Triglochin palustre_ } Buller in New Zealand found in the gullet
} of Anas superciliosa, the Grey Duck, numbers
} of the fruits of Triglochin triandrum.

NOTE 18 (page 35)

THE BUOYANCY OF THE SEEDS OR FRUITS OF THE BRITISH LITTORAL PLANTS THAT
ARE CONFINED TO THE BEACH

_Arenaria (Honckeneya) peploides._ The seeds float for many months in
sea-water unharmed, 75 per cent. floating after a year. They never
germinate in sea-water; but on being transferred to fresh water after
many months in sea-water they germinate healthily in a few days. These
seeds only float a few days in fresh water, all sinking within 10 days,
and even after a year’s drying they sink in a week or two. Precisely the
same results were produced in my experiments in 1892 on Cornish seeds,
and in 1904 on Devonshire seeds. In the great contrast between their
floating capacity in sea-water and in fresh water the seeds of this
plant defy the general rule that seeds that float a long time in
sea-water float also a long time in fresh-water. According also to
Sernander the seeds float a long time in the sea. He says that the
capsules float, but since they ultimately dehisce this could scarcely be
efficacious in dispersal. Floating portions of the plant also aid in its
dispersal, according to the same authority (p. 174). The plant forms
great extended masses on the pebbly shores of Spitzbergen (Ekstam, p.
28).

_Beta maritima._ Thuret found that the dried fruits of this plant
floated only two or three days in sea-water; whilst in my sea-water
experiments the freshly gathered fruits floated only one or two days.
Sernander speaks of them as fitted for dispersal from shore to shore;
but this could only be to a limited extent. Martins and Thuret
established by experiment the capacity of the germination of seeds of
other species of Beta after long immersion in sea-water; and the first
seems to imply that those of Beta vulgaris float for many weeks; but I
am inclined to think an error lies here.

_Cakile maritima._ The fruits, even after long drying, float, as a rule,
only a week and sink within ten days, the same results being afforded in
my sea-water experiments in 1893 on fruits from Cornwall, and in 1904 on
fruits from Devonshire. The fruits are common in the stranded drift on
the north coast of Devonshire and may often be seen germinating there.
They are also frequent in the beach drift of the Scandinavian coasts
(Sernander, p. 156).

_Crambe maritima._ The fruits were kept floating by Sernander more than
13 days (p. 165). Martins implies that they floated for 45 days. Darwin
says that they germinated after 37 days’ immersion in sea-water, but
does not specify that they floated all the time.

_Crithmum maritimum._ The ripe fruits readily separate into the two
carpels, which are very buoyant and float in sea-water for months. In my
experiments, 95 per cent. remained afloat after 10 months. It is
remarkable that whilst in sea-water the spongy covering of the carpels
retains its vitality, in fresh-water it becomes sickly and decays and
the carpels lose their floating power, so that they float weeks instead
of months as in the sea-water. The carpels are extremely light, being
washed up in the spray and blown up by the wind amongst the lightest of
the stranded drift of the Devonshire beaches. In a moderate gale they
are often blown off the beach and up the cliff-faces.

_Convolvulus soldanella._ From 40 to 50 per cent. of the seeds float
after six months in sea-water, and about 30 per cent. float after
eighteen months, retaining up to the end their germinating capacity.
Sernander implies that the plant is not found on the Scandinavian coast
to the north of Nissum Fjord in Denmark. It is known, however, to occur
in the south of Scotland. (I am indebted to Mr. Millett for his
extremely kind assistance in experimenting on this plant about ten years
since.)

_Eryngium maritimum._ The fruits float in sea-water, as a rule, only 3
or 4 days and all sink within a week. After drying for three months, the
floating period is only increased by a day or two. Though not at all
suited for transport for any distance by the currents, the carpels, on
account of their long prickly calyx teeth, would readily become
entangled in a bird’s plumage, and doubtless they are dispersed usually
in that fashion.

_Euphorbia paralias._ The seeds float a long time unharmed in the sea.
In my experiments at least 90 per cent. remained afloat after six weeks
in sea-water. On account of their small size they are liable to be
overlooked in beach drift; but they are to be found stranded on the
sands of our southern coasts, and they came under my notice in abundance
in the seed-drift of the Sicilian beaches.

_Glaucium luteum._—The seeds have no proper buoyancy even after
prolonged drying. On account of their oiliness they will float at first
on still water; but they can be made to sink at once or in a day by
dropping water upon them. The mode of dispersal is problematical.

_Lathyrus maritimus._—The seeds are evidently able to float a long time.
They were, according to Sernander (p. 178), found in quantities by J.
Schmidt cast up on some sand-islets near Falster in Denmark; and the
plant is regarded by Norman as distributed over the coasts of Arctic
Norway through the agency of the currents. They have, as observed by
Schmidt, considerable floating powers. Some small leguminous seeds,
seemingly of this species, which I found in the beach drift of
Woollacombe Sands, Devonshire, floated uninjured for many weeks in
sea-water.

_Matricaria maritima,_ maritime variety of M. inodora. The fruits
floated in my experiments unharmed after eight months in sea-water. In
an experiment made some years since on the fruits of the inland form I
noted that they had little or no buoyancy; but it is necessary to repeat
the observation. Sernander (p. 181) supports Norman’s view that these
plants are spread by the currents in Arctic Norway. The fruits occur in
the Baltic sea-drift and also in fresh-water drift. M. inodora is found
on sandy beaches in Nova Zembla. I am inclined to regard the maritime
form from the dispersal standpoint as a distinct species.

_Polygonum maritimum._—I have made observations on this plant in
Devonshire, the Lipari Islands, and the coast of Chile. As in the case
of several other species of Polygonum tested by me the fruits have
little or no buoyancy, but inclosed in the perianth they float three or
four days. The entire plant floats; but portions placed in sea-water
sank within five or six days. Shore-birds can alone explain the wide
distribution of this species.

The structural characters of some of these fruits or seeds are in their
relation to buoyancy discussed on page 115. It may be here observed that
the valuable results obtained by Prof. Martins in testing the
germinating capacity of the fruits and seeds of several of the
shore-plants above mentioned, after long immersion in sea-water, are at
times not to be depended on for the flotation indications, the
persistence of the seed’s vitality being the special purpose of his
research. His negative results as regards germination are not, however,
always conclusive, since the period employed from April to June was
quite insufficient. In many of my experiments seeds after long flotation
in sea-water did not germinate for a year or more afterwards. If his
investigation had been extended, the opinion that the Ranunculaceæ, the
Malvaceæ, and the Convolvulaceæ are apparently least able to resist the
action of sea-water would never have been formed. A very large amount of
evidence now shows that most seeds or fruits that are at all well
protected will germinate after long immersion in sea-water. But all
experiments must be well safeguarded and extended over a year or two.
The necessity of this was long since shown by Thuret. By employing
double sets of seeds he ascertained that in a third of the species
germination failed not only in the case of the seeds immersed in
sea-water, but also in those that had not been placed in sea-water at
all. Future investigators may, however, regard the buoyant qualities of
seeds or fruits with their associated structural characters as offering
now the true line of research. Observers beginning with Berkeley and
Darwin down to the present time have quite established the fact that
seeds as a rule germinate freely after long sea-water immersion.

NOTE 19 (page 35)

ON GERMINATION IN SEA-WATER

During my experiments on the buoyancy of about 270 British plants, about
a fourth of them (including most of those with buoyant seeds or fruits)
were subjected to prolonged immersion in sea-water from periods varying
from six to thirty-three months. If we except plants like Aster
tripolium, Salicornia herbacea, Triglochin maritimum, &c., that live
normally in salt marshes, or on the muddy banks of estuaries, only one
of the whole number, namely, Ranunculus sceleratus, displayed the
capacity of germination in sea-water. Amongst the plants that failed may
be mentioned the following that are confined to the sea-beach—Arenaria
peploides, Cakile maritima, Convolvulus soldanella, Eryngium maritimum,
Euphorbia paralias, Glaucium luteum, and we may here include Crithmum
maritimum of the rocky coasts. Of the beach-plants that also grow
inland, Silene maritima and Spergularia rubra (excepting the form found
on muddy coast flats) likewise failed. Amongst the plants of
miscellaneous inland stations that failed were Atriplex patula, Bidens
cernua, B. tripartita, Calla palustris, several species of Carex both
from dry and wet situations, Convolvulus arvensis, C. sepium,
Hydrocotyle vulgaris, Iris pseudacorus, several species of Juncus,
Lycopus europæus, Mentha aquatica, Ranunculus repens, Rhinanthus crista
galli, several species of Rumex, Scutellaria galericulata, Sparganium
ramosum, &c.

In nearly all the plants that failed to germinate in sea-water the
capacity of readily germinating in fresh water was displayed. The
restraining power of immersion in sea-water was illustrated over and
over again in my experiments. During the course of an experiment seeds
removed from the sea-water vessel and placed directly in a vessel of
fresh water kept beside the other germinated in a few days, whilst those
left in the sea-water never germinated, though often kept there for
months after. It was also noticeable that a previous sea-water immersion
favoured early germination in fresh water. It may be added that most of
the experiments were on floating seeds and seedvessels, though
germination also occurred in the sunken state.

It was ascertained in the exceptional case of Ranunculus sceleratus,
that although germination took place in sea-water, it was only after a
prolonged soaking of months had prepared the way. Of a number of its
seed-like fruits placed in fresh water and in sea-water in April and
kept under the same conditions, those in fresh water germinated freely
in a week or two, whilst those in sea-water did not begin to germinate
until the following October. Whilst the floating seedlings produced by
germination in fresh water grew vigorously and developed roots, those
resulting from germination in sea-water and left in the vessel only
attained a length of four millimetres in two months, developed no roots,
and showed only the first leaf. The sea-water seedlings were pale green,
and in their stout fleshy appearance contrasted greatly with the slender
fresh-water seedlings.

With regard to the germination in sea-water of the plants of the salt
marsh and of the mud-flats of estuaries, the following observations may
be made. With Aster tripolium the seeds germinate readily in sea-water
even when its density is raised by evaporation to 1·040; and I think
that by a carefully graduated series of experiments they could be
induced to germinate in brine. The seeds of Salicornia herbacea
germinate in sea-water more readily than in fresh water; and the
sea-water seedling is much the more vigorous and healthy of the two. I
kept the floating seedlings in sea-water for about ten weeks from the
date of germination, when they had developed the second joint and were
throwing out rootlets. After that, unless placed in salt-mud, they
became sickly and died. The floating seedling can evidently disperse the
species. I found with Spergularia marina, the maritime form of S. rubra,
that seeds of the plants growing on a sandy beach did not germinate in
sea-water, only those from plants growing on muddy coast-flats doing so.
But the sea-water seedlings, unlike those of Salicornia herbacea, but
like those of Ranunculus sceleratus, when left in sea-water did not
thrive. The seeds of Triglochin maritimum, as well as those of T.
palustre, behave very similarly in sea-water, germinating readily, the
liberated seedlings thriving afloat and producing the plumule. The
ultimate test of the capacity for germinating in sea-water seems to lie
in the behaviour of the seedling when left in the sea-water. Unless it
belongs to a characteristic plant of the salt marsh or of the estuary,
like Salicornia, it makes but little attempt at growth whilst afloat in
sea-water, showing no rootlets, though at times developing the plumule.

The germination of seeds in sea-water also attracted the notice of
Darwin; but his results in some respects are scarcely those I should
have looked for (_Gardener’s Chronicle_, May, 1855, and _Journ. Linn.
Soc._, vol. i., p. 130, 1857). Out of the seeds of 87 plants placed in
sea-water to test their capacity of germination when afterwards planted,
in three cases, those of Tussilago farfara, Convolvulus tricolor, and
the garden Orache (Atriplex), the seeds germinated under the water, the
freed seedlings, as with the two first named plants, living in the
sea-water for some time after. Darwin was evidently himself surprised at
these results, and I am quite unable to understand them. In England and
in the tropics I have carried on prolonged sea-water experiments on the
seeds of at least fifteen species of Convolvulus and Ipomœa (including
the beach plants C. soldanella and I. pes capræ) and have never obtained
such a result. The seeds will nearly always germinate well in fresh
water; but in sea-water the process begins, as indicated by the swollen
seed, and then aborts, the embryo dying (see page 83). The seeds of
Atriplex patula, though a long time in sea-water in my experiments, made
no attempt to germinate there. Neither Prof. Martins, who experimented
upon the effects of sea-water immersion on the seeds of nearly 100
plants, including many coast species, nor M. Thuret, who experimented in
sea-water on the seeds of 251 plants, the experiments being in some
cases prolonged for more than a year, make any reference, as far as I
could gather from their writings, to any cases of germination in
sea-water. Darwin’s results, however, are always significant in matters
of dispersal; and perhaps one of my readers will be able to experiment
again on his three plants.

When in Hawaii, I made some observations on the germination of Batis
maritima in sea-water, a plan with which I was also familiar in its home
in the salt-water pools of the coast of Peru. The mature fruits, on
being freed from the parent plant in sea-water, float away, and in from
one to two weeks they break down from decay, setting free the seeds. The
seeds float in sea-water indefinitely, their buoyancy only terminating
with their germination, the first seeds germinating afloat about six
weeks after the breaking down of the fruit, whilst the rest continue to
float in the sea-water during the next three months, some of them
germinating at intervals, and all of them doing so eventually. Strange
to say, although the seedlings remained healthy whilst afloat in the
sea-water, they made no effort either to separate the cotyledons or to
produce a plumule.

NOTE 20 (page 42).

ON THE MAXIMUM HEIGHTS REACHED BY SOME SHORE PLANTS IN THEIR EXTENSION
INLAND IN VANUA LEVU, FIJI

Since they occupy the “talasinga” districts described in the following
note, these shore plants would be expected to extend as high as those
districts extend, namely, to about 1,500 feet above the sea. This indeed
represents their limit excepting in one instance; but many fall
considerably short of this elevation.

_Canavalia obtusifolia_, variety, 700 feet, rare.
_Cassytha filiformis_, 950 feet.
_Cerbera Odollam_, 1,200 feet: 2,600 feet in one exceptional case on
the slopes of Mbatini.
_Colubrina asiatica_, 400 feet.
_Cycas circinalis_, 1,100 feet.
_Derris uliginosa_, 1,000 feet, rare.
_Ipomœa pes capræ_, 1,300 feet.
_Morinda citrifolia_, 700 feet.
_Scævola Kœnigii_, not common inland, and rarely over 100 feet above
the sea; but it may occur miles from the beach, as near Vatu Levoni,
where a few stunted plants were growing five miles from the coast.
_Vitex trifolia_, 1,300 feet, usually more or less unifoliolate and
procumbent.

Unless otherwise stated all the plants above named are common inland, as
also are _Premna tahitensis_, _Tacca pinnatifida_, _Tephrosia
piscatoria_, _Hibiscus tiliaceus_, &c.; but I have made no note of
_Thespesia populnea_ occurring far off the beach.

NOTE 21 (pages 42, 43)

ON THE DWARFING OF SHORE PLANTS WHEN EXTENDING INLAND INTO THE
“TALASINGA” PLAINS IN VANUA LEVU.

_Premna tahitensis_, 9 or 10 feet high at the coast, may here be only 3
feet high. Other trees like _Morinda citrifolia_ become also stunted.
_Cerbera Odollam_, a moderate-sized tree at the coast, may in the
“talasinga” plains be only 4 to 6 feet high, but it here displays
distinct varietal characters. Whilst the shore trees of Cerbera Odollam
have broad leaves (length 3 times the breadth) with obtuse points, and
short, stout flower-peduncles (1-1/2 - 2 inches), the inland or
“talasinga” species has long lanceolate leaves (length 7 or 8 times the
breadth), and long, slender flower peduncles (3 inches). However,
intermediate forms are common, the broad-leaved coast tree approaching
the inland plant and _vice versâ_.

NOTE 22 (page 43)

THE “TALASINGA” PLAINS OF VANUA LEVU, FIJI

Amongst the most conspicuous features of the north and north-west or lee
sides of the large islands of Vanua Levu and Viti Levu are the extensive
rolling plains that extend from the sea-border for some miles inland to
the foot of the mountains. It is to those of the first-named island that
the following remarks strictly apply; but no doubt they will serve
equally well for those of the other island. In the first volume on the
geology of Vanua Levu, reference is frequently made to this subject, and
the reader may profitably look at the remarks there made.

Here the mountain-forests more or less abruptly cease, and we have an
undulating region of grass, reeds, and ferns dotted over with
Casuarinas, Pandanus trees, Cycads, Acacias, and shrubby growths. Though
the list of plants characteristic of these plains is not small, they are
not, as a rule, numerous in any one locality, and the general appearance
is one of aridity. A dry, crumbling soil, often deeply stained by
iron-oxide, is plentifully exposed; and blocks of basic volcanic rocks
in all stages of disintegration are strewn over the surface in many
localities. Rivers, fed by the heavy rainfall of the forested slopes of
the mountains, traverse these regions, but, as a rule, receive no
tributaries; and the districts have, in fact, well earned the name given
to them by the natives of the “talasinga,” or sun-burnt, lands.

The vegetation, though sparse and scanty in comparison with that of the
forests, is sufficiently varied when it comes to be more closely
examined. In one locality we may have extensive tracts covered with
Gleichenia, Pteris, and other ferns of the bracken habit. In another,
tall reeds (Eulalia) and grasses cover large areas. Here, more than one
species of Tacca (T. pinnatifida and T. maculata) thrive. There, the
Turmeric (Curcuma longa) abounds. Trailing over the soil in one place we
notice Ipomœa pes capræ, in another the Yaka (Pachyrrhizus trilobus),
and in another the procumbent unifoliolate form of Vitex trifolia.
Amongst the shrubs and small trees we observe in different localities
the Sama (Commersonia echinata), the Mbulei (Alstonia plumosa—one of the
rubber plants), Mussænda frondosa, Melastoma denticulatum, and Nelitris
vitiensis, the Nunga-nunga. Dodonæa viscosa, found in similar regions in
Australia and New Zealand, abounds in places; and here and there may be
seen species of Hibbertia, another Australian genus. Fagræa Berteriana,
the Mbua tree, grows abundantly in certain districts, as in the Mbua
plains, and Gardenias are at times abundant. One or two characteristic
beach-plants have been already mentioned, and amongst others
particularly frequent in these plains are Cassytha filiformis, Cerbera
Odollam, Morinda citrifolia, and Premna tahitensis.

When these talasinga districts approach the forests, patches of wood
occur at intervals, and we observe here the Candle-nut Tree (Aleurites
moluccana), the Vunga (Metrosideros polymorpha), and the Thau-kuro
(Casuarina nodiflora). Such are some of the botanical features of these
districts; but the reader will acquire a sufficiently correct general
notion of the floral physiognomy of these regions if he bears in mind
their most conspicuous characters, those of an undulating region more or
less covered with ferns, tall reeds, and grass, and dotted over, either
separately or in clumps, with Casuarinas (C. equisetifolia), Screw-pines
(Pandanus odoratissimus), Cycads (C. circinalis), and Acacias (A.
Richii, &c.).

However, the peculiar vegetation of these plains often ascends the lower
slopes of the mountains, reaching to various elevations. In Vanua Levu
it often ceases at 900 or 1,000 feet, but it may only reach to 400 or
500 feet, and, on the other hand, not uncommonly it ascends to as much
as 1,500 feet, the greatest elevation recorded by me being 1,600-1,700
feet in the Sealevu district. It extends miles inland, and where
conditions are suitable it may reach the heart of the island.

Different explanations have been offered of the origin of the peculiar
vegetation of the leeward slopes of these islands. It is, however, a
phenomenon that is presented over much of the globe by islands lying in
the track of regular winds, the weather, or wet, side being densely
wooded, whilst the lee, or dry, side is covered with grass, ferns, and
similar vegetation. The predisposing cause must be climatic; and
although Mr. Horne’s explanation attributing it to the effect of fires
and to a faulty system of native cultivation (pp. 80, 132) may be
doubtless true in certain localities, the influences at work here must
be the same as are at work in other islands and on continental coasts in
other parts of the world.

But for all that it is not easy to give a definite explanation even from
a meteorological standpoint. Those who are interested in this subject
will recall the desert districts of Australia and the dreary sandy
wastes of the coast of Northern Chile and Peru; and they will be
cautious in venturing on a definite explanation even with such
relatively unimportant examples of the same principle as are exhibited
by the islands of Fiji. Dr. Seemann, writing of these “talasinga” plains
(p. xii), remarks that “their very aspect is proof that rain falls in
only limited quantity,” the mountainous backbone of the islands
intercepting, as he holds, much of the rainfall. But the subsequent
observations of Mr. Holmes, at Delanasau, in the “talasinga” district on
the north-west side of Vanua Levu, have shown that there is by no means
a small rainfall in this locality, the average rainfall, for instance,
for the seven years ending December, 1877, being 113 inches, which must
be quite two-thirds or three-fourths of the fall on the weather side of
the island (see p. 215); whilst the average number of days on which rain
fell was 156. The true cause would seem to lie in the excessive dryness
of the air on the lee side of the islands between the rains, and the
whole matter may, perhaps, be one rather for the hygrometer than for the
rain-gauge. I have no comparative data bearing on this point; but Mr.
Holmes, whose observations as here quoted are from Horne’s _Year in
Fiji_, found that the mean relative humidity for 1875 at 1 P.M. was 63,
which is certainly very low for the tropics. I may remark that, as far
as personal experience goes, the climate on the lee side of Vanua Levu
is much more enervating, much less healthy, and the air is far more
“drying” than on the side exposed to the trade-wind.

Geological characters, as I found, explained nothing in this connection,
the “talasinga” vegetation sometimes occurring on basaltic areas, at
other times on the “soapstone” or calcareous mud-stone, and again on
coarser tufaceous rocks. In my volume on the geology of Vanua Levu (p.
57), it is pointed out that the extensive disintegration of the basaltic
rocks, that are exposed on these plains in places, affords evidence of
the great antiquity of these “talasinga” districts in their present
unforested condition. The extent to which these rocks have weathered
downward is remarkable. In some places they are decomposed to a depth of
ten feet and more. The same inference is to be drawn from the occurrence
of fragments of limonite, or bog-iron ore, over these plains, marking as
they do original swampy tracts that, with a few exceptions, have long
since disappeared. Such deposits indicate that these plains have been
for ages in the same condition. ... It may be added that, according to
Mr. Lister and Mr. Crosby, the features of the “talasinga” plains occur
in the Tongan Group on the leeward sides of the islands of Eua and
Vavau.

NOTE 23 (page 43)

SCHIMPER’S GROUPING OF THE INDO-MALAYAN STRAND-FLORA

It is divided into four formations—the Mangrove, the Nipa, the
Barringtonia, and the Pes-capræ. The two last make up my
Beach-formation, the Barringtonia formation comprising the trees,
shrubs, &c., immediately lining the beach, and the Pes capræ including
the creepers and bushes of the beach itself. In the Pacific islands it
is not always easy to preserve this distinction. The Nipa formation
corresponds in some respects with my Intermediate or Transition
formation, lying as it does between the mangrove-belts and the woods of
the interior; but the swamp-palm (Nipa fruticans) that forms it in the
mass is not found in Fiji or, indeed, in the Pacific islands, excepting
the Solomon and Caroline Groups.

NOTE 24 (page 44)

GROUPING OF SOME OF THE CHARACTERISTIC PLANTS OF THE STRAND-FLORA OF
FIJI

(a) _Beach-formation._—Calophyllum inophyllum, Thespesia populnea,
Triumfetta procumbens, Carapa moluccensis, Canavalia obtusifolia, Vigna
lutea, Pongamia glabra, Sophora tomentosa, Cæsalpinia Bonducella, Acacia
laurifolia, Barringtonia speciosa, Terminalia Katappa, Gyrocarpus
Jacquini, Pemphis acidula, Morinda citrifolia, Guettarda speciosa,
Wedelia biflora, Scævola Kœnigii, Cordia subcordata, Tournefortia
argentea, Ipomœa pes capræ, Cassytha filiformis, Hernandia peltata,
Pandanus odoratissimus, &c.

(b) _Mangrove-formation._—Carapa obovata, Rhizophora mucronata,
Rhizophora mangle, Bruguiera Rheedii, Lumnitzera coccinea, Scirpodendron
costatum, &c. (See below.)

(c) _Intermediate or Transition-formation._—Hibiscus tiliaceus,
Heritiera littoralis, Smythea pacifica, Derris uliginosa, Entada
scandens, Barringtonia racemosa, Cerbera Odollam, Clerodendron inerme,
Vitex trifolia, Excæcaria Agallocha, &c.

N.B.—It is not possible to draw a definite line between the plants of
the mangrove swamp and those of the tracts around. Several of the plants
placed in the intermediate formation, such as Heritiera littoralis,
Entada scandens, Excæcaria Agallocha, &c., are just as much at home
amongst the mangroves. In the same way it is often difficult to
distinguish between the Beach and the Intermediate formations, and
plants like Cerbera Odollam, Hibiscus tiliaceus, and Vitex trifolia
belong equally to both.

NOTE 25 (page 47)

THE STRAND-FLORA OF THE TAHITIAN REGION

Drake del Castillo’s _Flore de la Polynésie française_ deals mainly with
the Society or Tahitian Islands, but also with the Marquesas, Paumotus,
Gambier Islands, and Wallis Island. The last-named, however, lies in
Western Polynesia, and is not dealt with in this connection. There is no
reason to believe, judging from the general character of the islands and
from Cheeseman’s memoir on the Rarotongan flora, that the strand-plants
of the islands of the Cook and Austral Groups, which also belong to this
region, differ materially from those of the Tahitian islands proper.
Rarotonga, however, possesses Entada scandens, not recorded as a growing
plant from any other part of East Polynesia, excepting perhaps Mangaia
in the same group.

NOTE 26 (page 48)

THE FIJIAN SHORE-PLANTS NOT FOUND IN TAHITI

Although most of these plants, such as Barringtonia racemosa,
Clerodendron inerme, Entada scandens, Excæcaria Agallocha, Heritiera
littoralis, Smythea pacifica, &c., have fruits that float for months,
and could have reached Tahiti as readily as some of the beach-plants
that have successfully established themselves, there are a few like
Dalbergia monosperma, Derris uliginosa, and Scirpodendron costatum, the
fruits of which only float for weeks, and it is possible that they may
have been unable to reach there.

NOTE 27 (page 49)

THE INTRUDERS INTO THE BEACH-FLORA FROM THE INLAND PLANTS OF TAHITI

Drake del Castillo mentions several, such as species of Boerhaavia, that
could only be occasional intruders; but it is noteworthy that Gardenia
tahitensis appears to be a genuine recruit from inland. The xerophilous
habit of the Pacific Gardenias and their station, usually near the
coast, however, would render this possible.

NOTE 28 (page 52)

THE LITTORAL PLANTS OF THE HAWAIIAN ISLANDS

Column headings:

A: Indigenous.
B: By aborigines in ancient times.
C: By Europeans soon after discovery.
D: Old World.
E: New Worlds.
F: Both Worlds.
G: Large.
H: Small.
I: Float for months.
J: Sink at once or in a few days.
K: Attract frugivorous birds.

+---------------------------+---+---+---+---+---+---+---------------------+
| | Origin. | Distri- | Characters of fruit |
| | | bution. | or seed. |
| +---+---+---+---+---+---+-------+---------+---+
| Species.[5] | |Intro- | | | | Size. |Buoyancy.| |
| | |duced. | | | | [6] | | |
| | +---+---+ | | +---+---+-----+---+---+
| | A | B | C | D | E | F | G | H | I | J | K |
+---------------------------+---+---+---+---+---+---+---+---+-----+---+---+
| Acacia Farnesiana |...|...| + |...|...| + |...| + |weeks|...| ? |
| Cæsalpinia Bonducella | + |...|...|...|...| + |...| + | ... | + |...|
| Calophyllum inophyllum |...| + |...| + |...|...| + |...| + |...|...|
| Cassytha filiformis | + |...|...|...|...| + |...| + | + |...| + |
| Colubrina asiatica | + |...|...| + |...|...|...| + | + |...|...|
| Cordia subcordata |...| + |...| + |...|...|...| + | + |...|...|
|E Cuscuta sandwichiana | + |...|...|...|...|...|...| + | ... | + |...|
| Cocos nucifera |...| + |...|...|...| + | + |...| + |...|...|
|P Gossypium tomentosum | + |...|...|...|...|...|...| + | ... | + |...|
|P Heliotropium anomalum | + |...|...|...|...|...|...| + | ... | + |...|
| Heliotropium curassavicum| + |...|...|...|...| + |...| + | ... | + |...|
| Herpestis Monnieria | + |...|...|...|...| + |...| + | ... | + |...|
| Hibiscus tiliaceus |...| + |...|...|...| + |...| + | + |...|...|
| Ipomœa glaberrima | + |...|...| + |...|...|...| + | + |...|...|
| Ipomœa pes capræ | + |...|...|...|...| + |...| + | + |...|...|
|E Jacquemontia sandwicensis| + |...|...|...|...|...|...| + | ... | + |...|
|E Lipochæta integrifolia | + |...|...|...|...|...|...| + | ... | + |...|
| Morinda citrifolia |...| + |...| + |...|...|...| + | + |...| ? |
| Mucuna gigantea | + |...|...| + |...|...| + |...| + |...|...|
| Pandanus odoratissimus |...| + |...| + |...|...| + |...| + |...|...|
| Portulaca oleracea |...| + |...|...|...|...|...| + | ... | + |...|
| Scævola Kœnigii | + |...|...| + |...| + |...| + | + |...| + |
| Sesuvium Portulacastrum | + |...|...|...|...|...|...| + | ... | + |...|
| Tacca pinnatifida |...| + |...| + |...| + |...| + | + |...|...|
| Tephrosia piscatoria | + |...|...| + |...|...|...| + | ... | + |...|
| Terminalia Katappa |...|...| + | + |...|...| + |...| + |...|...|
| Thespesia populnea |...| + |...| + |...|...|...| + | + |...|...|
| Tribulus cistoides | + |...|...|...| + |...|...| + | ... | + |...|
| Vigna lutea | + |...|...|...|...| + |...| + | + |...|...|
| Vitex trifolia | + |...|...| + |...|...|...| + | + |...| + |
+---------------------------+---+---+---+---+---+---+---+---+-----+---+---+
Footnote 5:

There are three endemic species here included which are preceded by E.
Two species preceded by P are confined to Polynesia. Most of the
plants are at present typically littoral, though often also occurring
inland.

Footnote 6:

All fruits or seeds, an inch or over in size, that could not have been
transported to Hawaii by birds are regarded as large.

NOTE 29 (page 54)

BOTANICAL NOTES ON THE COAST-PLANTS OF THE HAWAIIAN ISLANDS

[The following remarks have been extracted from my journals and
represent some of the field-notes of journeys made in the more
interesting localities.]

(1) _Walk along the Puna Coast, Hawaii, from Punaluu to Hilo_ (Dec. 26,
1896, to Jan. 6, 1897).—For the first two to three miles to Kamehame
Point, the following plants were noticed on the flows of smooth ropy
lava that formed the cliff-bound coast—Capparis sandwichiana,
Jacquemontia sandwicensis, Ipomœa insularis, Lipochæta lavarum,
Portulaca villosa, Tephrosia piscatoria, Tribulus cistoides, Waltheria
americana, &c. Beyond this point Scævola Kœnigii was abundant in places
on the old lava-flows near the sea, and further on patches of Myoporum
sandwicense growing, not as a tree 20 to 30 feet high, as in the
mountains, but as a prostrate shrub with fleshy leaves. Vegetation
similar to that above described occurred on the surface of the old
lava-flows that constituted the cliff-bound sea-border as far as
Kapapala Bay. On the sandy beach at Kapapala Bay grew Ipomœa pes capræ,
serving as host to Cuscuta sandwichiana. In the vicinity of the house at
Keauhou there were a few Coco palms and Pandanus trees, whilst Capparis
sandwichiana and Morinda citrifolia were growing on the adjacent
lava-fields.

Morinda citrifolia and Tephrosia piscatoria grew on the lava flows
between Keauhou and Apua. On the beach at Apua, Ipomœa pes capræ and
Scævola Kœnigii were abundant, the last extending a few hundred yards
inland on the lava. Further east the inland bush, made up of Cyathodes
tameiameiæ, Metrosideros polymorpha, &c., descended to the coast to
within a few hundred yards of the sea. In crossing the lava coast plains
to Kapa-ahu I observed Morinda citrifolia growing frequently out of the
cracks in the bare lava-rock, and an occasional solitary tree of
Erythrina monosperma growing also from the fissures.

Before reaching Kapa-ahu we passed the site of an old coast village,
named Laepuki, where there were growing from forty to fifty Coco-nut
palms, as well as another village, represented by a solitary house, and
named Kamomoa, where there were 27 Coco-nut palms and a few Pandanus
trees. Kapa-ahu, with its numerous Coco-nut palms, was more like a South
Sea coast village than any before seen; and the coast vegetation
suddenly acquired a South Pacific character.

At Pulama, for instance, about a mile west of Kapa-ahu, where the
ancient lava-flows, fairly vegetated, terminate at the sea in cliffs 20
or 25 feet high, there is a curious and quite unexpected development of
a littoral flora such as we should see in the South Pacific. Here,
growing on the broken lava surface at the brink of the cliffs and
overlooking the sea, thrive Cæsalpinia Bonducella, Cocos nucifera,
Ipomœa pes capræ, Ipomœa glaberrima, Morinda citrifolia, Pandanus
odoratissimus, Scævola Kœnigii, Sesuvium Portulacastrum, Thespesia
populnea, and Vigna lutea. This shore-belt of characteristic littoral
plants is backed by vegetation more inland in its character, amongst
which Aleurites moluccana, Dodonæa viscosa, Erythrina monosperma, Ipomœa
insularis, I. bona nox, Osteomeles anthyllidifolia, &c., are to be
observed. Such a shore-belt of typical littoral plants is rarely to be
found in the large island of Hawaii; and its usual position at the
margin of cliffs, and raised 20 or 25 feet above the sea, is rather
suggestive of an uplift in recent times of this part of the coast.

Between Kapa-ahu and Kalapana is a low country occupied mostly by
Guavas, and often turfy. At Kalapana, which is a large village situated
on a grassy plain by the sea, Coco palms and Pandanus trees abound, and
Mucuna gigantea and Cæsalpinia Bonducella are frequent near the coast,
whilst Ipomœa pes capræ is common on the beach. Calophyllum inophyllum
is planted near the houses. Here Osteomeles anthyllidifolia in its
dwarfed form descends to the edge of the cliffs. About half a mile
beyond Kalapana is the hamlet of Kaimu, and here among the Coco palms
close to the beach I noticed four Loulu palms (Pritchardia
Gaudichaudii). Beyond Kaimu the trees and shrubs of the inland wood,
Metrosideros polymorpha, Cyathodes tameiameiæ, &c., descend on the spurs
of old lava-flows close to the coast; whilst Pandanus and Morinda
citrifolia with Mucuna gigantea are common near the sea as far as
Kehena, where there are plenty of Coco palms. I approached Opihikao
through as fine a Pandanus forest as I have ever seen, the large Bird’s
Nest Fern (Asplenium nidus) growing half-way up their trunks, adding
picturesqueness to the scene, whilst Mucuna gigantea was a common
climber. Beyond Opihikao the inland woods descend to the coast. Thence
on to Makuu the coasts are mostly occupied by Pandanus forests, and the
lower coast road from Makuu to Hilo traverses a region where these
Pandanus trees abound, extending far inland. Scævola Koenigii and Ipomœa
pes capræ are common on the coast near Coco-nut Island, Hilo Bay.

It may be added that the agency of the wild goat explains the dispersal
of Myoporum sandwicense, Morinda citrifolia, Tephrosia piscatoria,
Waltheria americana, &c., over the almost bare surfaces of the lava
flows on the Puna coast. Goat droppings were frequent under the patches
of Myoporum and Waltheria. In some of them I found the entire seeds of
Portulaca oleracea and the small cocci of Euphorbia pilulifera, weeds
common in the district.

(2) _Coasts of the Kalae Promontory and its Vicinity, Hawaii._— This is
the most southerly portion of the group, and it is on the eastern coasts
of this district that many of the North American drift logs are embayed
and stranded. At Kamilo, to the east of the promontory, there is a long
beach of calcareous sand where Heliotropium anomalum, Scævola Kœnigii,
and Tribulus cistoides grow in abundance, whilst Sesuvium Portulacastrum
thrives on the beach and in brackish pools. Portulaca lutea (Sol.),
Ipomœa glaberrima (Boj.), and Jacquemontia sandwicensis also occur.
Where the beach-sand has encroached on the adjacent lava surface, the
Scævola covers extensive tracts off the beach, and is stunted. I noticed
a solitary thicket of Thespesia populnea on the beach.

The actual headland of Kalae is wind-swept and covered with grass,
amongst which Portulaca villosa and Sida fallax thrive. By the sea occur
Scævola Kœnigii and Ipomœa pes capræ, and there is some Sesbania
tomentosa near the point. Waiheiaukini beach is shut in between the
lofty arid slopes of the promontory on one side and a modern lava-flow
on the other side. Here Scævola Kœnigii grows in quantity, together with
Ipomœa pes capræ, Tribulus cistoides, Sida fallax, and Jacquemontia
sandwicensis, whilst Cuscuta sandwichiana is abundant, finding its hosts
in the first four plants just named.

(3) _South Kona Coast, Hawaii._—The coast here, as exemplified by that
between Kapua and Hoopuloa, is mostly bare lava. Here and there, a
little coral sand collects amongst the lava blocks of the rubbly shore,
and it is in such places that Scævola Kœnigii and Ipomœa pes capræ find
a home and apparently thrive, whilst Hibiscus tiliaceus and Morinda
citrifolia grow behind. I observed Cordia subcordata and one or two
specimens of Pritchardia Gaudichaudii by the coast on the south side of
Milolii. Around a brackish pool at Kapua I observed Heliotropium
curassavicum, and Acacia Farnesiana was to be seen growing on the beach
at Okoe. On the lava coast between Hoopuloa and Papa, two miles to the
north, Tephrosia piscatoria was very abundant.

(4) _North Kona Coast, Hawaii._—I examined the coast between Kailua and
Kiholo. White beaches are common south of Keahole Point, the coast
further north being usually lava-bound with sandy beaches here and
there. Heliotropium anomalum, Ipomœa pes capræ, and Sesuvium
Portulacastrum are the commonest beach plants on this coast. Scævola
Kœnigii is also abundant in places, whilst Tribulus cistoides and
Morinda citrifolia are also fairly common on the beaches. The Morinda
also grows on the adjacent lava flats; but on both sand and rock it is
evidently usually self-sown, since seedlings are to be seen near the
older plants. Heliotropium curassavicum is to be seen here and there on
the sand all along the coast, but nearly always associated with H.
anomalum. Jacquemontia sandwicensis occurs occasionally on the beach;
and Cuscuta sandwichiana is abundant in places, growing generally on
Ipomœa pes capræ, but sometimes on Scævola Kœnigii. Brackish water ponds
are common on the coast inside the beaches, Ruppia maritima flourishing
in the water, with Sesuvium Portulacastrum growing at the edges.
Sometimes Hala trees (Pandanus odoratissimus) fringe the borders of the
pools. I noticed Pritchardia Gaudichaudii on the coast at Kiholo, and I
learned that Cordia subcordata was once common here as on other parts of
the Kona coast; but it has died out as in most other localities.

(5) _Kohala Coast, Hawaii._—Several littoral plants are scantily
represented on the beach of black sand at the mouth of the Waimanu
valley, especially Ipomœa pes capræ, Morinda citrifolia, Pandanus
odoratissimus, and Scævola Kœnigii. The Pandanus covers the adjacent
precipitous slopes up to a height of several hundred feet above the sea.
Ipomœa pes capræ is abundant on the sand dunes backing the beach at
Waipio. I observed Naias marina in the Waipio River just inside the
mouth. No one seems to have recorded the plant from the group since
Chamisso found it in Oahu.

(6) _Hamakua Coast, Hawaii._—Not many opportunities presented themselves
on this cliff-bound coast of finding littoral plants. At the mouth of a
gulch between Ookala and Laupahoehoe I found growing at the coast Vitex
trifolia (var. unifoliolata) in quantity, together with Morinda
citrifolia, Scævola Kœnigii, and Pandanus odoratissimus, the last-named
clothing the hill-slopes overlooking the sea.

(7) _The Coasts of Oahu._—The littoral vegetation of the south-east
portion of the island from Diamond Head round to Waimanalo is, as a
rule, scanty. Ipomœa pes capræ and Tribulus cistoides prevail to Koko
Head, and on the rubbly coast between that headland and Makapuu Point
occur Tephrosia piscatoria, different species of Lipochæta, &c. Between
Makapuu Point and Waimanalo, Scævola Kœnigii and Vitex trifolia (var.
unifoliolata) are fairly abundant, the former growing on the rocky slope
at the base of the cliffs, and raised perhaps some 20 feet above the
sea. Along the whole east coast of the island the littoral vegetation is
rarely well represented. However, Ipomœa pes capræ is common everywhere,
whilst Scævola Kœnigii occurs frequently, and here and there a few
plants of Morinda citrifolia are seen on the beach, while thickets of
Hibiscus tiliaceus mark in some localities the mouths of streams.

On the north coast of Oahu, as on the Waialua and Waimea beaches, the
one-leaved variety of Vitex trifolia is common, together with Ipomœa pes
capræ and Euphorbia cordata; whilst Acacia Farnesiana is frequent on the
Waialua beach, its pods being much appreciated by the cattle.
Occasionally, as by the bridge at Waimea, Colubrina asiatica and
Thespesia populnea are to be noticed.

Shore vegetation is a little better represented on the beaches at and
near Kaena Point, the north-west corner of the island. Here on the sand
we find often in abundance Heliotropium anomalum, the same variety of
Vitex trifolia, Scævola Kœnigii, and Ipomœa pes capræ; whilst on the
rocks bordering the beach occur Gossypium tomentosum, Jacquemontia
sandwicensis, Tribulus cistoides, Vigna lutea, and more than one species
of Lipochæta, the last being derivatives from the inland flora.

On the west coast of the island true shore-plants play an inconspicuous
part. Ipomœa pes capræ is common on the beaches, and such plants as
Acacia Farnesiana, Jacquemontia sandwicensis, Gossypium tomentosum, and
Tribulus cistoides immediately border the beach. Ipomœa tuberculata is a
frequent intruder as well as the recently introduced Algaroba tree
(Prosopis dulcis). Acacia Farnesiana also extends inland, covering
entire large areas and forming in the Waianae valley extensive thickets
impenetrable for the cattle. It occupies great districts near the coast
in different parts of Oahu, and with Hibiscus tiliaceus is to be found
far inland. The cattle are active dispersers of its seeds. (See Note
30.)

True beach plants are infrequent at the mouth of Pearl Harbour, although
the coast is well suited for them. Here I found Heliotropium anomalum,
H. curassavicum, Jacquemontia sandwicensis, Lipochæta integrifolia (a
true beach plant), Herpestis Monnieria, &c. Batis maritima occurs in one
or two localities around Oahu, but it is, according to Hillebrand, of
recent introduction.

NOTE 30 (page 58)

THE BEACH-DRIFT OF THE HAWAIIAN ISLANDS

It was pointed out by Dole long ago in one of the Hawaiian Club Papers
(1868) that the existing currents bring to this archipelago only huge
pine logs from Oregon, but no tropical fruits; and Hillebrand (p. xiv.)
refers to the driftwood of pine logs from the north-west coast of
America, stranded on the shores of these islands. This drift seems to
collect in quantity in particular localities, as on the south-east coast
of Hawaii between Honuapo and the Kalae promontory (especially on the
Kamilo beach near Kaluwalu) and on the east coast of Oahu; and probably
there are other favourable localities for catching the drift on the
northern shores of Maui and Molokai.

It was on the south-east coast of Hawaii (on the beach at Kamilo and on
the eastern side of the Kalae promontory) that this drift came
particularly under my notice. Here the logs are stranded in abundance,
in sufficient quantity, in fact, to build a town, and they were employed
for building purposes by the manager of the neighbouring sugar-cane
plantation. Several of the logs are of huge size, as much as 4 feet in
diameter; and they are known locally as “white cedar” and “red cedar,”
and characterised as Oregon timber. Some of them are extensively
burrowed by the “teredo” and other boring mollusks. Others recently
stranded are covered with barnacles (Lepadidæ), whilst others that have
lain long on the beach are bare. I have seen these logs occasionally
washed up at Punaluu and at different places on the lava-bound Puna
coast. They apparently first strike the Puna coast, and are drifted
along until they become embayed near the Kalae promontory, and
ultimately stranded. Mingled with them on the beaches Pandanus trunks
occur in number; they evidently hail from those parts of the Puna coast
where Pandanus forests prevail, and thus they indicate the direction of
the drift on the coast of this island. In places there was a
considerable amount of small vegetable _débris_, sometimes partially
concealed by the sand, and containing seeds and fruits in fair quantity.

The following seeds and fruits were collected:—

_Pandanus_ drupes, common; most of them fresh-looking, but a few much
worn.
_Thespesia populnea_, a few seeds.
_Ipomœa pes capræ_, seeds, fair numbers.
_Ipomœa bona nox_, seeds, a few.
_Ipomœa glaberrima_, seeds, a few.
_Argyreia tiliæfolia_, fruits and seeds, a few.
_Strongylodon lucidum_, seeds, a few.
_Cæsalpinia Bonducella_, a single seed.
_Vigna lutea_, seeds, a few.
_Calophyllum inophyllum_, a few fruits.
_Ricinus communis_, a few, the seeds either free or in the cocci, and
often empty or decaying.
_Aleurites moluccana_, seeds, common, none sound, either empty or
containing a rotten kernel: also a single fruit.
One or two seeds not identified.

There was seemingly a total absence of the fruits or seeds of any
littoral plant not found in these islands, such as I was familiar with
in the South Pacific. In the mass this seed-drift could have been
derived from the neighbouring coasts of the island. This is especially
indicated in the cases of the fruits and seeds of Aleurites moluccana,
Ricinus communis, and Argyreia tiliæfolia. The sound seeds of Aleurites
do not float, the buoyant seeds being always empty, or nearly so; and
the presence of the seeds in beach-drift, as explained on page 419, is
due partly to the buoyancy of the empty seed and partly to the decay of
the stranded fruit, the fruits being able to float for a week or two.
So, also, the seeds of Ricinus, whether free or inclosed in the coccus,
do not, when sound, float longer than a week or ten days. The capsules
of the Argyreia can float two or three weeks, whilst the seeds vary in
their behaviour, as observed on page 20. I noticed in places where the
vegetable _débris_ was heaped up and exposed to the sun’s heat, that
some of the Ipomœa seeds were germinating. It is to be remarked that
horse-dung and goat-dung are always common in the beach-drift of these
islands. Seeds are sometimes to be seen in the stranded material; and it
was evident that the droppings of these animals can float for some weeks
before breaking down.... I may add that large sponges, apparently of no
value, are thrown up in quantities on the east side of the Kalae
promontory.

Excepting the pine logs, the only things coming under my notice in this
beach-drift that could be characterised without hesitation as
non-Hawaiian, were two well-worn pieces of acid pumice, less than an
inch in size. One of them was incrusted partially by the tubes of
annelids, and both of them had evidently been drifting about in the
Pacific for a long period, perhaps for years. They were such as occur in
abundance on the beaches of the South Pacific, and, in fact, on all the
shores of the Pacific Ocean, both temperate and tropical. Although I
carefully searched the stranded drift of many beaches in this group, no
other specimens of drift pumice were found.

On different parts of Oahu the beach-drift was always made up of
materials derived from the vegetation of the coast adjacent. Of most
frequent occurrence were the seeds of Ipomœa pes capræ and Vigna lutea,
and the fruits of Scævola Kœnigii, Vitex trifolia, and Pandanus
odoratissimus. In addition, the empty seeds of Aleurites moluccana were
numerous, and there were occasional seeds of Thespesia populnea,
Colubrina asiatica, and Mucuna gigantea. On one beach there were a
number of fruits of Terminalia Katappa, showing but little signs of
ocean travel, and evidently derived from trees in the vicinity. This
tree was introduced by Europeans; but it is not unlikely that in a
generation or two it will become, without man’s aid, one of the
characteristic beach trees of Oahu. It may be remarked that the pods of
Acacia Farnesiana, a shrub now growing abundantly in Oahu near the sea,
are washed up in great quantities on the beaches of the west coast of
this island, and the seeds are to be seen germinating in numbers on the
beach, the seedlings striking into the sand. The pods float unharmed in
sea-water for four or five weeks, but the seeds, when freed, sink.

Although the above evidence gives no indication of tropical drift of
non-Hawaiian origin on the beaches, it is probable, for reasons adduced
in Chapter VIII., that, in the winter, drift may be brought from
tropical America.

NOTE 31 (page 59)

THE INLAND EXTENSION OF THE SHORE-PLANTS OF HAWAII

_Cæsalpinia Bonducella._—According to Hillebrand, this plant, so
characteristic of the littoral floras of tropical regions, grows “in
gulches of the lower plains on all the islands,” no reference being made
to its occurrence on the beaches. It is very rarely to be seen on the
beaches of the large island of Hawaii; but it is to be found on the
lava-bound coasts, and from there it extends inland usually on old
lava-flows for five or six miles, and reaches sometimes considerable
elevations. In one locality I found it at 2,000 feet above the sea (see
page 188).

_Cassytha filiformis._—Though a typical shore-plant in Fiji and other
tropical localities, it is rarely so in these islands. Hillebrand says
nothing of its station. It grows well in the lower open wooded regions,
and is frequently found amongst the blocks of old lava-flows near the
coast.

_Cuscuta sandwichiana._—Unlike its fellow parasite Cassytha filiformis,
this species of Cuscuta, which is confined to this group, never came
under my notice away from the beach; and Hillebrand speaks of finding it
only at the coast (see page 366).

_Ipomœa pes capræ_, as I observed it in the islands of Hawaii and Oahu,
is confined to the beach or to neighbouring sand-dunes. Hillebrand makes
no reference to its occurrence inland. This species in these islands
offers thus a great contrast to its behaviour in Fiji.

_Scævola Kœnigii._—Whilst most at home on the sandy beaches, this plant
is also frequently met with in the island of Hawaii on scantily
vegetated lava-flows near the coast; but I never noticed it more than a
few hundred yards from the sea.

_Tephrosia piscatoria._—Though it may occur on the beach, it is
generally found as described by Hillebrand on the rocky or rubbly ground
at the back of the beach, as well as further inland. It is common on the
old lava-fields of the island of Hawaii near the coast; and, according
to the natives, its seeds are disseminated by the wild goats that
frequent these localities.

_Tribulus cistoides._—Hillebrand observes that this plant is found along
the sea-shore and on the lower plains. I found it most frequently on the
beaches and on the old lava-flows near the sea.

_Vitex trifolia, var. unifoliolata._—It is confined, as Hillebrand
remarks, to the beaches. Neither in Oahu nor in Hawaii did I ever find
it straying inland, which is the more remarkable since this variety, or
one closely similar to it, is one of the most characteristic inland
plants of the Fijian strand-flora.

_Vigna lutea._—This plant was found by me growing on the beaches and in
their vicinity. Hillebrand merely speaks of it as “growing at short
distances from the shore.”

Some of the trees, usually littoral in their station in the tropical
Pacific, which are regarded as having been introduced in early times
into the Hawaiian group by the Aborigines (see Chapter VII.), behave,
nevertheless, quite like indigenous plants in the inland regions and in
the lower levels. This is true, for instance, of Hibiscus tiliaceus and
Pandanus odoratissimus, the last-named forming forests at the sea-board
extending in places far up the mountain slopes. The same, however, may
be said of other plants known to have been introduced since the
discovery of the islands, as in the cases of Cactus Tuna and of Ricinus
communis; and it also applies to Aleurites moluccana, the Candle-nut
Tree, which, although it could only have been introduced by the
Aborigines, now forms forests on the lower slopes of the mountains.

NOTE 32 (pages 19, 112, 165)

THE FIJIAN SPECIES OF PREMNA

I was much interested in the small trees and shrubs of this genus in
Fiji, more especially on account of the relation between the shore and
inland species. This is an Old World genus containing some eighty
species mainly characteristic of tropical Asia and Malaya, and
represented in the South Pacific archipelagoes by two species, one
Premna taitensis or tahitensis, spread over the region and very near P.
integrifolia, an Asiatic species; the other Premna serratifolia, an
Asiatic plant found in Fiji, the Marquesas, and other groups. Without
endeavouring to give a precise value to the Fijian plants, I will merely
describe the prevailing forms, which are, however, connected by
intermediate varieties. These trees, I may add, are known by the same
name in the various Pacific groups, “Avaro” or “Avalo” in Tahiti,
“Alo-alo” in Samoa, “Yaro” and “Yaro-yaro” in Fiji.

The Fijian plants may be thus described.... (_a_) Premna serratifolia,
an inland tree, growing in open woods and on the outskirts of the
forest, 25 to 30 feet high, more or less hairy, leaves coarsely serrated
with long tapering points, putamen prominently tuberculated and
thick-walled.

(_b_) Premna taitensis or P. integrifolia, a low straggling coast tree
or shrub of the beaches, the coral islets, the swampy borders of the
estuaries, and the inland talasinga plains, its usual height being eight
to ten feet, except in the inland plains, where it is dwarfed, and three
to five feet high. It is more or less glabrous, the leaves being
typically entire with obtuse or retuse and mucronate apices. The putamen
is thin-walled and relatively smooth. (_c_) Intermediate forms found
generally in the inland plains or talasinga regions.

_On the Modes of Dispersal._—Speaking generally, the small drupes of
both species float at first, but the soft parts are soon removed by
decay, and the stone is freed. In the case of the coast species, P.
taitensis, the stones float indefinitely and are often found afloat in
rivers. In the case of the inland tree, P. serratifolia, most of the
stones sink at once, whilst the others sink in a few days. It is
probable that currents are one of the effective agencies in distributing
the coast species, but this could not apply to the inland tree. The
fruits of both the inland and the coast species would attract birds, and
the stones would resist injury in their crops. This is the agency
advocated by Prof. Schimper for the shore species, P. integrifolia, of
Indo-Malaya; and fruits referred with a query to this genus were found
in the collection of seeds and fruits obtained by me from the crops of
pigeons in the Solomon Islands (_Bot. Chall. Exped._, Introd. p. 46,
part IV. p. 312).

_On the Cause of the Buoyancy of the Stone or Putamen of the Coast
Species._—This is primarily connected with the empty seed-cavities, the
four-celled stone usually developing only one seed, the other cavities
being empty. This inference was established by the dissection of a large
number of stones, but it will be seen from the table below that
one-seeded stones are also frequent in the case of the inland tree (P.
serratifolia), where they as a rule sink. With either species the
substance of the stone has no floating power, but with the shore
species, on account of the thin-walled stone, the empty seed-cavities
cause it to be specifically lighter than water whilst with the inland
species the walls of the stone are so thick that the empty spaces of the
unfilled seed-cavities do not effect the same result. It may be remarked
that when the coast species grows in the inland plains the buoyancy of
the stone is preserved.

NOTE 33 (page 63)

DE CANDOLLE’S LIST OF PLANTS DISPERSED EXCLUSIVELY BY CURRENTS

Drepanocarpus lunatus; Ecastaphyllum Brownei; Mucuna urens, D.C.;
Tephrosia piscatoria; Hibiscus tiliaceus; Rhizophora mangle; Guilandina
Bonduc, Linn.; Ipomœa pes capræ; Canavalia obtusifolia.

I have experimented on the buoyancy of the fruits and seeds of all these
plants excepting the two first named. In five species the seeds float in
sea-water unharmed for several months. With Rhizophora it is the
floating seedling that disperses the plant. Neither the pods nor the
seeds of Tephrosia piscatoria are suited for dispersal by the currents.

NOTE 34 (page 64)

THE LITTORAL PLANTS OF THE EASTERNMOST POLYNESIAN ISLANDS

Except in the case of Hernandia peltata my authority here is the _Botany
of the “Challenger” Expedition_. Mr. J. H. Maiden gives some further
details of the flora of Pitcairn Island in a more recent paper
(_Austral. Assoc. Rep._, Melbourne, 1901, vol. 8), and Hernandia peltata
is included in his list.

NOTE 35 (page 68)

DISTRIBUTION OF THE LITTORAL PLANTS WITH BUOYANT SEEDS OR FRUITS THAT
ARE FOUND IN THE FIJIAN, TONGAN, SAMOAN, TAHITIAN, AND HAWAIIAN GROUPS

This list probably contains nearly all the species of the Polynesian
region, but it is not implied that these plants have been recorded from
all the groups (_vide infra_).

(a) _Species found only in the Old World._—Calophyllum inophyllum,
Hibiscus diversifolius, Thespesia populnea, Heritiera littoralis,
Kleinhovia hospita, Carapa moluccensis, C. obovata, Smythea pacifica,
Colubrina asiatica, Mucuna gigantea, Erythrina indica, Strongylodon
lucidum, Dalbergia monosperma, Pongamia glabra, Inocarpus edulis, Derris
uliginosa, Afzelia bijuga, Barringtonia racemosa, B. speciosa,
Rhizophora mucronata, Bruguiera Rheedii, Terminalia Katappa, T.
littoralis, Lumnitzera coccinea, Pemphis acidula, Morinda citrifolia,
Guettarda speciosa, Wedelia biflora, Scævola Kœnigii, Cerbera Odollam,
Ochrosia parviflora, Cordia subcordata, Tournefortia argentea, Ipomœa
glaberrima, I. grandiflora, I. peltata, Aniseia uniflora, Clerodendron
inerme, Vitex trifolia, Hernandia peltata, Excæcaria Agallocha, Tacca
pinnatifida, Cycas circinalis, Pandanus odoratissimus, Scirpodendron
costatum.

(b) _Species occurring in both the Old and New Worlds._—Hibiscus
tiliaceus, Suriana maritima, Ximenia americana, Dodonæa viscosa,
Canavalia obtusifolia, C. ensiformis, Vigna lutea, Sophora tomentosa,
Cæsalpinia Bonduc, C. Bonducella, Entada scandens, Gyrocarpus Jacquini,
Luffa insularum, Ipomœa pes capræ, Cassytha filiformis, Cocos nucifera.

(d) _Species found only in Polynesia._—Canavalia sericea, Mucuna
platyphylla(?), Cynometra grandiflora, Serianthes myriadenia, Parinarium
laurinum(?), Premna tahitensis.

_Remarks._—Of these seventy plants there is not one that has not come
within the scope of my observations and experiments. The West Coast of
Africa is included in the American region for reasons given in Chapter
VIII. For the other authorities on the buoyancy of these seeds and
fruits reference should be made to the list given under Note 2 and to
other parts of this work. About one or two of the plants, like Ipomœa
peltata, one scarcely knows whether they are most characteristic of the
coast-flora or of the inland-flora.

NOTE 36 (page 72)

HAWAIIAN PLANTS WITH BUOYANT SEEDS AND FRUITS KNOWN TO BE DISPERSED BY
THE CURRENTS EITHER EXCLUSIVELY OR, AS IN A FEW SPECIES, WITH THE
ASSISTANCE OF FRUGIVOROUS BIRDS

_Colubrina asiatica._—Usually regarded as confined to the Old World; but
since nearly all the species are American, that continent may be
considered as the probable home also of this species. Hillebrand gives
it a locality in the West Indies.

_Dioclea violacea._—Tropical America.

_Mucuna gigantea._—Old World.

_Mucuna urens._—America, and extending to the African West Coast, which
is to be included in the American region of shore-plants.

_Strongylodon lucidum._—Old World.

_Vigna lutea._—Old and New Worlds.

_Cæsalpinia Bonducella._—Old and New Worlds.

_Scævola Kœnigii._—Usually regarded as confined to the Old World, but
according to the synonymy accepted by some authors it is also to be
ascribed to America. The genus is chiefly Australian, and it is possible
that the littoral species may have reached America through the agency of
birds, since all the species of the genus possess fruits that would
attract frugivorous birds.

_Ipomœa glaberrima_ (Boj.).—Old World.

_Ipomœa pes capræ._—Old and New Worlds.

_Vitex trifolia._—Old World. The genus is also dispersed by pigeons.

_Cassytha filiformis._—Old and New Worlds. Like Scævola the genus is
chiefly Australian, and here, also, the fruits of the littoral species
are not only dispersed by the currents, but are known to be also
disseminated by fruit-pigeons.

It is possible that birds may have taken a predominant part in the
dispersal of the species of Scævola, Vitex and Cassytha.

There thus remain nine species for consideration. Of these two are
exclusively American, three are found in both the Old and New Worlds and
four are usually regarded as exclusively Old World plants, but one of
them (Colubrina asiatica) has a fair claim to be regarded as of American
origin. Thus it is quite possible that six out of these nine plants were
brought to Hawaii from America through the agency of the currents.

NOTE 37 (page 78)

ON VIVIPARY IN THE FRUITS OF BARRINGTONIA RACEMOSA AND CARAPA OBOVATA

As observed by me in the Rewa delta, Fiji, there was no external
evidence of such a process in the case of the fruits on the trees; but I
did not pay very special attention to the matter, and it will be
gathered from Chapter XXX. that the initial stage of germination may
show no indication in the appearance of the fruit. More observation is
needed for both species. As indicated in Note 50, the structure of the
seed of Barringtonia racemosa is suggestive of a lost viviparous habit.
With regard to Carapa, Schimper (p. 43) remarks that he has never
observed vivipary; but Miquel, in his _Flora Indiæ Bataviæ_,
particularly speaks of the seeds germinating in the capsule. I think
this is very likely, and that perhaps even the rupture of the capsule
may be partly due to this cause.

NOTE 38 (page 78)

ON THE TEMPERATURE AND DENSITY OF THE SURFACE-WATER OF THE ESTUARIES OF
THE REWA RIVER IN FIJI, AND OF THE GUAYAQUIL RIVER IN ECUADOR

(a) _The Rewa Estuary._—My observations were made mostly in the warm,
wet seasons, from October to January, 1897-99, and generally in the
vicinity of the Roman Catholic Mission. The density varied usually
between 1·000 and 1·010, the water being quite fresh after heavy rains
inland. Though the density was usually greatest at high water, this was
by no means always the case. The temperature of the water in dry weather
varied from 79° to 84° F. With the river in flood after heavy rains it
fell to 75° and 76°. As a rule, the fresher the water the lower the
temperature, but this was not invariable. There was evidence of
super-heating in the estuary, the water there having sometimes a
temperature of 82° or 83°, when the water higher up the river as far as
Viria was two or three degrees cooler, the sea-temperature being 79° to
80°. The average temperature of the water of the estuary during the
season would be 80 to 81°.

(b) _The Estuary of the Rio Guayas, also known, as the Guayaquil
River._—My observations were made in the last week of February and in
the first half of March, 1904. Whilst the sea-temperature a few miles
off the Ecuador coast varied from 76° to 80° F., the water of the
estuary from the mouth up to Guayaquil ranged from 79° to 86°, whilst
rather higher up the river the temperature was about 79° or 80°. The
super-heating of the estuary is thus directly indicated. It was well
marked in the lower part of the estuary during one of my ascents of the
river.

_Surface-temperatures of estuary of the Guayaquil River, March 13, 1904,
11 a.m. to 4 p.m.; tide running up._

Sea-temperature 5-10 miles off the mouth 79·7
Estuary-temperature at the mouth, off Puna 82·7
Estuary-temperature 3 miles above Puna 84·4
Estuary-temperature 15 miles above Puna 86·5
Estuary-temperature 25 miles above Puna 82·5
Estuary-temperature off Guayaquil 81·8

The water of the estuary was, as a rule, cooler with the ebbing tide.

The density of the estuary-water at the mouth opposite Puna during the
two days the ship was in quarantine ranged from 1·004 to 1·016, being
generally about 1·010, and salter with the up-going tide. Off Guayaquil
the water during the ebbing tide was quite fresh and, from an Ecuadorian
standpoint only, potable, whilst at high water it may be a little
brackish. The sea-water has much freer access to the channels in the
mangrove-district at the back of the city of Guayaquil, where at high
water I found the density to be 1·014.

Off Puna, on Feb. 25, I noticed that the surface-current which was
running down the stream was from one to two fathoms deep, whilst below
it was a strong current running up the river which carried my
thermometer up against the surface-current.

NOTE 39 (page 82)

ON THE PACIFIC SPECIES OF STRONGYLODON

Hillebrand in his _Hawaiian Flora_, following Seemann, regards S.
lucidum, Seem., and S. ruber, Vogel, as one species found in Fiji,
Hawaii, and Tahiti, and by the former placed also in Ceylon. Hillebrand
and Seemann are followed by Drake del Castillo as regards the Tahitian
species. Taubert, in his monograph on the Leguminosæ (Engler’s _Pflanz.
Fam._, Teil 3, Abth. 3, 1894), takes the same view of the Polynesian
species and of its wide distribution. However, in the _Genera Plantarum_
and in the _Index Kewensis_, the Asiatic and Polynesian species have
been always kept apart. The two species of the genus mentioned in the
first work are increased to five in the _Index Kewensis_, viz., one in
Fiji (S. lucidum), one in Hawaii (S. ruber), two in Madagascar, and one
in the Philippines.

NOTE 40 (page 88)

PRECAUTIONS IN TESTING SEED-BUOYANCY

Many seeds and fruits require a few hours’ soaking before they sink; and
when small they will rest a long time on the surface of still water, but
a touch with the finger or a drop of water will send them to the bottom.
A few will float a few days (3 or 4) before sinking; but such are
included in the non-buoyant group. Only in rare cases does prolonged
drying increase the period of flotation by more than a few days,
examples being given at the end of the Table of Buoyancy results under
Note 10. Adherent air-bubbles, a common cause of adventitious buoyancy,
must always be removed.

NOTE 41 (page 91)

THE BUOYANCY OF THE SEEDS OF CONVOLVULUS SOLDANELLA IN FRESH WATER AND
SEA-WATER COMPARED

The experiments were commenced at the close of September, 1894, and
covered six months. At the end of this period in Mr. Millett’s
experiment, 56 per cent. of the seeds were afloat in fresh water, and 62
per cent. in sea-water; whilst in my own experiment 72 per cent. floated
in fresh water, and 65 per cent. in sea-water. I was indebted to Mr.
Millett’s courtesy for the seeds.

NOTE 42 (page 96)

ON SECULAR CHANGES IN SEA-DENSITY

Exact data bearing on this subject are not at my disposal; but it would
seem that geologists have formed conflicting conclusions from similar
premises. There is the view that the composition of the ocean water was
very different in early geological periods (_Encycl. Brit._, x., 221);
but I should imagine that the character of the crustacean fauna of those
seas would negative any great divergence from the present condition.
Suess implies that the ancient seas carried the same minerals in
solution that they do now, and it is to be inferred in a similar
proportion (E. de Margerie’s French edition of _Das Antlitz der Erde_,
ii., 343 and 345).

NOTE 43 (page 102)

ON THE MUCOSITY OF SMALL SEEDS AND SEED-LIKE FRUITS WHEN WET

I paid considerable attention to this subject from the standpoint of
dispersal some years ago, and published most of the results in _Science
Gossip_ for Sept., 1894. This peculiar quality of seeds had been noticed
by Dr. Kerner in his _Pflanzenleben_ (vol. i., 1887-91), and was
regarded as illustrating a mode of dispersal of seeds by adherence. As a
rule, such seeds when placed in water become coated with mucus in a few
minutes, or within an hour, and when allowed to dry on feathers they
adhere as firmly as if gummed. I found that this quality is not affected
by prolonged drying, as in the cases of Nepeta glechoma and Salvia
verbenaca, where it was exhibited to the same degree after the seed-like
fruits had been kept from one to three years. I especially tested about
110 British plants that were likely to display this quality, and found
that about a dozen exhibited it in a marked degree, and if to these we
add those plants with seeds that display it to a limited extent so that
they merely become adhesive when wetted, the total would be nearly
twenty. It will be noticed from the list subjoined that the plants
showing marked mucosity belong to twenty genera and to ten families, the
Labiatæ and Cruciferæ predominating. Although in some genera, like
Plantago, there is reason to suppose that the seeds of all the species
would behave in this fashion, it would be wrong to infer that this is
usually the case, six genera being indicated below to which such a rule
would not apply, and doubtless the number could be extended. These
plants in England mostly occur at the roadside, on waste ground, and in
dry meadows. It may be added that although in most cases the seeds
appear in water to emit mucus, “exuded mucilage” being the expression
used in the English edition of Kerner’s work, in some instances, as with
Helianthemum vulgare, there appears to be a dissolving process affecting
the outer seed-covering.

I. _Plants with Seeds or Seed-like Fruits that emit Mucus to a Marked
Degree when placed in Water._

Arabis thaliana, G. Cruciferæ.
Camelina sativa, K. Cruciferæ.
Teesdalia, K. Cruciferæ.
Capsella bursa-pastoris, G. Cruciferæ.
Lepidium sativum, D. Cruciferæ.
Helianthemum vulgare, G. Cistaceæ.
* Viola tricolor (Field Pansy), G Violaceæ.
Linum usitatissimum, D. Linaceæ.
Linum, K. Linaceæ.
* Matricaria chamomilla, K. G. Compositæ.
* Senecio vulgaris, G. B. Compositæ.
Collomia, K. Polemoniaceæ.
Gilia, K. Polemoniaceæ.
* Veronica beccabunga, S. Scrophulariaceæ.
Ocimum basilicum, K. Labiatæ.
Salvia verbenaca, G., &c. Labiatæ.
Salvia, K. B. Labiatæ.
* Nepeta glechoma, G. Labiatæ.
* Dracocephalum, K. Labiatæ.
Prunella vulgaris, G. Labiatæ.
Plantago, K. Plantagineæ.
Plantago major, lanceolata, maritima, G. Plantagineæ.
Luzula campestris. G. Juncaceæ.

_Explanation of Abbreviations._—The capital letter following the name
indicates my authority, which is not necessarily the oldest in each
case: B = Beal; D = Darwin; G = Guppy; K = Kerner; S = Scott Elliot. The
respective works quoted will be found at the end of this volume. The
papers of Darwin quoted will be found in _Journ. Linn. Soc._, “Botany,”
vol. i., 1857, and in the _Gardeners Chronicle_ for 1855.

The asterisk is placed before those genera of which other species
examined by me exhibited no mucosity; these species are Arabis hirsuta,
Viola canina, V. palustris, Matricaria inodora, Senecio aquaticus,
Veronica agrestis, V. arvensis, Nepeta cataria, Dracocephalum
canariensis.

II. _Plants with Seeds or Seed-like Fruits which in my Experiments only
exhibited Mucosity in a Slight Degree, becoming merely “Sticky” or
Adhesive when placed in Water._

Arabis albida, Chrysanthemum leucanthemum, Lamium purpureum
(occasionally), Thymus sp., Juncus bufonius, J. communis, J. glaucus, J.
squarrosus.

III. _Plants with Seeds or Small Fruits that exhibit Adhesiveness in the
Dry State and are apt to stick to one’s fingers._

Adenostemma viscosum, Lycopus europæus, Piper Macgillivrayi, &c. One may
include here also Lagenophora (see page 276) as well as the familiar
instances of Pisonia (page 347) and Boerhaavia (page 356).

NOTE 44 (page 121)

ON THE EFFECTS OF INLAND EXTENSION ON THE BUOYANCY OF THE SEEDS OR
FRUITS OF LITTORAL PLANTS

When in Fiji I experimented on the buoyancy of the following
beach-plants that had extended far into the interior of Vanua Levu, as
will be found described in Note 22. Those tested were Cassytha
filiformis, Cerbera Odollam, Ipomœa pes capræ, Morinda citrifolia,
Premna tahitensis, Scævola Kœnigii, and Tacca pinnatifida. In all but
Cerbera Odollam, where I contented myself with establishing that the
fruits floated buoyantly in sea-water, the experiments were prolonged
for many weeks and often for several months; and in some cases, as with
Ipomœa pes capræ, three or four experiments were made on seeds from
different inland localities. The result was to establish in all cases
that the floating powers were as great with the inland as with the coast
plants of the same species; nor could any structural difference of
importance be noticed. It should be observed that there is every reason
to believe that the “talasinga” plains of Fiji have been occupied by the
intruding beach-plants for many ages.

NOTE 45 (page 122)

TABULATED RESULTS OF THE CLASSIFICATION, ACCORDING TO SCHIMPER’S
APPLICATION OF THE NATURAL SELECTION THEORY, OF THE BUOYANT SEEDS AND
FRUITS OF THE TROPICAL LITTORAL PLANTS ON THE BASIS OF THE STRUCTURAL
CHARACTERS CONCERNED IN BUOYANCY

Column headings:

#: Number.
%: Per cent.
A: Percentage of non-adaptive species.
B: Percentage of adaptive species.

+------------------+-----------------------------+---------------+--------+
| | Classification of species. | | |
| +-------------------+---------+ Proportion of | Total |
| | Non-adaptive. |Adaptive.| non-adaptive | number |
| +---------+---------+---------+ and adaptive | of |
| Region. | First | Second | Third | species. | species|
| | group. | group. | group. | | dealt |
| +----+----+----+----+----+----+-------+-------+ with. |
| | # | % | # | % | # | % | A | B | |
+------------------+----+----+----+----+----+----+-------+-------+--------+
|Pacific Islands | 27 | 40 | 10 | 15 | 30 | 45 | 55 | 45 | 67 |
| | | | | | | | | | |
|Pacific Islands, | | | | | | | | | |
|tropical America, | | | | | | | | | |
|and Indo-Malaya | 28 | 35 | 12 | 15 | 40 | 50 | 50 | 50 | 80 |
+------------------+----+----+----+----+----+----+-------+-------+--------+

NOTE.—If to the last we add the eight British shore plants, the buoyant
fruits of which are described in Chapter XII., three non-adaptive and
five adaptive, we get a proportion of adaptive species for temperate and
tropical regions of fifty-one per cent. This is probably fairly typical
of the world generally; but it must be remembered by the reader that the
author regards them all as non-adaptive. In that case, the table can be
used for the numerical results of the three groups which are based only
on structural characters without reference to any theory.

NOTE 46 (page 124)

ON THE MODES OF DISPERSAL OF THE GENUS BRACKENRIDGEA.

Seed-vessels of this genus found afloat in the New Guinea drift are
described by Mr. Hemsley as having two curved cavities crossing each
other one containing a seed, the other empty. “This empty cavity,” it is
stated “gives the fruit its buoyancy” (_Bot. Chall. Exped._, iii., 289;
plate 54) Dr. Beccari, in the English edition of his _Wanderings in
Borneo_, p. 187, speaks of the closed air-containing cavities in the
seed-vessels, or rather “stones,” of this genus as probably giving them
buoyancy and thus enabling them to be dispersed by currents. He points
out that the fleshy covering of these fruits would also aid their
dispersal by birds. The Italian botanist implies that the two Bornean
species grow in swamps. The Fijian species, as observed by me in flower
in Vanua Levu, grew in the dry talasinga districts bordering the
Mathuata coast, the locality where Seemann found the plant. One of the
most recent accounts of the genus is given by Van Tieghem in his memoir
on the Ochnaceæ in _Ann. des. Sci. Nat. Bot._, tome 16, 1902. According
to him there are nine species, all from Malaya and New Guinea, with the
exception of one in Fiji. Previous authors have also referred to
Queensland and Zanzibar species. However, all the species have a limited
distribution, a fact which plainly assigns to birds the principal share
in the dispersal of the genus.

NOTE 47 (page 125)

ON THE TRANSPORT OF GOURDS BY CURRENTS

Small calabashes or bottle-gourds are not uncommonly to be found
floating in the Fijian estuaries and stranded on the beaches; and I have
also found them in the sea off the coasts. They are usually more or less
globular, 3 or 4 inches across, and are evidently able to float for very
long periods and to carry the seeds unharmed. Most of those I examined
from the drift were dry inside and contained the seeds dried together
into a loose ball about an inch in size. The seeds are not those figured
in Gaertner’s _De Fructibus et Seminibus_, as belonging to Lagenaria
vulgaris, and more resemble those of Cucurbita, but are non-buoyant. One
of these gourds, picked up by me in the sea in Fiji, was placed in
sea-water, and two months later was still floating buoyantly. After
being then kept dry for seven months, it was broken open; and ten of the
seeds were put in soil, two of them germinating in a few days.

In Ecuador gourds similar in size and shape were frequently observed by
me floating in the drift of the Guayaquil River and stranded on the
sea-beaches. The seeds are similarly caked together in a loose mass in
the cavity of the fruit. Their characters indicate that they belong to
another species of gourd; and they differ also from the Fijian seeds in
their buoyancy, some of them in my experiments floating two months and
afterwards germinating.

It has been known since the days of Ström and Gunnerus, two Norwegian
naturalists of the 17th century, that gourds and calabashes are from
time to time stranded with other Gulf-stream drift on the coasts of
Norway. We learn from Sernander that those found are usually worked
calabashes; but he alludes to one that was unworked and contained
several seeds (see Sernander, p. 119).

It is scarcely likely that a seed-carrying gourd stranded on a beach
would be able to establish the plant without the aid of man; but it
seems highly probable that gourds have often been introduced into new
countries by the currents and that man has afterwards cultivated them.
These plants may be contrasted with that remarkable Cucurbit, Luffa
insularum, a genuine littoral plant, the seeds of which, and not the
fruits, are dispersed in the Pacific by the currents (see page 426).

NOTE 48 (page 126)

ON THE USELESS DISPERSAL BY CURRENTS OF THE FRUITS OF THE OAK (QUERCUS
ROBUR) AND OTHER SPECIES OF QUERCUS, AND ALSO OF THE HAZEL (CORYLUS
AVELLANA)

The fruits of different species of Quercus are of not infrequent
occurrence in the seed-drift both of the temperate and tropical regions,
being brought down by the rivers to the sea and then stranded on the
neighbouring beaches. They were amongst the drift gathered by Mr.
Moseley in the open sea, 70 miles off the New Guinea coast (_Bot. Chall.
Exped._, iv., 294). I found them on the beaches of Keeling Atoll where
no oak exists, and on the beaches of the south coast of Java; whilst
Prof. Schimper noticed them among the stranded drift of the Java Sea,
and Prof. Penzig found them stranded on the shores of Krakatoa. They
also came under my notice on the Sicilian beaches and on the Italian
coast at Cumæ. Those of Quercus robur are to be found on the English
beaches and in the autumn drift of the Thames, but they soon sink and
disappear from river-drift. They are referred to by Dr. Sernander as
frozen with other floating seeds in the ice of the Scandinavian rivers;
but he evidently does not regard them as possessing much independent
floating power.

Some years ago the author made a number of experiments on the buoyancy
of the acorns of Quercus robur, and he formed the conclusion that when
freshly collected not more than 4 to 8 per cent. of mature fruits will
float in fresh-water, and not more than about 10 to 12 per cent. in
sea-water, but that in either case they all sink in a day or two.
Immature acorns float much longer, and it is these that mostly figure in
the drift. However, unlike most fruits of little initial buoyancy the
mature fruits gain considerable floating power by drying. Of some that
had been kept for seven months 20 per cent. floated after four weeks in
sea-water and 15 per cent. after 10 weeks.... It may be added that,
according to Thuret, the fruits of Quercus ilex have little or no
floating power.

The buoyancy of the fruits of Quercus is due entirely to the cavity left
by the shrinking of the kernel. I never remember to have found one with
a sound seed amongst the drift in England and Sicily; and I should doubt
much whether those in the tropical drift retain their germinating
powers. But, apart from this, the genus Quercus finds in its own
constitution or habit the greatest obstacle in most species to the
adoption of a littoral station. However, there are exceptional
tendencies displayed by the evergreen oaks; and this is very
significant, since in their xerophilous leaves they possess the
preliminary qualification for a station near the sea. Quercus ilex, it
is well known, shows a partiality for the sea-air, and Q. virens, the
“live oak,” flourishes near the sea in the southern states of America, a
maritime variety being distinguished by botanists. One of the
willow-oaks of America, Q. phellos, which grows in swampy land, also has
a beach variety.

The Hazel-tree (Corylus avellana) must be placed in the same category
with Quercus. I found the empty nuts commonly amongst the stranded drift
of the Sicilian and English beaches. The fruits were also frequently
noticed by Dr. Sernander in the Scandinavian sea-drift; but he says
nothing of their empty condition. Mr. Darwin remarks, in the _Origin of
Species_, that he found that fresh hazel-nuts sank, but that after
drying a long time they floated for ninety days and subsequently
germinated. The floating-power is no doubt due to the cavity arising
from the shrinking of the kernel, and it is to this cause that Dr.
Sernander attributed the slight initial buoyancy observed by him.
However, the hazel, like the common oak, lacks the habit that would fit
it for a station by the sea, and, whatever capacity its fruits may
possess for dispersal by currents, it is quite useless for the spread of
the species.

NOTE 49 (page 131)

ON THE DISTRIBUTION OF IPOMŒA PES CAPRÆ, CONVOLVULUS SOLDANELLA, AND
CONVOLVULUS SEPIUM

Whilst Ipomœa pes capræ is cosmopolitan in the tropical zones,
Convolvulus soldanella is cosmopolitan in both the north and south
temperate zones; but, as might be expected, the two species at times
meet and their areas overlap. Thus, according to Mr. Cheeseman (_Trans.
New Zealand Inst._, xx., 1887), they meet in the Kermadec Islands, in
the South Pacific, in about latitude 30°. From my observations on the
coast of Chile it would seem that C. soldanella in its northward
extension fails somewhere between Valparaiso and Coquimbo, that is to
say, between 33° and 30° S. lat. Gay merely refers to the plant as
existing in North Chile, which in his time would include the coast
between 33° and 24° S. lat. It intrudes within the “thirties” on the
coast of California and is found in Madeira in about 33° N. lat. Ipomœa
pes capræ in its turn extends into subtropical regions, being recorded
from the Kermadecs, as above noted, and from the Bermudas in 32° N. lat.
Owing probably to special physical conditions of the coast, which are
referred to in Chapter XXXII., this plant is evidently limited to the
tropics on the west coast of South America. It did not come under my
notice on the beaches of North Chile, and it is apparently not mentioned
by Gay in his work on the Chilian flora.

Convolvulus sepium, the frequent inland associate of the littoral C.
soldanella over the temperate regions of the globe, belongs to the same
section of the genus (Calystegia). Its extraordinary occurrence by
itself in the island of St. Paul, in the Southern Ocean, about fifty
yards from the shore (_Bot. Chall. Exped._, ii., 153, 264), almost
suggests that we have here a dimorphic species with a littoral and an
inland form; and its existence in the Azores is in this connection very
remarkable. It may be here noted that of three plants raised from seeds
found in the beach-drift near Palermo two had the foliage of C. sepium
and one of C. soldanella. Perhaps one of my readers, in imitation of De
Vries with Œnothera, might be able to settle this point by raising some
hundreds of seedlings from the seeds of the beach species. It is
possible that the relation between these two species of Convolvulus may
be in some respects akin to that between Cæsalpinia Bonducella and C.
Bonduc, two littoral plants that accompany each other over much of the
tropical zone.

The student of dispersal will, however, find some curious gaps in the
distribution of Convolvulus soldanella even in the temperate regions;
and it will be curious to observe how they affect the distribution of C.
sepium. He will have to answer the query of De Candolle:... “Admitting,
if one wishes, that the currents have transported this marine species,
how comes it that it chances to be in the Pacific and in Europe, without
occurring on the east coasts of America and on the east and west coasts
of Africa?” (_Geographie Botanique_, ii., 1056). He will have to explain
why some botanists give C. soldanella a habitat in the tropics, as in
the Indian region. Schimper, who investigated this point, says that he
arrived at no certain result (p. 127). See Notes 13 and 41 and pages 29,
91, for further remarks on these two species of Convolvulus.

NOTE 50 (pages 79, 132)

ON THE STRUCTURE OF THE SEEDS AND FRUITS OF BARRINGTONIA

As regards the fruits and their coverings, the littoral and inland
species of Fiji evidently fall into different sections, the first named
(B. speciosa and B. racemosa) being distinguished by their outer fibrous
husk, to which the buoyancy is due, the last-named (B. edulis and an
undescribed species) possessing a hard stone surrounding the seed, and
here the fruits sink or float only for limited periods.

The fruits of B. edulis have an outer almost fleshy covering, a little
fibrous at the outside, and the hard ligneous “stone,” containing an
edible seed, requires a hammer to break it. They float heavily for three
or four weeks, whereas those of the littoral species float for many
months. In the case of another inland species found by me growing as a
small tree 12 feet high on the slopes of Mount Seatura in Vanua Levu at
an elevation of 1,000 feet above the sea, the seed was similarly
protected by a hard “stone” that could only be broken with an axe, and
the fruit was non-buoyant, with thin and perishable outer coats.

This mountain species of Fiji, which I may name Barringtonia seaturæ,
has the general habit of B. racemosa, with which the natives persisted
in linking it; whilst the fruit and foliage come nearer to those of B.
edulis. The leaves are entire, taper at the base, and have a petiole 1
inch long. The fruits are oblong, at least 3 inches in length, and are
obscurely angled.

It would appear from Schimper’s description (p. 173) that the fruits of
the Malayan Barringtonia excelsa possess both the hard stone-shell of
the inland Fijian species and the dry air-bearing fibrous husk of the
littoral species. This is of special interest, since the tree is both a
coast and an inland species.

The following notes on the structure of the seeds of Barringtonia were
made whilst I was drifting about in my canoe in the creeks of the Rewa
delta in Fiji; and whatever may be their deficiencies they have the
merit of having been written in the home of the plants.... When we cut
across a seed like that of B. racemosa or B. speciosa, we observe that
the different parts of the embryo are indistinguishable, being united
into a homogeneous, firm, fleshy mass. But if we look closely we notice
a central fusiform portion marked out from the surrounding parts by a
faint line, along which a delicate membrane of vascular tissue has been
developed. When “germination” begins, though, as the reader will
subsequently perceive, this term is here hardly appropriate, the real
nature of this singular structure becomes more apparent, as is indicated
in the accompanying figure. The central fusiform portion proves to be
the young plant without cotyledons and growing at either end to form the
root and the stem. The delicate investing membrane becomes thicker and
more apparent as germination proceeds, extending upwards and downwards
with the growth of the stem and root and forming a cortical covering in
either case. The investing fleshy portion of the seed, which is now
separable with the fingers, remains attached to the lower part of the
seedling for some time, being evidently a source of nutriment, and gives
a bulbous appearance to the young plant. Young bulbous plants of B.
racemosa, 1 to 2 feet high, are very common on the edge of Fijian
mangrove swamps where the parent tree thrives. The seedlings of B.
speciosa have the same appearance, but the outer fleshy part of the bulb
is not so thick.

[_To face page 574._

[Illustration: _B. racemosa._, _B. speciosa._]

Diagrams illustrating the structure of the growing seeds of Barringtonia
(two-thirds the natural size). That of B. speciosa represents a seed
removed from a fruit displaying the young plant protruding two or three
inches. That of B. racemosa represents the lower end of the seedling
when the plant is eighteen inches high.

_a_ = the exorhiza.
_b_ = the neorhiza invested by the medullary sheath.

This structure of the seeds of Barringtonia speciosa and of B. racemosa
was for a long time meaningless to me, until one day, whilst seated on
the banks of the Lower Rewa, with a number of the sected seeds and
bulbous seedlings gathered around, I reflected that the fruits of the
latter species that floated past me in the river-drift were nearly
always germinating. This called up “vivipary” to my mind; and as I
looked at the Rhizophora seedlings dangling from the branches of the
mangrove-trees close by, it occurred to me that this seed-structure
might be the result of a lost viviparous habit. One apparently had to
deal here not with an ordinary seed containing an embryo in the midst of
albumen, but with a seed in an arrested stage of germination surrounded
by a body that might perhaps prove homologous with the “cotyledonary
body” of Rhizophora. The process of development that goes on without a
break in Rhizophora, from the fertilisation of the ovule to the
detachment of the seedling from the branch, was here, as I considered,
arrested after germination had begun, but before the protrusion of the
seedling from the fruit. With nearly all plants, as I reflected, there
is a rest-stage of varying length, which might be called the seed-stage.
With the mangrove-genera, Rhizophora and Bruguiera, I had convinced
myself by a long series of observations, the results of which are given
in Chapter XXX., that this rest-stage does not exist. It occurs, I
argued, in Barringtonia, but only after germination has begun, and,
therefore, displaced when compared with the typical seed-stage of most
plants.

In this connection it may be noted that a difference in germinating
behaviour might be expected between the two shore species on account of
their difference in stations, Barringtonia speciosa growing on the sandy
beach, and B. racemosa in the wet ground around a mangrove-swamp. There
is a strong suspicion that the rest-stage in B. racemosa is very short,
though I never found germination in progress on a tree (see Note 37).
There is no doubt, on the other hand, that the rest-stage of B. speciosa
is often, as with most other plants, very long. This, then, was my
lesson from the Barringtonia fruits on the banks of the Rewa, and the
question arose whether this interpretation of these curious
seed-structures accorded with the opinion formed of their nature by
botanists.

Curious seed-structures of this kind must have their significance in the
history of the plant; and on returning to England I looked a little
further into the matter. To follow up this kind of inquiry, however,
would carry me far beyond the limits prescribed for this note, and I
have only treated it here in a tentative fashion. Different botanists of
eminence have paid attention to this subject, amongst them Roxburgh,
Thomson, and Miers (see Dr. T. Thomson in _Journ. Linn. Soc. Bot._, vol.
ii., p. 47, 1858, and Mr. J. Miers in _Trans. Linn. Soc. Bot._, vol. i.,
1880). It would appear that the seed-structure of Barringtonia is also
found in Careya, a genus of the same Myrtaceous tribe, and in Garcinia
and other genera of the Guttiferæ, as well as in other inland plants.

Mr. Miers, after reviewing the opinions of his predecessors, gives the
results of his own investigations. The solid embryo found in
Barringtonia and many other genera consists, he observes, (_a_) of an
external portion, the “exorhiza,” which nourishes the germinating seed
and then dies away; (_b_) of an internal portion, the “neorhiza,” which,
growing at each end, forms the central portion of the stem and root; and
(_c_) the “medullary sheath” of Mirbel, that lies between the two, and
is composed of elementary vascular tissue, which ultimately gives origin
to the wood, bark, and leaves of the stem and yields woody fibre to the
root. The exorhizal portion in some cases, as in Barringtonia
acutangula, splits into four parts during germination. Mr. Miers
compares this seed-structure with that of Rhizophora, employing the same
terms, “neorhiza” for the internal portion which forms the seedling, and
“exorhiza” for the external portion which merely nourishes it. However,
I may add that the exorhizal portion in Rhizophora, as shown in Chapter
XXX., is now regarded as formed by the coalesced cotyledons, and is
termed the “cotyledonary body”; so that by implication the corresponding
part of a Barringtonia seed should be regarded from the same standpoint.

It may be apposite to notice here that Barringtonia racemosa displays
one capacity which does not appear to belong to B. speciosa. The
branches stuck in wet soil throw out roots and establish themselves.
This capacity of vegetative reproduction is turned to account by the
Fijians, who make “live-fences” of this tree in wet localities.

NOTE 51 (page 135)

ON A COMMON INLAND SPECIES OF SCÆVOLA IN VANUA LEVU, FIJI

This is a tall shrub, or small tree, nine or ten feet high, which
corresponds with S. floribunda, Gray, as far as Seemann describes it. It
has small, black, juicy drupes, well suited for dispersal by birds,
having no “suberous” mesocarp as in the shore species (S. Kœnigii), and
no capacity for dispersal by currents. It grows, much like the Hawaiian
inland species, in exposed situations where there is plenty of light, as
on mountain-peaks, at the borders of forests, in open-wooded districts,
and in the plains, and is to be found at all elevations from near the
sea up to the highest mountain summit (3,500 feet) when the station is
suitable. I noticed it on the higher slopes and frequently on the tops
of nearly all the principal mountains that I climbed. It is evident that
birds carry the “stones” from one mountain-peak to another, and no doubt
they explain the presence of the species in Tonga. Dr. Seemann speaks of
it as a beach plant in Viti Levu. The plant familiar to me in Vanua Levu
is only on very rare occasions to be seen as an intruder in the
beach-flora.

NOTE 52 (page 137).

ON THE CAPACITY FOR DISPERSAL BY CURRENTS OF COLUBRINA OPPOSITIFOLIA, AN
INLAND HAWAIIAN TREE

The seeds in my experiments sank within ten days; but they are not
readily detached from the fruit, as in the case of the buoyant seeds of
the littoral species (C. asiatica). The fruits, which may float for a
week or two, break down, as Hillebrand observes, tardily and
imperfectly, and could give but little assistance to dispersal by water.

NOTE 53 (page 141)

ON THE GENUS ERYTHRINA

We have in E. indica a widely distributed littoral species, ranging from
India through Malaya to eastern Australia, and over nearly all the
groups of the Pacific, reaching to Tahiti and the Marquesas, but not
occurring in Hawaii. It is associated in Fiji and Tonga with another
shore-species, E. ovalifolia, Roxb., found also in India and Malaya. I
did not come on the second species in Fiji, and according to Seemann it
is rare. It is possible that there is a genetic connection between the
two; and it is noteworthy that in one case Seemann was uncertain (p.
426) whether the species was E. ovalifolia or only a variety of E.
indica.

In different parts of their areas both these species may be found
inland. This no doubt is to be connected with their occasional
cultivation. The Polynesians who esteem E. indica for its handsome
scarlet flowers and its scarlet seeds often plant it near their houses;
but it is curious that if we look at the pages of Seemann, Horne, and
one or two other botanical authors who have written on the Pacific, we
find no reference to its littoral station, the first-named botanist
merely characterising it in Fiji as occurring “wild or planted.”

However, in various localities in Fiji, as on the shores of Natewa Bay
in Vanua Levu, Erythrina indica thrives as a characteristic beach tree.
Dr. Reinecke speaks of it as widely spread on the Samoan coasts; and the
French botanists refer to it as a tree of the Tahitian beaches. Prof.
Schimper frequently mentions the two littoral species of Erythrina as
amongst the components of the Malayan strand-flora. Dr. Treub, when he
visited Krakatoa in 1886, three years after the eruption, noticed some
young plants of Erythrina growing on the shore; whilst Prof. Penzig in
1897 found that both E. indica and E. ovalifolia had established
themselves on the beach. Mr. Kurz again is quoted by Prof. Schimper (p.
170) as including E. indica amongst the “beach-jungle” of Pegu.

There is abundant evidence in support of the dispersal of the genus by
currents. I have observed the seeds of Erythrina indica on the beaches
of Keeling Atoll. Schimper noted Erythrina seeds amongst the stranded
drift of the Java Sea. Treub remarked young plants of the genus growing
on the shore of Krakatoa three years after the great eruption, and
Penzig places Erythrina indica and E. ovalifolia amongst the
beach-plants brought to Krakatoa through the agency of the currents. The
seeds of E. indica not infrequently came under my observation stranded
on the Fijian beaches and floating in the Rewa estuary; and in an
experiment made in Fiji they still floated after five months in
sea-water. Mr. Hemsley years ago formed the opinion, from the drift
collections at Kew, that the genus was dispersed by the currents. I may
here add in further illustration of this point that Erythrina seeds were
found by me in South America floating in numbers in the Guayaquil
estuary and stranded on the beaches of Ecuador.

It is noteworthy that, unlike some of the other shore-plants, Erythrina
indica has at least three sets of names in the South Pacific. Thus it is
known as Rara and Ndrala in Fiji, Ngatae in Samoa, Futuna, and
Rarotonga, Atae in Tahiti, and Kenae in the Marquesas. The Samoan and
Tahitian name recalls the Burmese name of Ka-thit, whilst the Marquesan
word is suggestive of the Makassar name Kăne or Kanur. The Hawaiian name
of E. monosperma is Wili-wili, which evidently has arisen from the
screw-like movement of the open pod when thrown into the air. The same
name in the form of Wiri-wiri is applied for a similar reason to
Gyrocarpus Jacquini in Fiji. It is possible that the Polynesians have
assisted the dispersal of the coast-species (E. indica); but the
currents could have performed the distribution unaided, and the variety
of aboriginal names is not in favour of human intervention.

With reference to the possible extermination by insects of Erythrina in
Hawaii, it has been before remarked (p. 143) that this would not account
for the survival of an inland species, such as E. monosperma in Hawaii.
However, this species since the occupation of that group by the white
man is on the road to extinction. Dr. Hillebrand observes that the
species was much more common formerly than in his time (1851-1871), a
result evidently due to the ravages of the common tropical mealy bug, a
pest of relatively modern introduction (see Koebele in Stubb’s
_Agricultural Report on Hawaii_). It may be added here that Cordia
subcordata, a littoral tree, had been almost exterminated by the ravages
of a small moth even in Dr. Hillebrand’s time. During my examination of
the coasts of the large island of Hawaii, in 1896-7, I was shown several
places not long before occupied by this tree; and, as indicated in Note
29, it only came under my notice in a few localities.

NOTE 54 (page 145)

ON THE GENUS CANAVALIA

Of the three maritime species, C. obtusifolia, D.C., occurs on beaches
all round the tropical zone. I was familiar with it on North Keeling
Island in the Indian Ocean, in Fiji, and in Ecuador. C. ensiformis,
D.C., is just as widely spread; but it is both inland and maritime in
its station, and except when collecting it in the Solomon Islands I have
had but little acquaintance with it. C. sericea (Gray) is a
characteristic beach-plant in Fiji, but is infrequent. In Rarotonga,
according to Cheeseman, it is a common littoral plant. It was also found
in Tahiti by Banks and Solander, and is seemingly peculiar to the
Pacific islands.

Besides C. ensiformis, the other two shore species may at times be found
inland. Thus it is singular that the French botanists do not, as a rule,
speak of C. sericea as a Tahitian beach plant; and Nadeaud only remarks,
concerning its station, that it frequents the wooded slopes of the
valleys of the interior. In North Keeling Island C. obtusifolia
presented itself to me not only as a beach-creeper, its usual habit, but
as a climber over the branches of the coast trees. In one locality in
Vanua Levu I found a variety of this species growing on a hill a mile
inland and about 700 feet above the sea. On one of the beaches it
approached C. sericea in some of its characters, as in the form of the
calyx and in the hairiness.

Although the seeds of C. obtusifolia have long been known to be
dispersed by the currents, having been found in Moseley’s collections of
floating drift off the New Guinea coast (_Bot. Chall. Exp._, IV, 291),
they displayed remarkable fickleness when experimented on by me in Fiji.
As a rule, however, about 10 per cent. sank at once in sea-water, 50 per
cent. floated after three weeks, and 10 per cent. after twelve weeks. Of
seeds that had been kept three years, 50 per cent. floated after eleven
weeks. The seeds are to be found in numbers amongst the stranded drift
of the Fijian and Ecuador beaches, and I noticed them also afloat in the
Rewa estuary of Fiji.

I tested the floating-power of the seeds of C. sericea in Fiji, and
found that half of them remained afloat after sixty days. On the seeds
of C. ensiformis I have not experimented; but their buoyancy is
indicated by the frequent occurrence of the plant on the Solomon Island
coral islets (Guppy’s _Solomon Islands_, pp. 290, 292, 296), and
probably the Canavalia seeds identified at Kew from my drift collections
on these islets belong to this species. Schimper (p. 166) refers to the
seeds of a Canavalia in Java that were still afloat after ten weeks.
These littoral plants are indebted for the floating capacity of the seed
to the buoyant kernel.

NOTE 55 (page 42 and Note 20)

THE INLAND EXTENSION OF SCÆVOLA KŒNIGII

Scævola sericea (Forst.), a hairy variety of this littoral plant, will
probably prove in some localities to be the inland form of the species.
Dr. Reinecke, who mentions only this variety for Samoa, says that it is
found in very moist ground in river-ravines, and no other station is
referred to. It would seem that both the glabrous and hairy forms occur
in Hawaii. Dr. Seemann speaks of the hairy variety as littoral in Fiji.

NOTE 56 (page 149)

ON THE CAPACITY FOR DISPERSAL BY CURRENTS OF SOPHORA TOMENTOSA, S.
CHRYSOPHYLLA, AND S. TETRAPTERA

(1) _Sophora tomentosa, Linn._—The moniliform pods will float for few
weeks, but it is to the seeds liberated by the breaking down of the pod
that the wide dispersal of this beach-plant by the currents is due. When
experimenting on the freshly obtained seeds in Fiji I found that
four-fifths of them floated after three months in sea-water. With seeds
that had been kept for three years, half floated after twelve months and
retained their sound condition. The seeds owe their floating power to
the buoyant kernel.

(2) _Sophora chrysophylla, Seem._—The dry pods of this Hawaiian mountain
species float between one and two weeks in sea-water, but being brittle
they readily break down and the seeds escape. The seeds have no buoyancy
even after drying for four years.

(3) _Sophora tetraptera, Ait._, from the coast of Chile.—After floating
from ten to fourteen days in sea-water, the dry pods become sodden and
begin to break up, the seeds escaping. Since, however, the pods tend to
decay and break open on the tree they would not be available for
dispersal by currents. Out of a number of freshly gathered seeds all
floated buoyantly after a month in sea-water, when the experiment ended;
and of seeds that had been kept over a year six out of ten floated after
four months in sea-water, two of them germinating afterwards in soil.
Like those of S. tomentosa the seeds possess buoyant kernels to which
the floating power is due. On account of the hardness of the tests the
seeds to ensure rapid germination require to be filed.

NOTE 57 (page 153)

ON THE SPECIES OF OCHROSIA

Schumann distinguishes the following species:

(_a_) O. parviflora, Hensl., widely spread in the Pacific islands.

(_b_) O. compta, Schumann, confined to Hawaii and corresponding to var.
B. of O. sandwicensis as given by Hillebrand.

(_c_) O. borbonica, Spr., synonym O. oppositifolia, Lam., from Mauritius
and Madagascar to Java and Singapore.

(_d_){O. sandwicensis, Gray, of Hawaii. } Both probably varieties of
{O. elliptica, Lab., of New Caledonia.} O. borbonica.

(_e_) O. parviflora, Schumann, of New Guinea, probably identical with O.
mariannensis.

NOTE 58 (page 156)

ON PANDANUS (from Warburg)

(a) _The size (length) of the drupes of endemic species in oceanic
islands._—The drupes of P. reineckei of Samoa are 4-5 cm. (1-3/5 - 2
inches). Those of P. joskei and P. thurstonii in Fiji measure
respectively 6 cm. (2-2/5 inch) and 2-1/2 cm. (1 inch).

Out of about sixteen species in the Mascarene Islands (Mauritius,
Réunion, and Rodriquez) quite half have drupes 2 - 3-1/2 cm.
(4/5 - 1-2/5 inch) in size, whilst they run up to 8 or 10 cm. (3-4
inches), and may be less than a centimetre (2/5 inch).

(b) _The affinities of the Fijian and Samoan species._

NOTE 59 (page 188)

SEEDS IN PETRELS

Darwin, in his correspondence (1859) with Sir Joseph Hooker, refers to
the occurrence of large West Indian seeds in the crops of some nestling
petrels observed by Sir William Milner at St. Kilda (_Life and Letters_,
II, 147, 148). Mr. Charles Dixon in _Ibis_ (1885) refers to Sir W.
Milner’s observation in the case of the Fulmar Petrel (Procellaria
glacialis) and speaks of them as Brazilian seeds brought by the Gulf
Stream, adding that he himself found a nut in the crop of one of these
birds in the same locality. He supposes that the birds pick them up from
the water. Mr. Hemsley very kindly wrote to Sir Joseph Hooker recently
on this point with the object of obtaining some idea of the nature of
the seeds; but after this lapse of time it has not been found possible
to satisfy my curiosity. I live in the hope of their proving to be
Cæsalpinia seeds.

NOTE 60 (page 202)

SCHIMPER ON THE HALOPHILOUS CHARACTER OF LITTORAL LEGUMINOSÆ AND OF
SHORE PLANTS GENERALLY

As a result of extensive microchemical investigations, this eminent
German botanist arrived at the conclusion that plants living on the
sea-shore, or in inland stations rich in chlorides, are able, as a rule,
to store up in their tissues a large quantity of these salts, a capacity
enabling them to live in localities where the subsoil is rich in these
materials. This inference, as shown in his experiments, is just as
applicable to the shore-plants of temperate regions, such as Aster
tripolium, Crambe maritima, and Eryngium maritimum, as it is to such
typical littoral plants of the tropics as Barringtonia speciosa, Ipomœa
pes capræ, Scævola Koenigii, Tournefortia argentea, &c. However, with
the Leguminosæ experimented upon, this capacity of storing up chlorides
was often exhibited but slightly or not at all; and characteristic
Pacific beach-plants, such as Canavalia turgida, Pongamia glabra, and
Sophora tomentosa are especially cited as examples (Schimper’s _Ind.
Mal. Strand-flora_, pp. 140-151; Wolff’s ash-analyses are here quoted).

NOTE 61 (page 215)

METEOROLOGICAL OBSERVATIONS ON THE SUMMIT OF MAUNA LOA

The summit is formed of bare rock and sand, the phanerogamic vegetation
ceasing a couple of thousand feet below. Some low plant-forms doubtless
occur under the moist, warm conditions near the steam-cracks, since
Wilkes mentions his finding a small moss; but with this exception the
surface may be described as sterile.

_Dryness of the Air and Electrical Phenomena._—Wilkes refers to the
association of these conditions more than once in his narrative.
Whenever, as sometimes happened, the dew point could not be obtained
with Pouillet’s hygrometer, electricity was easily excited, and was
developed in large sparks. On taking off the clothes at night, sparks
would appear. As shown in the table subjoined, electrical phenomena were
noticed during the first few days of my sojourn on the summit when the
relative humidity was very low. My red blanket at night crackled in my
hands and emitted sparks, and a glowing line was produced by drawing the
finger along. Whilst the air was in this condition I observed that the
wings of dead butterflies lying on the ground stuck to my fingers
tenaciously like a needle to a magnet. The adhesiveness disappeared when
the excessive dryness gave place to humidity. The physiological effect
on me of the associated dryness and electrical state of the air was
displayed in a hot, dry, sweatless skin (cracking and chapping rapidly),
severe headache and sore-throat, general lassitude, and great
irritability. When the weather changed and the air became humid, these
unpleasant symptoms quickly disappeared.

As a result of these dry conditions on the summit of Mauna Loa,
decomposition does not occur. I found in one place on the top, on the
site of an old camp, the remains of a quarter of beef, the meat fresh
but dried up. From a water-bottle left behind by one of the party and
subsequently restored to him, I learned that the visit had been made in
the previous summer. This non-decomposition seems a little strange,
since, as remarked below, flies and other insects were not infrequent on
the summit. However, as Hann remarks, when speaking of mountain
climates, everything dries much more quickly at great altitudes; animals
that have been shot, or killed by falling, become mummies without
undergoing decay (Schimper’s _Plant-Geography_, 697).... The scorching
power of the sun in a sky usually cloudless, or nearly so, was a trying
feature of my daily experiences; and I found that when I faced it with
unshaded eyes during my walks I suffered from severe pain in the
eyeballs at night.

_Insects on the Summit._—It may seem a strange thing to relate, that in
a region apparently absolutely sterile, the flies and other winged
insects caused me much discomfort in my small tent when I was confined
to it through illness. When lying down one morning I noticed the
house-fly, the blue-bottle, and two or three other flies, small beetles
not over a fifth of an inch in size, a moth, and a wasp. They were no
doubt quite happy in the heat, as the temperature inside was over 80°
F., and the sun’s rays felt almost scorching through the thin duck
canvas. Butterflies (and occasionally large moths) were often observed
flying in a drowsy condition about the summit and were easily caught.
They were fond of fluttering around the steam-holes. In places, numbers
were to be seen dead and dried up on the ground, the detached wings
lying about. In the case of a recently dead butterfly I found its
carcase already attacked by numerous small bugs. The butterflies were
most frequent when there was a fresh southerly breeze, and were
doubtless blown up the slopes from the forests below.

Whymper in his _Travels amongst the Great Andes of the Equator_ gives
many particulars of the occurrence of insects at great elevations. He
noticed beetles, diptera, butterflies, moths, and several other insects
at altitudes of 15,000 to 16,000 feet. At 16,500 feet he obtained a
small bug of the genus Emesa. He quotes Humboldt and Bonpland as showing
that insects are transported into the upper regions of the atmosphere
16,000 to 19,000 feet above the sea, and he remarks that the
transportation of insects by ascending currents of air has occasionally
been observed in operation. These facts bear directly on the dispersal
of insects.

_The Winds._—My tent, which was pitched near the middle of the western
border of the crater, happened to be situated in the battle-ground of
the northerly and southerly winds, in a region of gusty winds, fitful
airs, and dead calms. The northerly winds were usually from N.-N.N.W.
and the southerly winds from S.W.-S.S.W., easting in either case being
rarely observed, the northerly winds rather prevailing at night. As a
result of this location miniature whirlwinds were frequent in the
vicinity of my tent, which carried sand into the air and more than once
threatened to lift up my tent bodily and carry it off into the crater
below. At the north end of the crater-border north-easterly winds
prevailed, and at the south end southerly winds occasionally showing
easting. When on one occasion I walked round the crater-margin, a fresh
south-easterly wind prevailed at most parts of the circumference except
in the vicinity of my camp, where there was a light S.S.W. wind both at
8 a.m. and 6 p.m. when I started and returned. The local character of
the winds was often displayed in my walks. On one occasion, having left
my camp, where a southerly wind was blowing, and walked half a mile to
the north, I found a bitterly cold N.N.E. gale in my face which so
impeded my progress that I returned to my camp where the same southerly
breeze continued.

Commodore Wilkes was encamped on the east side of the crater, and there
(December and January) he experienced strong south-west winds, on at
least three days having the force of a gale. These are the prevailing
winds in this season over the group; whereas in August, the time of my
sojourn, south-westerly winds are quite out of season, this being in the
midst of the period of the N.E. trades.

It will be gathered from the foregoing remarks that the mere record of
the winds is insufficient for the purpose of obtaining any definite
notion of the air-currents at this elevation (13,600 feet). It is to
close observation of the clouds that we must look for data of
importance.

_The Clouds._—The clouds on the summit of Mauna Loa were an unending
source of interest to me, and I will give briefly the results of my
observations. The highest clouds were wispy cirri, often arranged as in
a mackerel sky, and evidently at a great altitude. They were only
observed on four or five days. (The lower clouds are indicated in the
accompanying diagram.) Below them and at no great height above the
mountain were to be not infrequently observed isolated woolly clouds
that were carried in a few minutes across the sky and had a brief
existence, often forming and melting away as one gazed at them. Next,
there was a heavy bank of cumulus, which formed on the south-west slope
near the top of the mountain, from which lines of cloud extended along
each flank. Lowest of all was a broad belt, or rather a sea, of cumulus
that was developed on both sides of the mountain about one-third way
down its slopes, and during the day-time isolated the peak from the
world below. It is with the last two cloud formations that we are most
concerned, and I will first describe the sea of cumulus.

The sea of cumulus, as in the case of similar cloud-formations of most
other isolated mountains, when viewed from above, as from the
mountain-top, presents a cloud-field of dazzling whiteness, sparkling in
the sun. Seen from below, as from the coast, it has the dark lowering
appearance of the rain-cloud and indicates the rain-belt. Disappearing
during the night, this broad belt begins to form again between 8 and 9
a.m., and by 10 or 11 a.m. the lower regions are completely hidden and
the mountain’s summit, cut off from the world, rises above the level of
the sea of clouds like an island in an Arctic ocean. As the day
progresses the clouds become more compact and dense. The usual altitude
of this broad belt of cloud is between 7,000 and 8,000 feet. This level
is indicated by the burying of the Kohala mountains, which rise to a
height of 5,500 feet in the distant north-west corner of the island, and
by the usual emergence of the highest summit of Hualalai, which rises,
still nearer, to an elevation of 8,275 feet. On some days, however, it
attains a height of nearly 9,000 feet. On such occasions the highest
peak of Hualalai kept reappearing and disappearing during the day, but
the distant summit of Haleakala in East Maui, 10,032 feet in elevation
and 80 miles away, was always visible.

Words fail to describe the magnificent aspect of this sea of cloud which
shuts off the spectator from the world below. From the summit of the
mountain he gazes down on its surface lit up by a sun shining in a
typically cloudless sky. At one time it appears as an undulating Arctic
land covered with snow of dazzling whiteness. At another time it looks
like a hummocky frozen Polar sea sparkling in the sunshine. Through
occasional rifts, however, one can discern a dark dismal region of mist
and rain-cloud beneath. Miss Bird, who passed a night on the summit in
June, 1874, well describes this sea of cloud in her book on the Sandwich
Islands as “all radiance above and drizzling fog below.”

[_To face page 585._

[Illustration: Diagram illustrating the prevailing cloud-formations of
Mauna Loa during August, 1897.]

The heavy bank of cumulus, that forms at noon on the south-west slope at
an altitude of 10,000 to 13,000 feet above the sea, and sometimes rises
above the mountain, is one of the most conspicuous of the
cloud-phenomena on the summit of Mauna Loa. Apparently extending from
it, but in reality moving towards it, are two lines of small cumuli that
follow the same level along either flank above the sea of cumulus, as is
indicated in the accompanying diagram. It was observed by Wilkes in
mid-winter, 1840-41, but at a lower level. “Clouds would approach us (he
writes) from the south-west when we had a strong north-east trade wind
blowing, coming up with their cumulus front reaching the height of about
8,000 feet, spreading horizontally and then disappearing.” During my
sojourn this bank formed a very striking feature in the landscape during
the early afternoon. On two or three occasions when I visited the south
side of the summit and descended for about a thousand feet I passed
through this bank, being then exposed to a driving mist coming up the
slopes from the south-west. Though its upper surface viewed from a
distance is dazzling white, below it is dark and nimboid.

It is to an updraught of warm moist air on the south or south-west
slopes of the mountain, and to the prevailing cool north-east trade that
strikes the north side of the summit, that we must look for the
explanation of the development and situation of this bank. Although the
trade-wind is markedly stronger than the south-west updraught, some of
the warm, moist, southerly air-currents find their way, as shown by the
observations at my camp, along the sides of the summit, and a line of
condensation is produced where they come into contact with the cool air
of the north-east trade as it sweeps past the flanks of the mountain.
Sometimes at my camp, when there was a light southerly breeze blowing, I
have noticed the line of small cumuli moving south along the mountain
side towards the bank of cumulus.... I may remark that on a few days a
small bank of cumulus formed under similar conditions on the north-west
side of the summit.

From my study of the clouds I arrived at the conclusion that there were
three prevailing air-currents on the summit of Mauna Loa:

(1) The updraught of warm moist air on the south and south-west slopes
of the mountain.

(2) The north-east trade wind, the upper limit of this air-current being
probably not far above the summit.

(3) An upper air-current from the south-east (E.S.E.-S.S.E.), which,
from the velocity of the clouds it carried, was often probably not over
a couple of thousand feet above the summit. It may be observed that on
the coast at the base of the southern slope of the mountain in the
middle of September, when the wind was N.E. and carried the lower clouds
with it, the upper clouds were, on several occasions, noticed travelling
in the opposite direction, namely, from the south.

The volcano was quiescent during my visit and could have exercised but
little influence on the air-currents.

_The Shadow of the Mountain._—Every morning and evening, in clear
weather, for about twenty minutes after sunrise and before sunset, the
shadow of the mountain was thrown back against the sky of the opposite
horizon. It seemed as if some Titanic brush, at work in the sky far
away, had painted in the profile of the mountain with a very uncanny
blue. At sunset the peak was the last to disappear. Commodore Wilkes,
who only records it once, namely, at sunset on the 1st of January,
describes it as “a beautiful appearance of the shadow of the mountain
projected on the eastern sky ... as distinct as possible, its vast dome
seemed to rest on the distant horizon.” This phenomenon is, of course,
well known in the case of other isolated mountains. According to
Murray’s _Handbook of Southern Italy_ (1892), the correct thing for a
visitor to Stromboli is to make an early ascent of the cone to observe
“the very curious triangular shadow of the mountain cast by the rising
sun upon the _sea_.” Unfortunately I neglected my opportunity when on
the island. The shadow of the mountain is also one of the sights of
Etna, a dark-violet, triangular shadow (Baedeker) being thrown at
sunrise over the surface of West Sicily, that is, on the _land_. I saw
the shadow but imperfectly outlined, as the weather was not favourable
at the time of my ascent. When at Nicolosi, on the south slope of Etna,
I noticed at sunset a faint shadow of the mountain thrown against the
eastern sky. I gathered from a short conversation with Prof. Ricco, the
director of the Catania Observatory, when I told him of the shadow of
the Hawaiian mountain, that the interest lay in its projection against
the sky. It is doubtless akin to the spectre of the Brocken and other
mountain spectres.

_Some Previous Meteorological Observations on Mauna Loa._—.... Mr.
Douglas, the botanist, who was subsequently found dead in a cattle-pit
on Mauna Kea, spent a day on the summit of Mauna Loa in the middle of
January, 1834. He mentions that a little way below the top the
thermometer fell at night to 19° F. The wind on the top was N.W. The air
at 11.20 a.m. was 33°, the hygrometer registering 0·5. He remarks that
the great dryness of the air was evident without the assistance of the
hygrometer (_Hawaiian Spectator_, vols. I and II, 1838-9).

Commodore Wilkes, in vol. IV of his _Narrative of the United States
Exploring Expedition_, gives the following observations on the
temperature and winds on the top of Mauna Loa between Dec. 23, 1840, and
Jan. 13, 1841. Those on the temperature are incomplete, but they give a
fair idea of the prevailing conditions. The degrees are in Fahrenheit’s
scale.

Dec. 23, 1840: elevation, 13,190 feet; 3 p.m., 25° F.; strong S.W.
gale; night temperature, 15°.
Dec. 24, 1840: summit (13,600 feet); night minimum, 22°.
Dec. 26, 1840: summit (13,600 feet); violent S.W. gale; night min.,
17°.
Dec. 27, 1840: summit (13,600 feet); sunrise temp., 20°; night min.,
17°; wind, S.W.
Dec. 29, 1840: summit (13,600 feet); noon temp. in shade, 47°; night
min., 20°.
Dec. 30, 1840: summit (13,600 feet); noon temp., 55°; night min., 13°.
Dec. 31, 1840: summit (13,600 feet); night min., 17°.
Jan. 2, 1841: summit (13,600 feet); sunrise, 20°; wind, N.E.
Jan. 3, 1841: summit (13,600 feet); night min., 17°.
Jan. 4, 1841: summit (13,600 feet); daylight, 20°.
Jan. 8, 1841: summit (13,600 feet); S.W. gale.
Jan. 10, 1841: summit (13,600 feet); night temp., 16°.
Jan. 12, 1841: summit (13,600 feet); night temp., 17°.
Jan. 13, 1841: summit (13,600 feet); strong S.W. wind.

The usual variation of temperature in the twenty-four hours is given as
17°-50°. The south-west was evidently regarded as the prevailing wind,
and the clouds are spoken of as sometimes moving from opposite
directions towards the same centre.

When Miss Bird spent a night on the summit of Mauna Loa during the
eruption of June, 1874, the cold was described as intense, eleven
degrees of frost (21° F.).

_Observations on the Summit of Mauna Kea._—.... When Prof. Alexander
with a party of scientists ascended this mountain (in the summer of
1892), the thermometer at night fell to 13° F., and the trade-wind was
found to be blowing as strongly on the summit as down below (Whitney’s
_Tourist Guide to Hawaii_). It is to be inferred that the party camped
by the small lake which is a few hundred feet below the actual summit
(13,800 feet). This lake, which I visited on May 20, 1897, is about 120
yards across, and evidently shallow, probably not more than three or
four fathoms deep. A carpet of algæ covered the bottom. At noon, by the
lake, the air in the shade was 53° F., whilst the temperature of the
surface-water was 51°. The lower clouds were moving from S.S.E. This
lake is said to be permanently frozen over in the winter, and to have
been visited by skaters.

_Permanent Water Supply on the Summit of Mauna Loa._—In this barren
rocky region water derived from the winter-snow is to be found all the
year through at the bottom of the deep cracks or fissures in the
lava-rock. Such fissures are from two to four feet wide, and in the case
of that near my tent the bucket had to be lowered to a depth of
seventeen or eighteen feet to reach the water, or rather the ice, since
it was often necessary to break the surface ice. In these deep, narrow
fissures, which the sun scarcely penetrates, the water would probably be
frozen over all through the seasons; but in those of less depth it would
remain liquid in summer.

REGISTER OF OBSERVATIONS ON WIND, RELATIVE HUMIDITY, CLOUD, RAIN, AND
TEMPERATURE, MADE BY H. B. GUPPY ON THE SUMMIT OF MAUNA LOA AT AN
ELEVATION OF 13,500 FEET ABOVE THE SEA, AUGUST 9TH TO 31ST, 1897.
(CAMP ABOUT MIDDLE OF WEST SIDE OF CRATER MARGIN)

+-----+-------+----------+----------+----------+----------+----------+----------+------------------+--------------+
| | | | | | | | | Air in shade. | |
| | | | | | | | | | |
|Date.|Obser- | 12-4 | 4-8 | 8-12 | 12-4 | 4-8 | 8-12 |------------------+ Remarks. |
| |vation.| A.M. | A.M. | A.M. | P.M. | P.M. | P.M. | | | | |
| | | | | | | | | Min.| Max.|Range.| |
+-----+-------+----------+----------+----------+----------+----------+----------+-----+-----+------+--------------+
|9 |Wind |S.S.W- |Variable | ... |W.S.W.- |N.N.W.- |N.N.W.- | F. | F. | F. |A beautifully |
| | |S.W. 2 | | |W.N.W. 3 |N. 1 | N. 3 | | | |coloured lunar|
| |Rel. | ... | ... | ... | ... | ... | ... | 27·5| 61·2| 33·7 |halo at 1 A.M.|
| | hum. | | | | | | | | | |Electrical |
| |Cloud | 0 | 0 | 0 | 0 | 0 | 0 | | | |condition of |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | |the atmosphere|
| | | | | | | | | | | |(see text). |
| | | | | | | | | | | | |
|10 |Wind | ... |N.N.W. 2 |S.S.W.- |S.S.W.- |Calm | ... | | | |Electrical |
| | | | |W.S.W. 3 |W.S.W. 2 | | | | | |condition of |
| |Rel. | ... | ... | 34 | 42·5 | 46 | ... | 33·5| 59·7| 26·2 |the |
| | hum. | | | | | | | | | |atmosphere. |
| |Cloud | ... | 0 | 1 | 2 | 0 | ... | | | |Faint lunar |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | |halo at 8 P.M.|
| | | | | | | | | | | | |
|11 |Wind |Calm |W.S.W. 1 | ... | ... |S.S.W.- | ... | | | |Electrical |
| | | | | | |S.W. 1 | | | | |condition of |
| |Rel. | ... | ... | {28·5 | ... | ... | ... | | | |the |
| | hum. | | | {21 | | | |}22·5| 61·2| 38·7 |atmosphere. |
| |Cloud | 0 | 0 | 0 | 1 | 0 | ... |} | | | |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | | |
| | | | | | | | | | | | |
|12 |Wind |Calm |Calm |N.N.W.- |N.N.W. 1 |N.N.W.- |N. 2-4, | | | |Earth tremors.|
| | | | |N. 2-3 | |N. 2-4 |S.W.- | | | |Total rain, |
| | | | | | | |W.S.W. | | | |10/100. At |
| | | | | | | | 3-5 | | | |sunset, wind |
| |Rel. | ... | ... | 45 | 79 | 89 | ... | 23 | 54·7| 31·7 |N.W., wet |
| | hum. | | | | | | | | | |canvas of tent|
| |Cloud | 0 | 0 | 5 | 10 | 10 | 5 | | | |froze hard. At|
| |Rain | 0 | 0 | 0 |Rain |Rain |Rain | | | |10 P.M., |
| | | | | | | | | | | |strong |
| | | | | | | | | | | |southerly |
| | | | | | | | | | | |wind, canvas |
| | | | | | | | | | | |thawed, rain |
| | | | | | | | | | | |with strong |
| | | | | | | | | | | |gusty wind |
| | | | | | | | | | | |until 4 A.M., |
| | | | | | | | | | | |when wind |
| | | | | | | | | | | |less. |
| | | | | | | | | | | | |
|13 |Wind |S.W. |S.W. 3-5 | S.W.- |S.W. 3 |Calm; | ... | | | |Earth tremors.|
| | | 4-6 | | S.S.E. | |N.W. 1 | | | | |Total rain |
| | | | | 3-4 | | | | | | |10/100. |
| |Rel. | ... | ... | 86 | 86 | 78 | ... |} | | | |
| | hum. | | | | | | |}33·6|48·7 |15·1 | |
| |Cloud | ... | 10 | 7 | 10 | 9 | ... |} | | | |
| |Rain |Rain |Rain | 0 |Rain |Rain | 0 | | | |Butterflies |
| | | | | | | | | | | |flying about |
|14 |Wind |N.N.W.- |Northerly,|N.N.W. 3 |N.N.W. 2 |N.N.W. 1 | ... | | | |in a semi- |
| | |N. 1 |3 | | | | | | | |torpid state, |
| |Rel. | ... | 61·5 | 47 | 42 | 45·5 | ... |} | | |and easily |
| | hum. | | | | | | |}32·5|52·2 |19·7 |caught with |
| |Cloud | 3 | 0 | 0 | 0-2 | 0 | ... |} | | |the hand. |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | | |
| | | | | | | | | | | | |
|15 |Wind |Northerly,|W.S.W. 1, |S.W.-W. |Calm; |Calm; |N.N.E.- | | | |Wind fitful |
| | |1 |N.N.W. 1 |2 |Southerly,|N.N.W. 1 | 1-2 | | | |during day; |
| | | | | |1 | | | | | |north-westerly|
| |Rel. | ... | ... | 38 | 44·5 |{62 | ... | | | |and south- |
| | hum. | | | | |{52·5 | |}28 |}54·7| 26·7 |westerly airs |
| |Cloud | 0 | 0 | 0 | 7 | 7-0 | 0 |} | | |with calms. |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | | |
|16 |Wind | Calm |N.N.W. 2 |N. 3 | ... |Southerly,| ... | | | |Carefully |
| | | | | | |1 | | | | |observed the |
| |Rel. | ... | ... | {32} | ... | 61 | ... | | | |shadow of the |
| | hum. | | | {28} | | | |}26 | 53.2| 27.2 |mountain |
| |Cloud | ... | 0 | 0 | 7 | 0 | ... |} | | |which, at |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | |sunrise and |
| | | | | | | | | | | |sunset, is |
| | | | | | | | | | | |projected |
| | | | | | | | | | | |against the |
| | | | | | | | | | | |opposite |
| | | | | | | | | | | |horizon. |
| | | | | | | | | | | | |
|17 |Wind | ... |N.N.W. 1 |N.N.W. 2, |N.- |N.N.W. 1 | ... | | | |Fitful |
| | | | |Southerly,|N.N.E 3, | | | | | |northerly and |
| | | | |2 |Southerly,| | | | | |southerly |
| | | | | |2 | | | | | |winds causing |
| |Rel. | ... | ... | 32 | 32·5 | ... | ... |} | | |miniature |
| | hum. | | | | | | |}20·5|58·7 |38·2 |whirlwinds |
| |Cloud | ... | 0 | 0 | 1 | 0 | ... |} | | |that carried |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | |dust and paper|
| | | | | | | | | | | |up into the |
|18 |Wind |Calm |S.W. 1 |Variable | ... |S.W. 2 | ... | | | |air. |
| |Rel. | ... | ... | 26 | ... | 47 | ... |} | | | |
| | hum. | | | | | | |}23 | 58 | 35 | — |
| |Cloud | 0 | 0 | 0 | 1 | 0 | ... |} | | | |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | | |
| | | | | | | | | | | | |
|19 |Wind | ... | ... |N.N.W. 1, |Northerly,|Northerly,|Southerly,| | | |Through the |
| | | | |W.S.W. 1 |3, |2, |3 | | | |day, fitful |
| | | | | |S.S.W. 3 |Southerly,| | | | |northerly and |
| | | | | | |2 | | | | |southerly |
| |Rel. | ... | ... | 20 | 23 |{24 } | ... | | | |breezes. |
| | hum. | | | | |{35·5} | |}22 | 58·7| 36·7 | |
| |Cloud | ... | ... | 0 | 0 | 0 | 0 |} | | | |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | | |
| | | | | | | | | | | | |
|20 |Wind |Southerly,|Northerly,|Northerly,|S.S.W.- |Northerly,|Southerly,| | | |Fitful |
| | |2 |1, | 1, |S.W. 4 |2, |3 | | | |northerly and |
| | | |Southerly,|Southerly,| |Southerly,| | | | |southerly |
| | | |1 |1 | |1 | | | | |airs, often |
| |Rel. | ... | ... | 26·5 | 25 | 29·5 | ... |} | | |reversing |
| | hum. | | | | | | |}22 |57·2 |35·2 |several times |
| |Cloud | 0 | 0 | 0 | 0 | 0 | 0 |} | | |in a few |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | |minutes. |
| | | | | | | | | | | | |
|21 |Wind |Southerly,|Southerly,|S.S.W. 3-4|S.S.W. 4-5|S.W. 4-5 |S.W. 4 | | | |At camp, |
| | |4 |3-4 | | | | | | | |strong |
| |Rel. | ... | ... | 35·5 | 31 | 47 | ... |} | | |southerly |
| | hum. | | | | | | |}26·5|53·7 | 27·2 |winds all |
| |Cloud | ... | 0 | 0 | 0 | 0 | ... |} | | |day. At 7 |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | |A.M., walked |
| | | | | | | | | | | |half-mile |
| | | | | | | | | | | |north and |
| | | | | | | | | | | |found a |
| | | | | | | | | | | |bitterly cold |
| | | | | | | | | | | |N.N.E. gale |
| | | | | | | | | | | |blowing there,|
| | | | | | | | | | | |which forced |
| | | | | | | | | | | |me to return |
| | | | | | | | | | | |to camp where |
| | | | | | | | | | | |the south wind|
| | | | | | | | | | | |still blew |
| | | | | | | | | | | |freshly. |
| | | | | | | | | | | | |
|22 |Wind | ... |S.S.W. 1 | ... | ... |Calm, | ... | | | |Walked round |
| | | | | | |Southerly,| | | | |the crater |
| | | | | | |2 | | | | |from 8 A.M. to|
| |Rel. | ... | ... | 58 | ... | ... | ... |} | | |6 P.M. |
| | hum. | | | | | | |}20·5|46·7 |26·2 | |
| |Cloud | ... | 0 | 0 | 1 | 0 | ... |} | | | |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | | |
| | | | | | | | | | | | |
|23 |Wind | ... |S.S.W. 2 | ... |N.N.W.- |Southerly,|Calm | | | |A few drops of|
| | | | | |N.E. 1-3, |1 | | | | |rain at 2 P.M.|
| | | | | |Southerly,| | | | | | |
| | | | | |3 | | | | | | |
| |Rel. | ... | ... | {60 | 52·5} | 70 | ... | | | | |
| | hum. | | | {54·5 | 64 } | | |}24 | 50·7| 26·7 | |
| |Cloud | ... | 0 | 3 | 3-8 | 3-6 | 0 |} | | | |
| |Rain | 0 | 0 | 0 |Rain | 0 | 0 | | | | |
|24 |Wind | ... |Northerly,|S.W. 1 |N.N.W.- |N.N.W.- |Northerly | | | | |
| | | |3 | |N. 2-3 |N. b. E. |2 | | | | |
| | | | | | |2-3 | | | | | |
| |Rel. | ... | ... | 54·5 | ... | 53 | ... |} | | | |
| | hum. | | | | | | |}20 | 52·7| 32·7 | |
| |Cloud | ... | 0 | 0-1 | 4 | 0 | ... |} | | | |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | | |
| | | | | | | | | | | | |
|25 |Wind |Calm | ... |Calms |Calms |Northerly |Northerly | | | |A few drops of|
| | | | |with N.W. |with N.W. |and |and | | | |rain at 3 P.M.|
| | | | |and S.W. |and S.W. |Southerly |Southerly | | | | |
| | | | | airs | airs |airs |airs | | | | |
| |Rel. | ... | ... | {42·5 | 73·5} | 61 | ... |}17 | 52·2| 35·2 | |
| | hum. | | | {37·5 | 61·5} | | |} | | | |
| |Cloud | 0 | 0 | 0-1 | 2-9 | 0 | 0 | | | | |
| |Rain | 0 | 0 | 0 |Rain | 0 | 0 | | | | |
| | | | | | | | | | | | |
|26 |Wind |Northerly,| N.N.W. 1 | ... |N.N.W.- |N.N.W. 2 | ... | | | |Descended |
| | |2 | | |N. 3 | | | | | |through the |
| |Rel. | ... | ... | 49·5 | ... | 64 | ... |} | | |bank of |
| | hum. | | | | | | |}19·5| 53·7| 34·2 |cumulus on |
| |Cloud | 0 | ... | 3 | 3 | 0 | 0 |} | | |S.W. slope and|
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | |found driving |
| | | | | | | | | | | |mist coming up|
| | | | | | | | | | | |the slope from|
| | | | | | | | | | | |S.W. |
| | | | | | | | | | | | |
|27 |Wind |Calms |Calms |S.W.- | W.S.W.- |Southerly |Calm | | | |Rain not |
| | |with light|with light| W.S.W. 2 | W. 1 | | | | | |measurable. |
| | |airs |airs | | | | | | | |Rain-clouds |
| |Rel. | ... | ... | 35 | 73·5 | ... | ... |} | | |poured into |
| | hum. | | | | | | |}18·5| 50·7| 32·2 |and filled the|
| |Cloud | 0 | 0 | 4 | 10 | 4 | 0 |} | | |huge crater. |
| |Rain | 0 | 0 | 0 |Rain | 0 | 0 | | | | |
| | | | | | | | | | | | |
|28 |Wind |Calms |Calms |N.N.W.- |S.S.W. 3 |Northerly,|Calms | | | |9 A.M., high |
| | |with |with |N. 2-3 | |3 |with | | | |stationary |
| | |northerly |variable | | | |variable | | | |cirrus; at |
| | |airs |airs | | | |airs | | | |noon, solar |
| |Rel. | ... | ... | {46·5 | 64·5} | | | | | |halo; in |
| | hum. | | | {59·0 | 73·0} | ... | ... |}15 | 49·7| 34·7 |afternoon, |
| |Cloud | 0 | 0 | 5 | 7 | 0 | 0 |} | | |nimbus partly |
| |Rain | 0 | 0 | 0 |Rain | 0 | 0 | | | |filling crater|
| | | | | | | | | | | |causing a |
| | | | | | | | | | | |rainbow there;|
| | | | | | | | | | | |a few drops of|
| | | | | | | | | | | |rain at 4 P.M.|
| | | | | | | | | | | | |
|29 |Wind |N.N.W.- |N.N.W.- |N.- |N.N.W. 3, |Calms with|Northerly,| | | | |
| | |N. 3 |N. 2 |N. b. E. 3|S.S.W. 2 |variable |1 | | | | |
| | | | | | |airs | | | | | |
| |Rel. | ... | ... | 45 | 44 | 60·5 | ... |} | | | |
| |hum. | | | | | | |}21·5| 48·7| 27·2 | |
| |Cloud | 0 | 0 | 0 | 0 | 0 | 0 |} | | | |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | | |
| | | | | | | | | | | | |
|30 |Wind |Calms with|Southerly,|N.N.W. |N.N.W. 3 |Calms with| ... | | | | |
| | |northerly |1 |2-3 | |variable | | | | | |
| | |airs | | | |airs | | | | | |
| |Rel. | ... | ... | 32·5 |{41·0} | ... | ... | | | | |
| | hum. | | | |{54·5} | | |}18 | 50·7| 32·7 | |
| |Cloud | 0 | 0 | 0 | 0 | 0 | 0 |} | | | |
| |Rain | 0 | 0 | 0 | 0 | 0 | 0 | | | | |
| | | | | | | | | | | | |
|31 | ... | ... | ... | ... | ... | ... | ... | 18·5| 50·0| 31·5 | |
+-----+-------+----------+----------+----------+----------+----------+----------+-----+-----+------+--------------+

_Method of Observation employed by the Author on the Summit of Mauna
Loa._—My camp was placed near the middle of the west margin of the
crater about 13,500 feet above the sea. The instruments employed were a
Sixe’s maximum and minimum thermometer made by Negretti and Zambra,
several unmounted thermometers, and a reference thermometer (with a Kew
certificate) by the above-named makers, which was used as a standard.
The freezing point was also tested for all the instruments on the summit
in melting powdered ice. The maximum air observations and those on the
relative humidity were taken in a small cave with a hole in the roof,
through which there was a steady flow of air. One day was occupied in
comparing the cave-observations with those obtained under a temporary
screen rigged up outside my tent, the only difference shown being as a
rule less than a degree. The minimum observations taken in my tent,
where there was no artificial heat, were usually only 1·5° higher than
those given by a thermometer outside the tent.

_Results of the Observations on the Top of Mauna Loa, Aug. 9-31, 1897_

Mean minimum temperature of air in shade 23·2° F.
Mean maximum temperature of air in shade 53·8°
Mean daily range of temperature 30·6°
Lowest reading 15·0°
Highest reading 61·2°
Mean temperature for the period 38·5°

} Many observations
Mean relative humidity, 8-9 a.m., 44·5 % } included which
Mean relative humidity, noon ... 43 % } are not given in
Mean relative humidity, 5-6 p.m., 56 % } the register.

On Aug. 11th, at 10 a.m., wet bulb, 33·2°; dry bulb, 52°; difference,
18·8°.

On Aug. 19th, at 11 a.m., wet bulb, 35·7°; dry bulb, 56°; difference,
20·3°.

Owing to the varying winds at my camp, the relative humidity fluctuated
greatly in a short time. Thus, on Aug. 12 it was 46% at noon, and 79% at
2 p.m.

_Average Cloudiness (10 indicating a Sky completely Overcast)_

12-4 A.M. | 0 | Cloudless during 12 out of 13 days
4-8 A.M. | 0 | Cloudless during 19 out of 20 days
8-12 A.M. | 1·3 | Cloudless during 13 out of 22 days
12-4 P.M. | 3·5 | Cloudless during 6 out of 22 days
4-8 P.M. | 1·5 | Cloudless during 17 out of 22 days
08-12 P.M. | 0 | Cloudless during 11 out of 12 days

The winds at the camp were extremely variable and local from north and
south, usually light, with force 1-3: see under Winds and Clouds in the
text.

Rain fell on six days, total 30/100 of an inch: but on four of the days
it was not measurable.

NOTE 62 (page 222)

ON THE RELATIVE PROPORTION OF VASCULAR CRYPTOGAMS IN FIJI

According to Seemann’s work, where about 617 indigenous flowering plants
and about 195 ferns and lycopods are enumerated, the vascular cryptogams
would form about 24 per cent. of the whole flora. (All weeds and
cultivated plants are here excluded.) The vascular cryptogams, however,
seem to figure too prominently in Seemann’s collections. From Horne’s
data, who says that he added 363 flowering plants to the flora, the
flowering plants would amount to about 980; and since Baker implies, in
_Trimen’s Journal of Botany_, 1879, that Horne added 42 species of ferns
and lycopods to the flora, this would increase the vascular cryptogams
to 237, which enables us to estimate the relative proportion of vascular
cryptogams in Fiji as about 20 per cent. of the whole flora of vascular
plants. This is probably near the truth.

NOTE 63 (page 222)

ON THE TABLE OF VASCULAR CRYPTOGAMS OF TAHITI, HAWAII, AND FIJI

In the case of Tahiti, I have gone carefully through the list given by
Drake del Castillo, comparing it with other Polynesian lists given by
Seemann, Horne, Hillebrand, Hemsley, &c., and have reduced his endemic
species from 19 to 13. The same thing has been done with Hillebrand’s
list for Hawaii, some of his species having been found in other parts of
Polynesia, thus reducing the endemic species from 75 to 70. The data
relating to Fiji are referred to in Note 62.

NOTE 64 (page 223)

ON THE DISTRIBUTION OF THE TAHITIAN FERNS AND LYCOPODS

I have arranged them as follows, according to the distributions given by
Drake del Castillo:—Cosmopolitan, 5; Tropics of Old and New Worlds, 33;
Tropics of Old World, mainly Indo-Malaya, 58; “Océanie,” including
Australia, 17; Polynesia, 26; South America, 2; peculiar to Tahiti, 13:
total, 154.

Out of 141 non-endemic Tahitian species, 107 at least have been recorded
from the Fijian area comprising Samoa and Tonga, and 42 from Hawaii. Of
the last, all but four occur also in Fiji. There is thus a very small
element peculiar to Hawaii and Tahiti alone. Some of them will no doubt
be found in the Fijian area; whilst two of them, Acrostichum squamosum
and Lycopodium venustulum, are high-mountain forms in Hawaii and Tahiti,
which have evidently failed to find a suitable elevation in Fiji.

NOTE 65 (page 225)

DISTRIBUTION OF SOME OF THE MOUNTAIN FERNS OF HAWAII THAT ARE NOT FOUND
EITHER IN FIJI OR TAHITI (mainly from Hillebrand)

NOTE 66 (page 226)

ENDEMIC GENERA OF FERNS IN HAWAII

Hillebrand gives two genera of ferns peculiar to Hawaii, one, Sadleria
of Kaulfuss, “scarcely distinct from Blechnum,” and containing four
species; the other, Schizostege, constituted by himself, and represented
by a single species found in only one or two of the islands.

NOTE 67 (page 241)

ON THE DISPERSAL OF COMPOSITÆ BY BIRDS

The goldfinch’s habit of pecking at the heads of thistles, and pulling
out the achenes in bundles, is well known. Gätke mentions two suggestive
instances of birds feeding on the fruits of a Composite plant. According
to this observer, the Scarlet Grosbeak (Pyrrhula erythrina), when it
alights on Heligoland, always feeds on the achenes of Sonchus oleraceus,
which it picks off the plant; whilst the Parrot Crossbill (Loxia sp.),
feeds in Heligoland on burrs and thistles (_Heligoland as an
Ornithological Observatory_, pp. 407, 409). See Note 91.

NOTE 68 (page 264)

ON SOME OF THE HAWAIIAN ENDEMIC GENERA, EXCLUDING THOSE OF THE COMPOSITÆ
AND LOBELIACEÆ

_Haplostachys, Phyllostegia, and Stenogyne, all Labiate
Genera._—Phyllostegia is not strictly peculiar to Hawaii, since out of
the 17 species enumerated in the _Index Kewensis_, 15 are Hawaiian, 1
Tahitian, and 1 is accredited to Unalaska (one of the Aleutian Islands).
The last locality appears to be an error. The species in question is P.
microphylla, Benth.; and on looking up the original authority in
_Linnæa_ (vi. 570, 1831), I find the locality is thus given: “insula
coralligena Romanzoffii,” which is either one of the atolls of the
Paumotu Islands in about lat. 15° S. and long. 144° W., or a coral
island of the Marshall Group, most probably the former.... I paid some
attention to the suitability of the fruits of these three Labiate genera
for dispersal by frugivorous birds, for which the fleshy nucules in the
cases of Phyllostegia and Stenogyne apparently fit them. Out of the
fruits of five species of Phyllostegia examined by me, the
seed-coverings in three species, after the removal of the fleshy
covering of the nucule, were too soft for the protection of the seed in
a bird’s stomach. Hillebrand also observes (p. 347) that the nucules
when dried are wrinkled, and absorb moisture easily, a quality which, if
true of all the species, would make the distribution of the genus by
birds impossible. However, in two species I found the seed-coverings
somewhat harder. It would seem that since birds have largely ceased to
disperse these plants, the soft-skinned nucules would in the absence of
their selective agency more frequently characterise the genus. It is
possible that the dry nucules of Haplostachys, which according to
Hillebrand are not affected by drying, represent the original condition
of those of Phyllostegia, and that the fleshy character has been
acquired in this archipelago. It will be seen in the list on page 263,
that Haplostachys is regarded by Gray as a section of Phyllostegia. The
remarks under Phyllostegia, regarding the softness of the seed-coverings
beneath the fleshy coat of the nucule, also apply to Stenogyne; and
Hillebrand, in contrasting its fleshy nucules with the dry nucules of
Haplostachys, implies that they absorb water, which, I may remark, would
render them quite unfit for dispersal by frugivorous birds.

_Touchardia (Urticaceæ)._—According to Hillebrand, the solitary species
is by no means common in the group now. In 1897 I found it growing
abundantly some miles up the Waipio gorge, Hawaii.

_Cheirodendron (Araliaceæ)._—C. Gaudichaudii, the well-known “Olapa”
tree, is common in the forests of all the Hawaiian Islands between 2,000
and 5,000 feet; but I noticed it occasionally at greater elevations, as
on the south-east slopes of Mauna Kea, where it extends to 7,000 feet.
As described on page 343, the “Olapa” often grows in close contact with
the Lehua (Metrosideros polymorpha), the two trunks appearing as one.
The drupes would attract frugivorous birds and the pyrenes are well
adapted for this mode of dispersal. Mr. Perkins states that the drupes
are much sought after by the various species of Phæornis, a genus of
birds peculiar to Hawaii.

_Deterioration of Fruits for Purposes of Dispersal._—Among fruits or
endemic genera that have evidently deteriorated in the Hawaiian Group as
far as fitness for dispersal is concerned, may be mentioned, in addition
to those of Phyllostegia and Stenogyne above noticed, those of the
Araliaceous genera, Pterotropia and Triplasandra, and the Amarantaceous
Nototrichium. The pyrenes of the first two genera on account of their
thin covering, and the seed of the last-named genus on account of its
thin testa, seem ill-fitted now for transport in a bird’s stomach, yet
we cannot doubt that their ancestors originally arrived in this fashion.
The same principle is also illustrated by some Hawaiian non-endemic
genera of later eras that possess peculiar species, such, for instance,
as in the case of Elæocarpus discussed in Chapter XXVI.

NOTE 69 (page 366)

ON THE GERMINATION OF CUSCUTA

My observations were made on the Hawaiian endemic species (C.
sandwichiana) and on a Fijian introduced species. Germination occurs
readily in fresh water, the floating seedling growing rapidly. When the
germinating seed is placed on wet soil in the shade, the seedling grows
at the rate of 3/4 inch (19 mm.) a day. The store of nutriment contained
in the swollen radicular end will support the seedling for a couple of
days, and if it has not then found a host it withers and dies. At first
lying prone the seedling then lifts its upper end into the air, and it
was almost pathetic to notice it moving round and round, endeavouring
vainly to find some object near. The seedlings make no effort to strike
into the soil, and when they are allowed to attach themselves to a plant
they ascend rapidly, growing at the upper end and dying at the lower
end.

NOTE 70 (pages 477, 480-1)

ON BEACH-TEMPERATURE

My data are rather scanty; but, judging from observations made in
Hawaii, in South America, and in the south of England, the following
scale would probably be true of typical beaches where the sand is found
relatively cool and moist at a depth of four or five inches. This
moisture seems to arise entirely from subsoil drainage seaward. When a
beach fronts an arid, rainless region, few if any plants grow on it; the
sand is loose, hot, and dry at the depth indicated; and the temperature
of the surface half-inch rises to between 130° and 140° F., whilst four
inches down it is 95° to 100°. Salt-marshes situated behind a beach even
in a desert-region change its thermal behaviour, and it would then be
more like a beach skirting a vegetated sea-border in the same latitude.
The method of observation was as follows:—An unmounted thermometer of
the size of a clinical thermometer, but graduated higher, was placed
horizontally in the sand half an inch below the surface and a reading
taken. It was then pushed vertically into the sand until the bulb was
four inches deep and another reading taken. Provided that the sand is
moist beneath, the colour does not seem to make much difference, except
perhaps in very dark sands, none of which were tested.

_Ordinary Beach-Temperatures with an Unclouded Sky in the Hot Season
during the Early Afternoon._

+-----------------------------------+------------------+-----------------+
| |Surface half-inch.|Four inches deep.|
| +------------------+-----------------+
|Temperate latitudes about 50-55°| 100-105° F. | 77° F. |
|Sub-tropical latitudes about 30-35 | 105-110 | 80 |
|Tropical latitudes about 10-20 | 110-120 | 85 |
+-----------------------------------+------------------+-----------------+

This illustrates only the average condition. On a calm day in the case
of a beach facing south in the South of England, I have obtained exactly
the same readings in July as at Valparaiso in January, 112° at surface,
80° four inches deep.

NOTE 71 (page 479)

On the Buoyancy of the Seeds or Seed-vessels of some Chilian Shore
Plants

(1) _Nolana_, probably _paradoxa_. Common on the beaches of Southern
Chile. The ripe drupes have a somewhat fleshy outer covering which they
lose when lying on the sand, and present themselves then as dark-brown
angular “stones,” often five to six millimetres across. Inside the outer
hard covering of the stone is a layer of spongy tissue which gives it
buoyancy; but since these coverings are wanting at the scars marking the
basal insertion of the drupe, the embryo seems insufficiently protected
against injury during flotation in sea-water; and the seed-vessel at
first appears to be only fitted for conveyance by the currents over a
limited tract of sea. However, in a preliminary experiment on
seed-vessels that had been kept a few weeks, I found that 30 per cent.
floated after three weeks in sea-water. Subsequently, after drying for a
year, the seed-vessels were again tested in sea-water, nearly all of
them floating after three months’ immersion. Two of them, removed after
six weeks’ flotation, germinated healthily. These fruits are common in
beach-drift between Corral and Valparaiso.

(2) _Raphanus_, near R. maritimus. Growing near beaches in South Chile,
and not infrequently represented in the stranded beach-drift by the
pods, which in my experiments floated seven to ten days in sea-water,
after drying some weeks.

(3) _Franseria._ A species common on the beaches of Valparaiso and
Talcahuano. Its prickly fruits, after being kept six weeks, floated only
two to four days. They are well suited for transport in birds’ plumage.

NOTE 72 (page 483)

THE SOUTHERN LIMIT OF THE MANGROVE FORMATION IN ECUADOR.

... The southern limit of the mangrove formation on the west coast of
South America is usually placed at 4° S. lat.; but it is probable that
the vicinity of Tumbez in lat. 3° 30ʹ S. would be more correct. Baron
von Eggers would place it rather further to the north-east, near the
frontier of Ecuador and Peru in lat. 3° 20ʹ S. I spent eight days in the
locality last named and saw no evidence of the beginning of the
mangrove-formation.

NOTE 73 (page 495)

ADDITIONAL NOTE ON THE TEMPERATURE OF THE DRY COAST OF ECUADOR BETWEEN
PUNA ISLAND AND THE EQUATOR.

... Baron von Eggers gives the mean annual temperature for El Recreo,
about half a degree south of the equator, at 75° F., which is near that
of Rio de Janeiro in lat 23° S. on the east coast of the continent. Mr.
F. P. Walker has kindly given me the results of temperature-observations
covering a period of ten years, taken in the room for testing cables at
Santa Elena Point (2° 10ʹ S.), usually about 6·30 a.m. The range of the
monthly means was 71° F. (August) to 79·1° (March), and the mean for the
year was 74·8°. In that locality a typical daily range would be 65° to
80°; and Mr. Walker believes that a minimum of 59° has been recorded.

NOTE 74 (page 495)

OBSERVATIONS ON THE TEMPERATURE OF THE HUMBOLDT CURRENT FROM ANTOFAGASTA
NORTHWARD, BETWEEN JANUARY AND MARCH, 1904 (Fahrenheit scale)

The observations were usually taken at the anchorages, but in some
places, as at Ancon and Puerto Bolivar, they were taken from a boat
outside the roadstead.

If we wish to ascertain how the Humboldt Current retains its cool
temperature as it advances through the tropics to the equator, a glance
at the following table will show that the surface-temperatures can aid
us but slightly, since they do not vary in accordance with the latitude,
a subject further discussed below. We can, however, obtain some valuable
indications from the deeper temperatures. Let us take for instance the
plane of 60°. Whilst south of Ancon (lat. 11° 45ʹ S.) it was rarely
deeper than four fathoms, north of this latitude it descends rapidly,
being probably about ten fathoms down at Salaverri and Eten and about
twenty fathoms deep at Payta, in latitude 5° S., where the Humboldt
Current leaves the coast. Within the Gulf of Guayaquil it is probable
that the plane of 60° would descend to nearer thirty fathoms, the region
being outside the influence of the current.

Some interesting facts are also elicited from the variation of the
surface-temperatures. When we were coasting along at a distance of five
or six miles from shore the readings were fairly constant from hour to
hour varying only a degree or so. But nearer the land, for instance,
about two or three miles away, the variation from hour to hour amounted
to two or three degrees, whilst within the limits of the anchorages, a
mile and less from the coast, the change from hour to hour amounted to
three or four degrees. Nor was there any uniformity at the same hour
over the surface of a roadstead. The temperature would often rise or
fall a degree every few boat-lengths. Sometimes the inshore water was
the coolest and sometimes it was the warmest. Thus at Iquique the
inshore water was three degrees warmer than the water half a mile out,
whilst at Mollendo, when the temperature one-third of a mile off the
shore was 70°, it was 63° close to the rocky coast. The same thing was
exhibited at Pisagua, where the surface-water two miles out at sea was
61°, whilst close inshore at the anchorage it was 58°. It was evident
that there was a considerable intermingling of the warmer surface and
the colder, deeper waters on the coasts of Chile and Peru. This was
particularly noticeable on a rocky, steep-to coast, or where there was
an uneven bottom. At some places, indeed, the warm upper layer did not
exist, the cold water welling up all along the coast. This was
especially the case between the 22nd and 19th parallels of latitude, a
tract of coast in which lie Tocopilla, Iquique, and Pisagua, and
probably the coolest part of the sea-border at this season of the year.

During a fortnight spent at Ancon (11° 45ʹ S.), between January 27 and
February 10, I paid considerable attention to the local climatic
conditions, and especially to the temperature of the inshore water. The
daily range of the air-temperature was only five or six degrees, the
average minimum and maximum being 71° and 75·9°, and the mean for the
period 73·5°. The mean temperature of the surface-water at the head of
the pier, from observations taken at about 7 a.m. and 4 p.m., was 68·6°,
or five degrees cooler than the air, the mean temperature in the morning
being 69·1° and in the afternoon 68°.

OBSERVATIONS ON THE TEMPERATURE OF THE HUMBOLDT OR PERUVIAN CURRENT

(Made by H. B. Guppy, January to March, 1904.
Those at Panama are added for the sake of comparison)

+--------------+----------+----------+--------+--------+---------+----------------------------------------------------------------------------------------------------------------------+
| | | Distance | | | | Depths in fathoms: temperature in Fahrenheit degrees. |
| | Depth | from | | | +--------+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+
| Locality. |(fathoms).| shore | S. lat.| Date. | Hour. | | | | | | | | | | | | | | | | | | | | | | | |
| | | (miles). | | | |Surface.| 1.| 2.| 3.| 4.| 5.| 6.| 7.| 8.| 9.| 10.| 11.| 12.| 13.| 14.| 15.| 16.| 17.| 18.| 19.| 20.| 21.| 22.|
+--------------+----------+----------+--------+--------+---------+--------|----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Antofagasta | 22 | 2/3 | 23°40ʹ |{Jan. 12| 5 p.m. | 71° | 70°| ...| ...| ...| ...| ...| 61°| ...| ...| 57°| ...| ...| ...| ...| ...| 56°| ...| ...| ...| ...| ...| ...|
| | | | |{Jan. 13| 6 a.m. | 70 | ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| 55 | ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Tocopilla | 18 | 1/2 | 22 0 | Jan. 14| 8 a.m. | 57 | ...| ...| ...| ...| ...| ...| ...| ...| ...| 56 | ...| ...| ...| ...| ...| 55 | ...| ...| ...| ...| ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Iquique | 9 | 2/3 | 20 15 | Jan. 15|5.30 a.m.| 59 | ...| ...| ...| ...| 58 | ...| ...| ...| 55 | ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Pisagua | 10 | 1/2 | 19 30 | Jan. 16| 8 a.m. | 57·5 | ...| ...| ...| ...| ...| ...| ...| ...| ...| 56 | ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Arica | ... | 1/4 | 18 25 | Jan. 16| 7 p.m. | 66 | ...| ...| ...| ...| ...| ...| ...| 57 | ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| | | |{17 0 | Jan. 17| 8 a.m. | 61·5 | ...| ...| ...| ...| ...| ...| ...| 57 | ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
|Mollendo | 22 | 1/3 |{17 0 | Jan. 17| 6 p.m. | 65 | ...| ...| ...| ...| ...| ...| ...| ...| ...| 59 | ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
| | | |{17 0 | Jan. 18|9.30 a.m.| 62 | ...| ...| ...| ...|57·5| ...| ...| ...| ...| 57 | ...| ...| ...| ...| 56 | ...| ...| ...| ...| ...| ...| 55 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Callao | 5 | 1 | 12 3 | Jan. 20| 6 a.m. | 60·5 | ...| ...| ...| ...|58·5| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Ancon | 17 | 1-3/4 | 11 45 | Jan. 29| 11 a.m. | 66 | ...| ...|6·15| ...|58·5| ...| ...| 57 | ...| 57 | ...| ...| ...| ...| ...| ...|56·5| ...| ...| ...| ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Salaverri | 6 | 3/4 | 8 15 |Febr. 21|6.30 p.m.| 65 | ...| ...| ...| ...| ...| 62 | ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Eten | 7 | 1/2 | 7 0 |Febr. 22| 6 a.m. | 66·5 | ...| ...| ...| ...| ...| ...| 63 | ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Payta | 16 | 1-1/2 | 5 0 |Febr. 23| 9 a.m. | 70 | ...| ...| ...| ...| ...| ...|67·5| ...| ...| ...| ...| ...| ...| ...|62·5| ...| ...| ...| ...| ...| ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Puerto Bolivar| 10 | 1 | 3 10 | Mar. 7| noon | 78 | ...| ...| ...| ...|71·5| ...| ...| ...| ...| 70 | ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|Panama | 8 | 3 |{(8 50) | Mar. 23| 4 p.m. | 82 | ...| ...| ...| ...| ...| ...|79·5| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
| | | |{( N.) | Mar. 24| 7 a.m. | 79·5 | ...| ...| ...| ...| ...| ...| ...|79·5| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...| ...|
+--------------+----------+----------+--------+--------+---------+--------+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+

The Ancon climate at this period is full of oddities and abnormalities,
and in this way typifies much of the coast of Peru. Thus, since the heat
of the day is tempered by the cool south-westerly winds which die away
in the evening and give place usually to warm, light, northerly and
north-westerly breezes, there is, as above remarked, but a small
difference between day and night temperatures. The coldest time of the
twenty-four hours is not in the early morning but at sunset. The sea off
the beach is, on the average, much cooler than the air, which is not a
normal state of things; and again, the water is often two or three
degrees colder in the evening than it is in the morning, which is very
unusual. Though the sea-border is practically a desert for the greater
part of the year and has no rain, it is frequently enveloped in
drizzling fogs or “garuas.” Judged from a European standard, things go
by contraries on the coast of Peru; and this is entirely the effect of
the Humboldt Current.

The temperature of the inshore waters of Ancon Bay varied considerably
during the twenty-four hours. During the day, with the prevailing
southerly wind, the cool waters of the current had free access to the
bay, and swept around its border in their course north; but in the
night, when northerly breezes occurred, the cold waters of the current
were pushed off the coast and their place taken by the warmer inshore
waters from the north; and this sometimes continued for a day or two.
When the current again got mastery and its clean, cool waters filled the
bay, the temperature of the water dropped suddenly five or six degrees,
and the bay was filled with fish. At such times men in boats leave the
beach, and in a few minutes, with hand-nets and baskets, they obtain
thousands of the small fry. Other men, fishing with lines from the
pier-head, seem ill-contented unless they can catch fish of the size of
small mackerel at the rate of one a minute.

There can be little doubt that on the coasts of Chile and Peru the
instincts of fish often lead them astray, on account of the sudden
changes of temperature arising from the conflict between the warmer
waters of the open sea and the cooler waters of the current. From the
preceding remarks it will be inferred that sometimes the current is
pushed off the coast for a while and its place taken by the warm waters
from the north. At other times it dives down, so to speak, and flows at
a deeper level, and warmer waters prevail both out at sea and inshore.
At other times again, and this must be most disconcerting to the fish,
the cold current suddenly appearing at the coast predominates at the
surface for days together, and we have stretches of coast which,
although lying within tropical latitudes, are washed by waters having
the temperature of the temperate zone. It is to such causes that we must
attribute the reckless habits of fish on these coasts. They are known to
throw themselves on the beaches in thousands, where by their decay they
taint the air long afterwards. Mr. Anderson Smith in his recent book on
_Temperate Chile_ vividly describes what goes on on such occasions at
the port of Valdivia. At times the scene must be indeed a strange one,
since huge octopi are rolled up on the beaches in numbers, and are
regarded by the indigenes as deliberately seeking their death. Whether
they commit suicide or not, “their beaks that blacken the edge of the
sea-wash in places” afford a melancholy proof that their instinct has
blundered.

_The Mode of Observation._—A thermometer made on the Sixe pattern which
I used several years ago for taking the bottom-temperatures of rivers,
was employed for the deeper temperatures, and at critical depths the
observations were always repeated. This instrument was compared after
each set of observations with an ordinary thermometer graduated on the
stem, which was compared with my standard thermometer provided with a
Kew certificate.... The observations in the Panama Roadstead have been
added for the sake of contrast.

NOTE 75 (page 496)

ON THE STRANDED MASSIVE CORALS APPARENTLY OF THE GENUS PORITES FOUND ON
THE COAST OF PERU AND NORTH CHILE, AT ARICA (18° 25ʹ S.), CALLAO (12° 3ʹ
S.), AND ANCON (11° 45ʹ S.)

At Arica they occurred on the beach only. At Callao they also extended
inland on the low spit at Punta for about 100 yards. At Ancon they were
found not only on the beach but also twenty or thirty paces inland on
the low adjoining plains. Their size varied from three inches to three
feet. They were all more or less rounded by wave action, and were
extensively burrowed by boring molluscs. Whilst some on the beach still
displayed the dried-up soft parts of the boring mollusc, others inland
were falling to pieces and undergoing chemical change. There was nothing
to indicate that the corals were recently alive; and at Ancon they
appeared to have been torn off a rocky spit of andesite that had become
exposed on the beach during a recent movement of emergence, of which
there is other evidence on this coast. Further particulars are given on
page 496.

NOTE 76 (page 429)

STRANDED PUMICE ON ENGLISH AND SCANDINAVIAN BEACHES

Sernander, in his description of the Atlantic drift of the Scandinavian
coast, refers to the occurrence of a small amount of true pumice. I have
found solitary fragments of acid pumice well rounded by wave-action at
Croyde Bay on the north coast of Devonshire, at the mouth of Salcombe
Harbour on the south coast of the same county, and at Maenporth, near
Falmouth, in Cornwall. Steamer slag, in some cases rudely simulating
pumice, is common on all the South of England beaches I have examined.
It is also common on the Scandinavian coasts, though seemingly regarded
by Helge Bäckström, who is quoted by Sernander, as derived from the
factories on the east coast of England. (See on these subjects a paper
by Helge Bäckström, “Über angeschwemmte Bimsteine und Schlacken der
nordeuropäischen Küsten”; Bihang till _K. Sv. V. A. Handl._ Bd. 16. Afd.
3, 1890; also a letter in _Nature_, about 1886, by H. B. Guppy.)

NOTE 77 (page 21)

ON THE MODE OF DISPERSAL OF KLEINHOVIA HOSPITA

This small tree has a very wide distribution in the tropics, ranging
from East Africa and the Mascarene Islands through India, South-eastern
Asia, Malaya, New Guinea, and the Solomon Islands to Fiji and Tahiti. It
is a plant that grows in inland open woods as well as amongst the
littoral trees on the beach; and it is always doubtful (in Malaya, Fiji,
and Samoa) whether to regard it as a shore plant or as an inland plant,
different authors varying on this point. In Vanua Levu I formed the
opinion that it is only an intruder amongst the littoral vegetation. In
accounting for its distribution we have to choose between man, the bird,
and the current. Though it may sometimes be noticed in native
plantations, as I observed in the Solomon Islands, the tree has no
special use; and the Solomon Island natives themselves indicated to me
that the parrots that fed on the fruits of the tree aided in
distributing the plant. The buoyant behaviour of the seeds, which are
freed by the dehiscence of the bladder-capsules on the tree, is not
constant. Whilst in the case of the seeds of littoral trees in Fiji I
found that 30 per cent. floated after ten weeks, Prof. Schimper
ascertained in the case (seemingly) of Malayan seeds that they sank at
once. The seed-structure connected with the buoyancy is, as shown on
page 105, accidental in character, and reference is made on page 20 to
other plants of doubtful littoral reputation, in which the buoyant
qualities are variable. The occasional buoyancy of its seeds will only,
as I think, explain its occasional station at the coast; and I agree
with Prof. Schimper (p. 156) when he attributes its wide distribution to
birds, the seeds being hard, crustaceous, and about three millimetres
across.

NOTE 78 (page 436)

ON THE “SEA”: AN UNIDENTIFIED WILD FRUIT-TREE IN FIJI

This is a fair-sized forest tree common in places in the lower forests.
I have never been able to identify it; but a “putamen” which was sent to
the Kew Museum was named Spondias with a query. It is to be hoped its
true botanical name will be discovered by one of my successors. Seemann
places it amongst the “desiderata” concerning which further information
is needed. The fruit is a drupe 2 to 2-1/2 inches long possessing a
pleasant fruity odour and inclosing a hard two-celled stone about 1-2/3
inch long, one cell containing a large fleshy seed covered with tawny
hair, the other filled with the hair only and containing no seed. The
Fijians say that these fruits, large as they are, are swallowed by the
fruit-pigeons, the stones being found in their gullet. The leaves are
distichous, alternate, lanceolate, eight or nine inches long, glabrous
and dark green above, and covered below with a whitish woolly matted
tomentum. The empty stones are not uncommon in the stranded beach-drift.

NOTE 79 (page 395)

ON WILLOW-LEAVED RIVER-SIDE PLANTS

A number of observers, beginning with Humboldt, in his _Ansichten der
Nature_, and including Seemann, L. H. Grindon, Ridley, Beccari, and
others, have referred to what is called “stenophyllism” in plants. These
willow-leaved river-side plants are found all over the globe, such
plants usually growing close to the water’s edge in situations where
they are liable to be more or less submerged when the river is in flood.
Seemann, Beccari, and Ridley mention more than two dozen genera
belonging to a great variety of orders, and including Acalypha,
Antidesma, Calophyllum, Eulalia, Eugenia, Fagræa, Ficus, Garcinia,
Ixora, Lindenia, Melastoma, Podocarpus, Psychotria, &c., all tropical,
and represented either in Fiji, Borneo, or in the Malay Peninsula;
whilst my readers will recall amongst temperate floras river-side plants
of the genera Epilobium, Lythrum, Salix, &c., possessing the same form
of leaf and the same station. The genus Eugenia comes under this
category in Fiji, Borneo, and the Malay Peninsula, with reference to one
or more of the species. In Fiji, species belonging to the genera
Lindenia and Dolicholobium especially attracted my attention in this
respect. It is noteworthy that several of the Bornean plants and some of
the Fijian plants here concerned are endemic. Just as I have remarked in
the question of the buoyancy of seeds and fruits, that not all
water-side plants have buoyant seeds or fruits, but that nearly all
plants thus endowed are found at the water-side, so we may say of the
willow-leaved plants, that not all river-side plants have the
willow-form of leaf, but that plants thus characterised gather at the
river-side. Beccari and Ridley regard this willow-form of leaf as the
result of adaptation. Seemann remarks that we have here the old question
whether the webbed feet of a duck are the cause or the effect of the
bird’s swimming; and I take the same position. (See Seemann’s _Flora
Vitiensis_; Ridley in _Trans. Linn. Soc. Bot._, vol. iii. 1888-94; and
Beccari’s _Nelle Foreste di Borneo_, 1902, or the English edition of
1904.)

NOTE 80 (pages 255, 504)

MR. PERKINS ON THE HAWAIIAN LOBELIACEÆ (_Fauna hawaiiensis_, vol. I.)

My view, that the early Hawaiian Lobeliaceæ acquired the monstrous form
of their flowers in the humid forests of a later age, is supported by
the observations of Mr. Perkins on the connection between the
highly-specialised nectar-eating Drepanids of Hawaii and the
highly-specialised flowers of the Tree-Lobelias, a subject further
discussed in Chapter XXXIII. This naturalist ascertained, in the case of
one of the trees, that fertilisation could only be effected by these
birds. So close is the biological connection between the Drepanid and
the Tree-Lobelia, that Mr. Perkins finds here in part the cause of the
development of the most remarkable forms of the birds. The botanist,
also, would not dissociate the plants from this conclusion. There would
be every reason to look for abnormal growth in birds and plants when the
bird depends on the flower for its food, and the flower is dependent on
the bird for its pollenisation. It is through such guises that the
zoologist and the botanist have to penetrate when establishing the
systematic affinity.

NOTE 81

ON THE VERTICAL RANGE OF SOME OF THE MOST TYPICAL AND MOST CONSPICUOUS
OF THE PLANTS IN THE FORESTS ON THE HAMAKUA SLOPES OF MAUNA KEA, HAWAII

During a descent of this mountain on its north side to near Ookala, the
conditions were unusually favourable for recording the range of altitude
for some of the plants easily recognisable.

Acacia koa began at 6,700 feet, and extended down to 2,300 feet.

Rubus (“akala”) began at 6,500 feet, and extended down to 2,500 feet.

Cheirodendron (“olapa”) began at 6,400 feet, and extended down to 2,200
feet.

Cyanea, a lobeliad growing on trunks of tree-ferns, began at 4,000 feet,
and extended down to 2,300 feet.

Freycinetia began at 3,850 feet, and extended down to 2,000 feet.

Asplenium nidus began at 2,800 feet, and extended down to 2,200 feet.

Aleurites moluccana began at 1,800 feet, and extended down to 50 feet.

Metrosideros polymorpha, ranging through all the zones.

NOTE 82 (page 416)

ABORIGINAL WEEDS[7]

(Found by Captain Cook’s Botanists, Banks, Solander, the Forsters,
Nelson, &c.,
in the Pacific Islands, 1768-80)

+-------------------------------+---------------------------------+----------------------------------+
| | Locality given by Cook’s | General distribution. |
| | botanists. | |
+-------------------------------+---------------------------------+----------------------------------+
|Cardamine sarmentosa | Tahiti | Polynesia. Introduced into Peru.|
|Sida microphylla | Tonga, New Hebrides | Old World tropics. |
|Sida rhombifolia | New Hebrides. H. | Tropics of Old and New World. |
|Urena lobata | Tahiti | Tropics of Old and New World. |
|Waltheria americana |{ Tahiti. H. |} Tropics of Old and New World. |
| |{ New Caledonia |} |
|Oxalis corniculata | Tahiti | Old and New World. |
|Cardiospermum halicacabum | Tahiti. H. | Tropics of Old and New World. |
|Desmodium polycarpum | Tahiti | Old World. |
|Phaseolus truxillensis | Tahiti. H. | Tropics of Old and New World. |
|Lablab vulgaris | Tahiti | Old World tropics. |
|Abrus precatorius | Tahiti | Tropics of Old and New World. |
|Cassia sophora | Tonga | Tropics of Old and New World. |
|Hydrocotyle asiatica | New Caledonia | Tropics of Old and New World. |
|Oldenlandia tenuifolia | New Hebrides | Old World? |
|Oldenlandia paniculata | Tonga | Old World tropics. |
|Geophila reniformis | Tahiti | Tropics of Old and New World. |
|Ageratum conyzoides | New Hebrides. H. | Tropics of Old and New World. |
|Adenostemma viscosum | Tahiti. H. | Tropics of Old and New World. |
|Eclipta alba | New Hebrides | Tropics of Old and New World. |
|Siegesbeckia orientalis | Tahiti | Tropics of Old and New World. |
|Bidens pilosa | Tonga | Tropics of Old and New World. |
|Dichrocephala latifolia | Tahiti, Tonga, New Hebrides | Old World tropics. |
|Sonchus asper | Tonga, New Zealand | Old World. |
|Ipomœa insularis | New Hebrides, Tonga, Hawaii | Australia and Polynesia. |
|Ipomœa bona-nox | New Hebrides, Tonga, Tahiti. H.| Tropics of Old and New World. |
|Solanum nigrum, var. oleraceum.| Tahiti. H. | Old and New World. |
|Physalis angulata | Tahiti | Tropics of Old and New World. |
|Vandellia crustacea | Tahiti | Tropics of Old and New World. |
|Leucas decemdentata | Tahiti | Old World tropics. |
|Teucrium inflatum | Tonga | New World tropics. |
|Amarantus melancholicus, var. | | |
| tricolor | New Hebrides, Tahiti | Old World. |
|Euxolus caudatus | Tonga, Tahiti | Old World tropics. |
|Achyranthes aspera | Tahiti | Old and New World. |
|Cyathula prostrata | Tahiti | Old World tropics. |
|Fleurya interrupta | Tahiti, Tonga | Old World tropics. |
|Commelina pacifica | Tonga, New Caledonia. H. | Tropics of Old and New World. |
|Eleusine indica | Tahiti. H. | Tropics of Old and New World. | |
+-------------------------------+---------------------------------+----------------------------------+
Footnote 7:

Seemann is the principal authority, the results of his examination of
the old collections being given in his _Flora Vitiensis_. Species
regarded by Hillebrand as indigenous in Hawaii or as existing in that
group at the time of its discovery by Cook are indicated by H in the
second column.

NOTES 83-89 omitted

NOTE 90 (page 29)

ON THE BUOYANCY OF THE SEEDS OF EUPHORBIA AMYGDALOIDES AND E. SEGETALIS

The seeds of both species have no proper buoyancy, and display no
structure in their testas suggesting it; though, through the shrinking
of the nucleus, a temporary floating power may be acquired with less
mature or imperfect seeds. They support the general principle indicated
for the British species on page 29.

NOTE 91

MR. E. KAY ROBINSON ON THE DISPERSAL OF ASTER TRIPOLIUM

According to this naturalist, the seeds of this plant are eaten in
winter by snow-buntings on the English east coast. In reply to my query
he tells me that the “draggled fluff still containing seeds” might
easily adhere to birds (_The Country-Side_, Sept. 30, 1905).

GENERAL INDEX

_Note._—Several subjects are worked up in this index, which, on account
of the plan of the book, are not dealt with connectedly in the text. As
examples may be cited the entries under the heads of “Hawaiian Flora”;
“Species, their development”; “Fruit-pigeons”; “Polymorphous Species”;
&c.

The figures in larger type indicate the pages where the subject is
treated at length or where the most important points are discussed. This
sign is not often used where the references can be classed, or where
several references of importance belong to the same subject.

Abrus precatorius, 531, 605

Acacia farnesiana, 478, 552, 555, 556, =557=, =559=

Acacia heterophylla, 200

Acacia koa, 151, =200=, 533, 604

Acacia laurifolia, 45, 133, 134, =164=, 166, 200, 529, 551

Acacia richii, 531, 549

Acæna exigua, 253, 275;
genus, 270-2, =275=, =276=

Acalypha, 395, 603

Acer campestre, 536

Achillea millefolium, 536

Achras, 373

Achyranthes aspera, 605

Acorus;
home of the genus, 396

Acrostichum squamosum, 593

Adaptation in relation to means of dispersal, 11, =99-103=, 105,
=119-129=, 324, 516, 522

Adenanthera pavonina, 159, 420

Adenostemma viscosum, 240, 417, 568, 605

Ægiceras, 470-1

Æthusa cynapium, 28, 536

Afzelia bijuga, 21, =93=, 107, =170-6=, 436, 529, 563

Agapetes, 265

Agathis: _see_ Dammara

Ageratum conyzoides, 417, 531, 605

Aglaia, 376

Agrostemma, 471

Agrostis, 272, 275, 538

Aira cæspitosa, 417, 418

Ajuga reptans, 28, 537

Alchemilla arvensis, 418, 536

Alchemilla vulgaris, 417

Aleurites moluccana, 59, 61, 361, =418=, 435, 438, 533, 549, 554,
558-9, 560, 604

Alexander, Prof., 587

Algaroba, 485, 557

Algerian beach-flora, 34

Alisma natans, 537

Alisma plantago, 38, 92, 537

Alisma ranunculoides, 537

Alliaria officinalis, 536

Alnus glutinosa, 31, 37, 430, 537

Alphitonia, 333, =346=, 357, 531

Alpine floras, 4, 34, 238;
_see_ under Mountain floras

Alpinia, 531

Alopecurus, 538

Alsinidendron, 262-3

Alstonia, 381, =384=, 548

Alyssum maritimum, 536

Alyxia, 334, =344=, 531, 533

Amarantus melancholicus, 605

Amarouria, 265

America, as the home of tropical shore-plants, 67-75;
_see_ under Hawaiian flora and under Tahitian flora for the American
plants in those islands

Amorphophallus, 412, 414

Anagallis arvensis, 537

Ancon (Peru), 482, 497;
climate, 492, 598

Angelica sylvestris, 28, 536

Aniseia uniflora, 530, 563

Anona paludosa, 68, 77, 109, 115, 435, 438, 486, 488-9, 498

Antarctic flora, represented in the Pacific islands, 271-3, 287, 292,
294, 305, 503-4, 518

Anthyllis vulneraria, 536

Antidesma, 371-2, 603

Antofagasta (Chile), beach-drift, 480

Apetahia, 252, 256-7

Apium graveolens, 28, 536

Apium inundatum, 28, 536

Apium nodiflorum, 28, 37, 536

Arabis albida, 568

Arabis hirsuta, 536, 568

Arabis thaliana, 536, 567

Arachis hypogæa, 479

Araliaceæ, 261-3

Araucaria, 298

Arcangeli, Prof., on the Italian species of Medicago, 431

Arenaria peploides, 35, 36, 107, =116=, 429-32, 536, =541=, 544

Argemone mexicana, 533

Argyreia tiliæfolia, 20, 106, 110, 533-4, 558

Argyroxiphium, 236-8, 240, 243-4

Arica (Chile), 481-2, 497

Armeria vulgaris, 33, 34, 36, 511, 537, =540=

Armeria maritima, 477

Artemisia, 238, 240, 269, 272, =278-9=, 540

Artemisia absinthium, 279, 536, 540

Artemisia maritima, 33

Artemisia tridentata, 279

Artemisia vulgaris, 279, 536, 540

Artocarpus incisa, 531

Artocarpus integrifolia, 531

Arum maculatum, 537

Arundel, Mr., 49, 179

Arundo phragmites, 538

Asparagus, 538

Aspidium aculeatum, 226

Aspidium caryotideum, 593

Aspidium filix mas, 225, 593

Asplenium adiantum nigrum, 225, 593

Asplenium aspidioides, 593

Asplenium contiguum, 593

Asplenium fragile, 593

Asplenium monanthemum, 593

Asplenium nidus, 554, 604

Asplenium trichomanes, 225, 593

Astelia, 270-2, 274, 290, =291=, 292-4, 305

Aster tripolium, 28, 34-6, 89, 536, 540, 545, 581, 605

Astrocaryum, 499

Astronia, 376

Atlas, Great;
flora, 238, 277

Atriplex patula, 537, 544;
genus, 284, 416, 546

Avicennia, 68, 69, 77, 78, 438, 484-5, 489, 498

Azolla, 488

Azores, 505

Bäckström, Helge, 601

Baillon, on the Lobeliaceæ, 251

Baker, Mr., on Fijian ferns, 224, 592

Bakeria, 265

Ball, Mr., 238, 476

Ballota nigra, 28, 537

Bananas: _see_ Musa

Banyans, 387

Barbarea vulgaris, 536

Bark, in beach-drift, 430

Barratt, Mr., 215

Barringtonia formation, 550

Barringtonia, genus;
buoyancy and structure of fruits, 17, 18, 121, 160;
mode of dispersal, 161;
seed-structure and vivipary, 132, 168, 573;
relation between coast and inland species, 134, 166

Barringtonia edulis, 19, 161, 531, 573

Barringtonia excelsa, 19, 108, 574

Barringtonia racemosa, buoyancy of fruits, 18, 161, 529;
structure of seeds and fruits, 108, 161, 564, 573-6;
fruits in drift, 76, 78, 435;
station, 43, 47, 551;
distribution and dispersal, 68, 160, 551, 563

Barringtonia samoensis, 19

Barringtonia speciosa, buoyancy of fruits, 18, 161, 529;
structure of seeds and fruits, 108, 114, 161, 573-6;
fruits in drift, 76, 78, 79, 435-7;
station, 43, 551;
distribution and dispersal, 49, 56, 57, 64, 68, 160, 563

Barringtonia, undescribed species;
in Fiji, 19, 161, 531, 574;
in Solomon group, 161

Barrows, Prof., 373

Bartsia odontites, 537

Bassia, 374

Batatas edulis, 415;
_see_ under Ipomœa batatas

Batis maritima, 482, 485, =546=, 557

Bats, as dispersing agents, 321, 343, 394, 510, 514

Battandier, M., 34

Bauhinia, 531

Beach flora: _see_ Littoral plants

Beach formation, 43, 550

Beach seed-drift, 31, =429=, 479, 480, 482, 489, 499, =557=

Beach temperature, 477, 480, 481, =595=

Beal, Prof., 568

Beccari, Dr., on the dispersal of Brackenridgea, 124, 569;
on the Cassowary as a seed-distributor, 152;
on Ficus, 388, 504;
on the retrocession of cultivated plants, 161;
on Sararanga, 156;
on willow-leaved plants, 603

Beech-nuts in drift, 429

Begonia, 394, 509

Begoniaceæ, 263, 394

Bell, Mr. Jeffrey, on Peruvian corals, 497

Bentham, Mr., 2, 423;
on the Compositæ, 236, 245, 248

Berberis vulgaris, 535

Berkeley, Mr., 539

Bermuda, 348, 351

Bernicla sandwicensis, 241, 275, 283

Berrya, 379

Beta maritima, 35, 537, 542

Betula alba, 537

Bidens, 379

Bidens cernua, 28, 31, 536, 540, 544

Bidens pilosa, 379, 533, 605

Bidens tripartita, 28, 536, 544

Bird, Miss, 584, 587

Birds, differentiation of, 5-8, 504-6, 514, 520-2;
crossing oceans, 506;
migrations, 505-6;
Polynesian, 67;
at high altitudes, 241;
biologically connected with plants, 504;
as seed-dispersers, 5, 205, 226, 241, 296, 321;
_see_ under Fruit-pigeons, Ducks, Geese, Sea-birds, Tetraonidæ, &c.

Bischoffia, 381, =386=, 531

Blysmus rufus, 537

Bobea, 262-3

Boehmeria, 263, 356

Boerhaavia, =355=, 552, 568

Bonin Islands, 54, 320

Boobies (Sula), as seed-dispersers, 188, 347, 356, 511

Borago officinalis, 537

Bourbon Island: _see_ Mascarene Islands

Bourne, Mr., 172

Brackenridgea, 113, 124, 569

Brandis, Dr., on the dispersal of Santalum album, 283

Brassica, 536

Breadfruit (Artocarpus), 412, 415, 531

Breweria, 362-3

Brighamia, 252, 255, 258

British flora, 23, 31, 115, 432, 535, 539-44

Brown, Dr. R., on seed-dispersal by sea-birds, 510

Brown, Mr. R., on a drift seed of Cæsalpinia, 189

Broussaisia, 263

Bruguiera, 43, 441, 551;
dispersal by currents, 48, 55, 77, 94, =461-2=, 467, 529;
distribution, 54, 68, 69, =461=, 563;
in beach-drift, 435, 437, 461;
seed-development and germination, =463-6=, 468-71;
fertilisation, =462=

Bryonia dioica, 536

Buller, Sir W., on the fruits and seeds eaten by New Zealand birds,
292, 296, 301, 321, 337, 347, 508, 541

Buoyancy of seeds and fruits, of Pacific plants, 12-22, 104-15;
of Fijian plants, 44-6, 529, 531-3;
of Hawaiian plants, 57, 533, 552;
of Tahitian plants, 49;
of British plants, 23-39, 115, 535-8, 539-44, 566;
its relation to sea-density, 88-98, 516;
structures concerned, 17, 104-18;
the question of adaptation, 119-29, 516, 569;
effect of inland extension, 121, 568;
the great sorting process, 16, 24, 30, 515;
long flotation experiments, 530, 539;
tables showing results of experiments, 529-38, 552;
effect of drying, 535, 538, 539-44, 571-2;
precautions in testing buoyancy, 566;
the risks of the floating seed in warm seas, 79-87

Burkill, Mr., on the Tongan flora, 224, 232, 335, 385

Butomus umbellatus, 537

Butterflies at high altitudes, 509, 583

Buttneria, 379

Button, Mr., 172

Byronia, 371

Cacao, 489

Cacti, 471, 485, 560

Cæsalpinia, general account of Pacific species, 183-97

Cæsalpinia bonduc, buoyancy of seeds, 21, 192-5, 529;
structures concerned with seed-buoyancy, 106, 191-2;
dispersal by currents, 189;
distribution, 186, 563;
station and extension inland, 49, 186-7

Cæsalpinia bonducella, buoyancy of seeds, 21, 192-5, 529, 530-1, 552;
structures concerned with seed-buoyancy, 106, 111, 191-2;
dispersal by currents, 57, 189, 430, 562, 563;
in beach-drift, 189, 430, 437-8, 558;
dispersal by birds, 57, 188, 511, 581;
station and extension inland, 42, 59, 186-8, 551, 552-4, 559;
germination, 191;
distribution, 68, 186, 563;
suggested relation to C. bonduc, 573

Cæsalpinia nuga, 183

Cæsalpinia, undescribed mountain species of Fiji, 184-5

Cakile maritima, 30, 35, 36, 109, =116=, 429-31, =432-3=, 536, 539,
=542=, 544

Calamintha officinalis, 28, 537

Californian current, 491

Calla palustris, 537, 544

Callao, 482, 492, 496-7, 599, 601

Callitriche, 38, 537

Calonyction: _see_ Ipomœa

Calophyllum;
relation between coast and inland species, 17, 18, 120, 134, 136, 533

Calophyllum amœnum, 534

Calophyllum burmanni, 18, 136, 531

Calophyllum calaba, 534

Calophyllum inophyllum;
buoyancy of fruits, 18, 529, 534, 552;
structures concerned in buoyancy, 107, 113, 115, 122;
fruits in beach-drift, 437, 558;
station, 42, 43, 49, 52, 550, 554;
its relation to inland species, 136

Calophyllum spectabile, 18, 136, 389, 531

Caltha palustris, 85, 535

Camelina sativa, 567

Campylotheca, 236-8, 240, 243-4, 533

Cananga odorata, 159, =393=, 531

Canarium, =400=, 532

Canavalia, 107, 201-2;
Pacific species, =145=, 578;
relation between the littoral and inland species, 20, 134, 145

Canavalia ensiformis, 145, 529, 563, 578-9, 581 (turgida)

Canavalia galeata, 20, =145-6=, 533

Canavalia sericea, 107, 145, 529, 563, =578-9=

Canavalia obtusifolia;
seed-buoyancy and dispersal by currents, 19, 83, 145, 529, 562,
=579=;
cause of buoyancy, 107, 113;
in beach-drift, 437-8, 489;
station and distribution, 42, 43, 488, 498, 547, 550, 563, =578=

Candolle, A. de, 25, 62, 80, 239, 418, 562, 573

Candolle, C. de, 376

Canna indica, 532

Canthiopsis, 265

Canthium, 355

Cape-pigeon (Daption capensis), 511

Capercailzie, 282

Capparis sandwicensis (sandwichiana), 533, 553

Capsella bursa pastoris, 536, 567

Carapa, seed-buoyancy, 108, 114, 529;
beach and river drift, 76, 435, 437;
germination, 76, 78, 564;
station, 43, 550-1;
distribution, 68-9, 562

Cardamine hirsuta, 536

Cardamine pratensis, 536

Cardamine sarmentosa, 604

Cardiospermum halicacabum, 417, 605;
_see_ Additions and Corrections

Carduus, 28, 536

Carex, 37, 272, 283, 538, 540, 544

Careya, 575

Carmichael, Captain, on Tristan da Cunha, 276, 286

Carpenter, Captain, 490

Carruthersia, 265

Caryophyllaceæ;
represented in the early flora of Hawaii, 261-3, 518

Caspary, R., on the dispersal of water-lilies, 512

Cassia gaudichaudii, 533

Cassia occidentalis, 533

Cassia sophora, 605

Cassowaries as seed-dispersers, 152, 160

Cassytha filiformis;
fruit-buoyancy and dispersal by currents, 56, 57, 71, 106, 111,
121-2, 530, 552, 563, 569;
dispersal by birds, 71, 123, 564;
station and distribution, 56, 59, 67, 122, 551-2, 563;
extension inland, 42, 49, 59, 121, 547, 548, 559, 569

Castillo (Drake del), on the Tahitian flora, 46, 49, 221, 231, 254,
285, 318, 347, 504, 551

Casuarina, 134, 136, 479

Casuarina equisetifolia, 42, 45, 136, 530, 548-9

Casuarina nodiflora, 136, 549

Cattle Plains of Hawaii, 208

Centranthus ruber, 536

Centropogon, 251, 259

Cerastium vulgatum, 536

Ceratophyllum, 38, =398-9=, 408, 537

Cerbera odollam, distribution, 64, 563;
station, 551;
extension inland, 41, 42, 49, 121, =547=, 548, 569;
fruit-buoyancy and dispersal by currents, 76, 108, 114, 121, 530;
in river and beach drift, 76, 435, 437

Chærophyllum sylvestre, 28, 536

Chagres River, 498

Chamisso, 367

Chancay coast (Peru), 482

Charpentiera, 263

Cheeseman, Mr., on the flora of Rarotonga, 50, 177, 208, 216, 232, 238,
256, 291, 293, 295;
on Kermadec plants, 295, 420, 572;
on Polynesian food-plants, 415, 420

Cheirodendron, 262, 263, 281, 343, 364, 533, =594=, 604

Chelidonium majus, 535

Chenopodium, 272, 283, 284, 537

Cherry (Cerasus) stones in beach-drift, 429, 431, 479

Chestnuts in beach-drift, 429

Chile;
the coast plants and beach-drift, 431, 474-80, 596;
the coast climate, 491-3, 598-601

Chloridops kona, 275

Christmas Island, 422

Chrysanthemum leucanthemum, 536, 568

Chrysanthemum segetum, 536

Chrysodium aureum, 48, 486, 498

Chrysophyllum, 362

Chrysosplenium, 536

Cicuta virosa, 28, 536

Citrus aurantium, 125, 532

Citrus decumana, 125, 126, 532, 533

Citrus, other species, 125, 436, 532

Cladium mariscus, 537

Clarke, Mr. C. B., on Cyrtandra, 316

Cleome, 362

Clermontia, 252-6, 258-9, 533

Clerodendron, 17, 121

Clerodendron inerme, 47, 76, 108, 114, 435, 530, 551, 563

Climate and currents, 491-5, 500, 597-601

Clouds, altitude of, on mountains;
on the Owen Stanley Range, 215;
observations on the summit of Mauna Loa, 584;
on the Chilian and Peruvian sea-borders, 492, 494

Coast and inland species of a genus, relation of, 16, 27, 133-169

Cocculus, 362

Cochlearia officinalis, 33, 536, 540

Coco-de-mer, 61

Coco-nut palm (Cocos nucifera), 67, 108, 413, 530, 552, 553-4, 563;
effective dispersal by currents, 436;
viviparous, 472

Coix lachryma, 532

Collomia, 568

Colobanthus, 263

Colocasia antiquorum, 412

Colon;
shore-plants, 498

Colubrina, 134, =137=

Colubrina asiatica;
station and distribution, 137, 552, 556, 562-4;
inland extension, 49, 547;
seed-buoyancy and dispersal by currents, 56, 57, 105, =137=, 529;
river and beach drift, 436, 559

Colubrina oppositifolia, 137, 533-4, 576

Columbæ, extinct;
of the Mascarene Islands, 152, 157, 159, 169, 200, 517

Comins, Rev. R. B., 379

Commelina nudiflora (syn. pacifica), 533, 605

Commersonia echinata (syn. platyphylla), 532, 548

Commersonia, 376, 380

Compositæ, age of, 9, 231-49, 304, 306, 503, 514, 517-20;
endemic Hawaiian and Tahitian genera, 236, 248;
arborescent, 235-49;
dispersal by birds, 241, 593, 605;
fruit-buoyancy of British species, 536

Coniferæ, age of, 303-6, 502-3, 514, 519-20;
Fijian, 294, 297-306;
New Zealand, 507-9, 514

Conocarpus erectus, 68, 108, 438, 498

Convolvulaceæ, seed-buoyancy, 28, 110, 117, 544;
abortive germination of floating seeds, 76, 79, 83, 85, 87

Convolvulus arvensis, 28, 110, 537, 544

Convolvulus sepium;
station, 29;
distribution, 417-8, 573;
seeding in England, 539;
buoyancy of seeds, 26, 29, 106, 110, 537, 539, 544;
suggested dimorphism, 573

Convolvulus soldanella;
seed-buoyancy, 26, 28, 35, 36, 83, 91, 106, 110, 115, 537, 542, 566;
non-germination in sea-water, 35, 544, 546;
seeds in beach-drift, 31, 429-31;
dispersal by currents, 432-3;
distribution, 131, 433, 476-9, 542, 572;
suggested dimorphism, 573

Convolvulus tricolor;
germination in sea-water, 546

Conway, Sir Martin, 241

Coprosma, 270-2, 274-5, 290-3, =294-6=, 305, 315, 321, 331, 533

Corals on the coasts of Chile and Peru, 496-7, 601

Cordia, 17, 121, 134, 137, 485, 488

Cordia aspera, 137

Cordia subcordata;
station and distribution, 52, 551-2, 555, 563;
fruit-buoyancy, 79, 108, 114, 530-1;
sea and beach-drift, 79, 437;
becoming extinct in Hawaii, 578

Cordyline, 420, 532

Coreopsis, 237

Coriaria, 270, 290, =291-2=, 305

Corks in beach-drift, 479

Corral (Chile), 478

Corvus tropicus, 321

Corylus avellana, 126, 429 (hazel), 431 (hazel), 537, 538, 572

Corynocarpus, 508

Cotula plumosa, 241

Cotyledon umbilicus, 417-8, 536

Couthovia, 265-6, =401=, 532

Crambe maritima, 35, 536, 542, 581

Cratægus oxyacantha, 536

Cratæva religiosa, 379

Crepis, 536

Crinum asiaticum, 530

Crithmum maritimum, 28, 35, 109, 116, 429, 433, 536, =542=, 544

Crocodile, in Fiji, 65

Crosby, Mr., 550

Croton, vivipary, 472

Cucumis acidus, 532

Cucurbita, 125, 479, 532

Cupania, 532

Curcuma longa, 548

Curlews, as seed-dispersers, 355, 356

Currents, as seed-dispersers, 4, 12, 44, 49, 57, 58, 61-75, 79-84, 179,
557, 562, 571;
Gulf-stream, 4, 80, 180, 189, 430, 570, 581;
currents reaching Hawaii, 58, 72-5, 557-9;
Humboldt or Peruvian current, 480, 483, 490-5, 500, 597-601;
_see_ under Climate and currents

Cuscuta, 58, =366=, 537, 552, 553, 555, 559, =595=

Cyanea, 252, 254-5, 258, 604

Cyathodes, 272, =282=, 284, 285, 290-1, =292=, 305, 533, 553-4

Cyathula prostrata, 605

Cycas circinalis, 42, 109, 115, 328, 413, 530, 547, 548-9, 563

Cynometra, 108, 529, 563

Cyperaceæ, dispersal by ducks, 513;
by purple water-hens, 296

Cyrtandra, 308-9, =316=, 331, 405, 520

Cyrtosperma, 413

Cytisus scoparius, 536

Dacrydium, 294, 297-8, =302=, 305-6

Dalbergia monosperma, 106, 435, 529, 551, 563

Damasonium stellatum, 537

Dammara, 294, =298=, 303-6, 532

Daption capensis (Cape-pigeon), 511

Darwin, Mr., 24, 150, 177, 347, 497, 509, 538, 539, 542, 544, 546, 568

Datura stramonium, 537

Davis, Prof., 491

Davis, Rev. S. H., 213

Dead Sea, density of, 89

Death and Evolution, 11, 523

Delissea, 252, 254-5, 258

Density of sea-water and seed-buoyancy, 88, 566

Derris uliginosa;
distribution, 68, 552, 563;
station, 44, 551;
inland extension, 42, 547;
fruit-buoyancy, 106, 111, 529, 552;
river-drift, 76, 435;
beach-drift, 437

Deschampsia, 272, 275

Desmodium umbellatum, 106, 529

Desmodium polycarpum, 605

Deterioration of capacity for dispersal, 262-3, 337, 350, 365, 507,
594-5

Deyeuxia, 272, 284-5

Dianella, =356=, 533

Dichrocephala latifolia, 605

Dickson, Mr., 493

Didunculus, 159, 393

Didymocarpus, 318

Differentiation of birds and plants, 505-7, 514, 520, 521-2;
of climate, 470, 473, 507, 521-2

Dimorphism, suggested, in Rhizophora, 449, 465, 521;
in Cæsalpinia, 573;
in Convolvulus sepium and C. soldanella, 573

Dioclea, 82, 107, 113, 436, 529, 531, 563

Dioscorea, 412-4, 532

Dispersal, agencies of, 5, 61, 502;
suspension, causes of, 5-9, 242-3, 365, =504=, 514, 521-2;
_see_ under Adaptation, Birds, Currents, Deterioration &c.

Dixon, Mr. C., on differentiation of birds, 505-6;
on seeds in petrels, 581

Döderlein, 54

Dodo, 8, 159, 522

Dodonæa, 67, 71, 106, 333, =338=, 357, 529, 548, 554, 563

Dole, Mr., 241, 557

Dolicholobium, =394=, 603

Douglas, Mr. D., 278, 282, 586

Dove: _see_ Pigeon

Draba verna, 536

Dracæna aurea, =367=, 533

Dracæna, vivipary in, 471

Dracocephalum, 568

Dracontomelon, =399=, 532

Dragon-flies, dispersed by winds, 510

Drepanididæ, 259, 343, 348, 504-5, 603

Drepanocarpus, 562

Drift: _see_ Beach seed-drift, River seed-drift

Drift-timber, 58, 72, 557

Drosera, 4, 253, 270, 272, 285-8, 536

Druce, Mr., on Sea-thrifts, 34

Drupe, rubiaceous, its first appearance in the Pacific, 262

Drying-winds, 491-4

Drymispermum, 45, 133, 134, =164=, 166, 530

Dryobalanops, vivipary in, 471

Dubautia, 236-8, 240, 243-4

Ducie Island, 49, 64

Ducks, as dispersers of seeds, 241, 277, 356, =369=, =370=, 375, 399,
506, =512-3=, 514, 541

Dusolier, M., 506

Dwarfing of shore-plants growing inland, 547

Dysoxylum, 376

Easter Island, 64

Ecastaphyllum, 562

Eclipta alba, 533, 534, 605

Ecuador, climate of sea-border, 476, 483, 489-91, 493-6, 500, 597;
influence of the Humboldt or Peruvian current, 490-1, 493-5, 500,
599;
mangroves, 3, 77, 445, 448, 474-6, 483-90, 521, 597;
beach-drift and beach-plants, 180, 488-9;
temperature of the Guayaquil estuary, 77, 78, 565;
drift of the Guayaquil River, 91, 435, 488-9;
Santa Rosa River, 486;
Machala plains, 485, 495;
Santa Elena coast, 483, 490, 494, 597

Eeka, mountain in Maui, 208, 214, 253

Eggers, Baron von, on the climate and mangroves of Ecuador, 449, 450,
476, 483-4, 487, 490, 493-5, 597

Ekstam, O., on seed-dispersal in Spitzbergen and Nova Zembla, 242, 277,
282, 434, =511=, =512=

Elæocarpus, =334=, 357-8, 389, 391, 401, 507-9, 532

Elatine hydropiper, 536

Elatostema, 317, 383, 391, =405=

Electrical state of the atmosphere, 582, 588

Eleusine indica, 605

Elizabeth Island, 64

Embelia, 362-4, 409, 520

Empetrum nigrum, 511

Endemic genera and species, tables of, 232, 233, 244, 252, 255, 263,
265;
_see_ under Fiji, Hawaii, Tahiti

Endemism: _see_ under Species

English beach-drift, 429-33;
_see_ Beach seed-drift

Entada scandens, =176=;
station, 44, 48, 50, =177=, 182;
distribution, 68, =177-9=, 182, 200, 499, 500, 551, 563;
dispersed by currents, =179-80=, 182;
seed-buoyancy, 82, 94, 106, 111, =181-3=, 529, 531;
river and beach drift, =180=, 430, 435, 437-8, 488-9, 499

Epilobium, 471, 536, 603

Epiphytic plants, 281 (Vaccinium);
291 (Weinmannia);
343 (Myoporum and Vaccinium);
383 (Loranthus);
402 (Myrmecodia)

Eranthemum, 532

Eriophorum, 537

Erodium maritimum, 33

Eryngium maritimum, 28, 35, 536, 539, =543=, 544, 581

Erythrina;
general discussion of the genus, =577=;
relation between coast and inland species, 19, 134, =141=;
seed-buoyancy, 107;
seeds in river and beach drift, 435, 437-8, 489

Erythrina indica, 19, 107, =141-4=, 529, 562, =577=

Erythrina monosperma, 20, =141-4=, 200, 533, 553-4, 578

Erythrina ovalifolia, 107, 577

Erythrina aurantiaca, 144

Erythrina vespertilio, 144

Estuaries, temperature of tropical, 78, 564

Etna, shadow of, 586

Eucalyptus, 479

Eugenia, 61, 134, =163=, 166, 334, =349=, 357, 507, 532, 603

Eugenia brackenridgei, 351

Eugenia corynocarpa, 350, 351, 532

Eugenia grandis, 163

Eugenia malaccensis, 349, 532

Eugenia monticola, 351

Eugenia neurocalyx, 350, 351

Eugenia rariflora, 164, 349-52, 532

Eugenia richii, 45, =164=, 350, 529

Eugenia rivularis, 350, 532

Euonymus europæus, 536

Eupatorium cannabinum, 536

Euphorbia, 29, 478, 537

Euphorbia amygdaloides, 537, 605

Euphorbia atoto, 106, 530

Euphorbia cordata, 556

Euphorbia helioscopia, 29, 537

Euphorbia paralias, 29, 31, 35, 109, =116=, 429, 431-3, 537, =543=, 544

Euphorbia peplus, 537

Euphorbia segetalis, 537, 605

Eurya, 371-3

Euxolus caudatus, 605

Everett, Mr., 388

Evolution, 11, 227, 522;
_see_ under Adaptation, Natural Selection, and Species

Excæcaria agallocha, 43, 109, 436, 530, 551, 563

Exocarpus, 271, 274-5

Fagræa berteriana, 42, 45, =385=, 532, 548, 603

Fagus, 508

Fanning Island, 377

Fernando Noronha, its flora and its birds, 8, 144, 366, 388

Ferns, =220-30=, 509, 517, 592, 593

Ficus, 61, 377, =387=, 395, =504=, 532

Fiji, the climate, 209, 215-8;
the seed-drift of rivers and beaches, 435-6;
the “talasinga” plains, 42, 43, 215, 386, 547, =548=, 569;
area and altitudes, 207-8

Fijian flora, the littoral plants, 13, 40, 528-30, 550-1;
ferns and lycopods, 220-30, 592;
endemic genera and endemic species, 231-5, 264-6;
mountain plants, 269, 293-7, 305;
conifers, 297-306
The age of Malayan plants comprising first the genera widely
dispersed over the Pacific, 307-58;
and then those that are locally dispersed, 359, 360;
the last comprising Fijian genera found in Hawaii and not in
Tahiti, 359, 371-4;
Fijian genera found in Tahiti and not in Hawaii, 359, 380-8;
and genera found in Fiji, but neither in Hawaii nor in Tahiti, 360,
399-408 ...
The number of Fijian flowering plants, 528, 592 ...
The Fijian “difficulty,” 155, 166, 169, 517 ...
The buoyancy of the seeds of inland plants, 531

Fish in the Humboldt current, 600

Fitchia, 236, 237-8, 240, 245, 248

Fleurya interrupta, 605

Flying-foxes: _see_ Bats

Focke, W. O., on the dispersal of Leguminosæ, 150, 417

Fogs on the Chilian and Peruvian coasts, 475, 481, 490, 492-3, 600

Forbes, Dr. H. O., 304, 347, 388

Forster on the “Antarctic” flora, 271

Fragaria chilensis, 4, 270, 272, 285, 287-8

Fragaria vesca, 536

Francolins at high altitudes, 241

Franseria, 131, 431, 478-9, 597

Freycinetia, 308-9, =319=, 331, 504-5, 509-10

Frigate-birds (Fregata), as seed-dispersers, 188, 511

Fritillaria meleagris, 537

Fruits “difficult” or “impossible” from the standpoint of dispersal,
262, 267, 372-3, 379, 388-9, 391, 409;
_see_ under the Fijian flora and the Hawaiian flora Fruit-pigeons
(Carpophaga, &c.), seeds and fruits eaten by them, limit of size,
337, 381, 389, 400-1, 410;
Achras, 373;
Areca, 330;
Cananga, 393;
Canarium, 372, 400;
Cassytha, 123, 564;
Coriaria, 292;
Corynocarpus, 508;
Couthovia, 266, 401;
Dracontomelon, 372, 400;
Elæocarpus, 337, 372, 508;
Eugenia, 350;
Ficus, 388, 504;
Gnetum, 404;
Kentia, 330;
Litsea, 508;
Myristica, 403;
Olea, 508;
Oncocarpus, 266;
Phyllanthus, 326;
Pisonia, 347;
Plectronia, 355;
Podocarpus, 301, 508;
Premna, 561;
Psychotria, 314;
Sapota, 373;
Sideroxylon, 373;
Spondias, 602;
Vitex, 564;
Ximenia, 113.
(For additional note on Litsea and Vitex, _see_ Additions and
Corrections)

Fruits, fleshy, 101

Fulmar-petrel, 581

Galapagos Islands, flora and avifauna, 505

Galeopsis tetrahit, 28, 537

Galium aparine, 27, 536, 539

Galium mollugo, 536

Galium palustre, 27, 536

Garcinia, 575

Gardenia, 308-9, =311=, 532-3, 548, 552

Garuas, 475, 481, 600

Gätke, H., 151, 506, 593

Gaudichaud, 385, 434

Gay, C., on Chilian plants, 477-8

Geese, as seed-dispersers, 241, 275, 283, 356, 511-2, 514

Geissois, 343, =393=, 510, 532

Geniostoma, 384

Geological time, 502, 520

Geophila reniformis, 417, 532, 605

Geranium, 4, 269, 272, 274-5

Germination, of floating seeds, 76;
its connection with vivipary, 78, 84, 87, 191, 468, 521;
effect of previous immersion in sea-water on the germinating capacity
of seeds in soil, 25, 539;
germination in sea-water in temperate latitudes, 35, 544;
germination in sea-water in tropical latitudes, 79-87;
the process in Cæsalpinia, 191, and in Cuscuta, 595

Giffard, Mr., 212

Gilia, 568

Gill, Rev. Wyatt, 48

Gizzard-stones in birds, 159

Glacial epoch, 503, 509, 513

Glaucium luteum, 35, 90, 536, 539, 543 544

Glaux maritima, 34, 36, 537, 541

Gleichenia, 548

Gnetum, 391, =404=, 532

Goats as seed-dispersers, 554, 558

Goebel, K., 453;
on vivipary, 469-72

Gossypium, 352, 533

Gossypium tomentosum, 58-9, 529, 533, 552, 556

Gouania, 371-3

Gouldia, 262-3

Goura pigeon, 8, 159, 392

Gourds dispersed by currents, 125, 570

Græffea, 265

Grape-seeds in beach-drift, 429

Gray, Asa, 34

Grewia, 381, 382, 532

Grindon, Mr., 603

Grouse-family as seed-dispersers: _see_ Tetraonidæ

Guavas, 554

Guayas or Guayaquil river and estuary: _see_ under Ecuador

Guettarda, relation between coast and inland species, 17, 121, 134,
=162=, 166, 532, 533;
germination, 79, 132

Guettarda speciosa, station at coast, 43, 551;
station inland, 162;
distribution, 64, 68, 563;
buoyancy of fruits and their dispersal by currents, 55-6, 108, 114,
163, 529;
beach-drift, 163, 437

Guilandina (synonym of Cæsalpinia), 188, 562

Gulf-stream drift: _see_ under Currents

Gulls (Laridæ) as seed-dispersers, 241, 511, 514

Gum-resins, native names of, 300

Gunnera, 269, 271, 272, 274, 275

Gunnerus, 570

Gyrocarpus jacquini, 2, 49, 68, 106, 111, =422=, 428, 529, 551, 563

Haberlandt, G., on the germination of mangroves, 453, 457, 465

Haleakala, 208

Hall, Mr. W. L., on the forests of Hawaii, 213

Halophily, 581

Hamilton, Mr., on the crop-stones of the moa, 159

Hann, Prof., on mountain climates, 582

Haplopetalon, 264, 265

Haplostachys, 263, 594

Hawaiian islands;
area and altitude, 207-8;
climate, 209-15, 217-8, 582-92;
temperature and relative humidity, 209-11, 218;
rainfall, 212-15, 218;
meteorology of the summit of Mauna Loa, 582-92

Hawaiian flora;
littoral plants, 15, 51, 71, 552, 553-7;
beach-drift, 58, 557;
ferns and lycopods, 220-30, 517, 592, 593;
the eras of the flowering plants, 234-5, 517;
tables of endemic genera and endemic species, 232, 233, 244, 252,
263;
the age of Compositæ, 235-49;
the age of the Lobeliaceæ, 250-60, 266-7, 603;
the endemic genera belonging neither to the Compositæ nor to the
Lobeliaceæ, 261-4, 594;
mountain flora, 269-88, 518
The age of Malayan genera, 307-78;
the Malayan genera widely dispersed in the Pacific, 307-58;
the Malayan genera locally dispersed, 359-69;
the residual genera (found only in Hawaii), 359, 361;
the genera occurring in Hawaii and Tahiti but not in Fiji, 359,
370;
the genera occurring in Hawaii and Fiji but not in Tahiti, 359,
371;
the absentees from Hawaii, 359, 375
American plants in Hawaii, 261, 334, 362, 372, 409, 517-8.
They include the following orders and genera:—
Compositæ, 237, 248, 260, 518;
Lobeliaceæ, 254, 260, 266, 518;
Caryophyllaceæ, 261-3, 518;
Sanicula, 272-3, 287;
Sisyrinchium, 272-3, 287;
Fragaria, 285, 288;
Rubus, 273, 285;
Pritchardia, 309, 326;
Lythrum, 362;
Perrottetia, 362;
Sicyos, 362, 365;
Chrysophyllum, 362;
Nama, 362;
Jacquemontia, 362, 365;
Sphacele, 362;
Phytolacca, 362, 364;
Urera, 362
The Hawaiian difficulty, 140, 165, 168, 517;
Hawaiian plants and birds, 505-6, 603;
buoyancy of fruits and seeds of inland plants, 533-4

Haynaldia, 251

Hazel: _see_ Corylus avellana

Hazlewood’s Fijian Dictionary, 400

Heather, 511

Hedera helix, 536

Hedley, Mr., 65, 66, 304

Hegelmaier, on Lemna, 408

Heilprin, Prof., 506

Helianthemum vulgare, 536, 567

Heliosciadium, 538

Heliotropium anomalum, 49, 56, 58, 365, 370, 528, 552, 554-7

Heliotropium curassavicum, 56, 58, 482, 552, 555, 557

Hemsley, Mr. W. B., 34, 232, 238, 301, 434, 499, 511, 530, 539, 581;
on dispersal by currents, 62;
on Cæsalpinia, 184;
on insular Leguminosæ, 198-9;
on the Pacific Compositæ, 242, 246-7;
on the Pacific Lobeliaceæ, 251, 257;
on Pisonia, 346;
on Gnetum, 404;
on ten widely spread British plants, 417;
on Rhizophora mangle in the Pacific, 441;
on vivipary in Dracæna, 471;
on Brackenridgea, 569

Heptapleurum, 263

Heritiera littoralis, station, 43, 551;
distribution, 47, 68, 375-6, 551, 562;
buoyancy of fruits, 106, 112, 529;
their occurrence in river-drift, 79, 435;
and in beach-drift, 437

Hernandia peltata, station, 551;
extension inland, 49;
distribution, 54, 64, 68, 562, 563;
buoyancy of fruits, 109, 115, 530;
beach-drift, 437

Herpestis monnieria, 552, 557

Hesperomannia, 236-7, 243-4

Hesselman, H., on plant-dispersal, 282, 511

Hibbertia, 402, 548

Hibiscus, 21, 134, 137

Hibiscus abelmoschus, 21, 124, 532, 533

Hibiscus diversifolius, 21, 105, 529

Hibiscus esculentus, 21, 532

Hibiscus tiliaceus, station, 43, 486, 498, 551, 555-7;
growing inland, 41-2, 419, 547, 557, 560;
distribution, 52, 486, 498, 552, 563;
seed-buoyancy, 21, 105, 529, 552;
beach-drift, 437-8;
river-drift, 435;
sea-drift, 489;
currents, 562

Hibiscus youngianus, 21, 533, 534

Hillebrand, Dr., on the Hawaiian flora;
introduced littoral trees, 51;
the agency of the currents, 73;
the endemism of the ferns, 224-7, 593;
the Compositæ, 236-47;
the Lobeliaceæ, 250-5;
the absence of Coniferæ, 303;
on Cyrtandra, 317;
other references, 231, 340, 341, 405, 557, 559, 560, 576, 578, 594,
&c.

Hillebrand, Mr. W. F., 224

Hillebrandia, 263, 394

Hilo, 53, 213

Hippomane mancinella (Manchineel), 68, 109, 498

Hochstetter, 336

Holland, Mr., 478, 489, 499

Holmboe, Jens, on Silene maritima, 280

Holmes, Mr., on the Fijian rainfall, 216, 549

Honckeneya: _see_ Arenaria peploides

Hooker, Sir J., 198, 238, 258, 271, 344, 369, 581

Horne, Mr., on the Fijian flora, 41, 172, 394, 395, 425, 549, 592

Horse-dung containing seeds transported by currents, 558

Hottonia palustris, 536

Hoya, 377

Humboldt on insects in the upper air-currents, 583

Humboldt current, 480, 483, 490-5, 500, 597-601

Hydnophytum, 402

Hydrocharis, 537

Hydrocotyle asiatica, 532, 605

Hydrocotyle verticillata, 533-4

Hydrocotyle vulgaris, 28, 536, 544

Hygrophytes, 32, 515

Hypericum, perforatum, quadrangulum, elodes, 536

Iceland;
plants, 505

Ilex aquifolium, 536

Impatiens fulva, 536

Impatiens parviflora, 536

Incas;
bones exposed on the Ancon plain, 497

Inocarpus edulis, 108, =421=, 435, 529, 563

Insects in the upper air-currents, 509-10, 514, 582-3

Ipomea: _see_ Ipomœa

Ipomœa, 20, 76, 109, 134, 137, 489, 532-3, 546

Ipomœa batatas, 20, 415 (Batatas edulis), 532

Ipomœa bona nox (Calonyction speciosum), 20, 106, 110, 533, 534, 554,
558, 605

Ipomœa glaberrima (Calonyction comosperma), 20, 52, 57, 106, 110, 472,
530, 552, 555, 558, 564

Ipomœa grandiflora, 20, 106, 530, 531

Ipomœa insularis, 20, 110, 532, 533, 553, 554, 605

Ipomœa peltata, 110, 435, 472, 532

Ipomœa pentaphylla, 20, 110, 533

Ipomœa reptans, 533

Ipomœa tuberculata, 20, 110, 533, 556

Ipomœa turpethum, 20, 106, 110, 532

Ipomœa pes-capræ;
seed-buoyancy, 20, 21, 83, 106, 110, 121, 530, 546, 569;
dispersal by currents, 56-7, 562;
seeds in river and beach drift, 435, 437-8, 489, 558-9;
distribution, 56, 68, 433, 488, 498, 552, 563, 564, 572;
station, 43, 551, 553-6, 560;
growing inland, 21, 41, 42, 121, 547, 560, 569

Iris;
home of the genus, 396

Iris fœtidissima, 24, 27, 537

Iris pseudacorus, 24, 27, 30, 430, 537, 540, 544

Isodendrion, 263

Isotoma, 255

Jacquemontia, 58, =365=, 533, 552, 553, 555-7

Jacquin, on Rhizophora mangle, 457

Jambeli Island, 488-9

Johnston, Sir H., 241, 252, 510

Jouan, H., 187

Juan Fernandez, 222, 316

Juncus, 89, 471, 537, 545, 568

Junghuhn, on the climate of the Java mountains, 211

Jussiæa villosa, 533

Kadua, 262, 263

Kandavu, 207, 208

Karsten, G., on Rhizophora, 453

Kauai, 207, 208, 214

Kauri pine, 299

Keeble, F. W., on the dispersal of Loranthus, 383

Keeling atoll, 81, 190, 510

Kerguelen, 241, 242, 276

Kermadec islands, 258, 276, 295, 420, 572

Kerner, Dr., 63, 101, 277, 408, 469, 567

Kidder, Dr., 241, 276

Kilimanjaro, 251

Kinabalu, Mount, in Borneo, 286, 295-6, 301

Kirk, T., on the forest-flora of New Zealand, 299, 347, 507

Kittlitz, on Nipa fruticans, 66

Kiwi, 522

Kleinhovia hospita, 21, 105, 376, 529, 562, =602=

Koebele, Prof., 212, 578

Kolpin-Ravn, on seed-buoyancy, 24, 38, 116, 538

Krakatoa beach-drift, &c., 180, 190, 206, 221

Krämer, Dr., 275, 283

Kurz, Mr., 577;
on Scirpodendron, 406

Labiatæ;
station and seed-buoyancy, 28, 537;
mucosity of seeds, 567-8;
Hawaiian endemic genera, 261-3, 518, 594

Lablab vulgaris, 413, 417, 605

Labordea, 262, 263

Labrador current and climate, 493

Lagenaria (bottle-gourds), dispersal by currents, 125, 570

Lagenophora, 240, 242, 270-2, 274-5, =276=, 293-4, 305, 568

Lagerheim, Prof., on the dispersal of Empetrum nigrum, 512

Lagopus: _see_ Tetraonidæ

Laguncularia;
station and distribution, 68-9, 484, 498;
dispersal by currents, 77;
fruit-buoyancy, 108;
fruits in beach-drift, 438;
in river-drift, 489;
in sea-drift, 489;
germination, 77-8, 469, 471

Lamium album, 28, 537

Lamium galeobdolon, 537

Lamium purpureum, 28, 537, 568

Lapsana communis, 536

Larch, 430

Lathyrus maritimus, 35, 107, 116, 430, 432, 536, =543=

Lathyrus pratensis, 536

Layard, Messrs., on birds and seeds, 143, 296, 388, 420

Leersia oryzoides, 538

Leguminosæ in littoral floras, 13, 19, 68, 76, =79-85=, 87, =107=, 111,
117, =140-51=, =170-97=, =198=, 438-9;
dispersal by birds, 150, 417, 581

Lemnaceæ, =407=, 488, 537

Leontodon autumnalis, 536

Lepidium sativum, 567

Lepinia tahitensis, 378-9

Leucæna forsteri, 425, 529

Leucas decemdentata, 605

Ligustrum vulgare, 537

Limnanthemum, =396=, 537

Linaria cymbalaria, 537

Linaria vulgaris, 537

Linden, A., on Pritchardia, 326-7

Lindenia, =395=, 532, 603

Lindman, C., on the Gulf stream drift of the Scandinavian coast, 180,
430, 531

Linum, 536, 567

Lipochæta, 58, 236-7, 239, 243-4, 552-3, 556-7

Lister, Mr., 327, 356, 550

Lithospermum officinale, 537

Litsea, 508;
_see_ Additions and Corrections

Littoral plants;
distribution in tropics, 68, 516, 562;
causes of the buoyancy of the seeds and fruits, 104, 119, 569;
long flotation experiments, 530;
the littoral plants and the currents of the Pacific, 61;
relation of littoral and inland species of the same genus, 130-70;
the inland extension of beach plants, 34, 40, 49, 59, 121, 547-50,
559, 568, 579;
Fijian shore plants, 13, 15, 40, 528, 547-50;
Tahitian shore plants, 14, 47, 551;
Hawaiian shore plants, 15, 51, 552-7, 559, 563;
shore plants of west coast of South America, 474-88, 500, 596;
shore plants of the Panama isthmus, 498;
British shore plants, 33-6, 106, 107, 109, 115, 432-3, 540-4;
halophily, 581

Lobelia, 252-6, 258, 272-3, 279

Lobelia dortmanna, 536

Lobeliaceæ, Age of, 250-60, 266-7, 304, 306, 517-20;
arborescent, 250-60;
Hawaiian, 253;
Tahitian and Rarotongan, 256;
affinities with American forms, 251, 267;
relation between the flowers and birds, 504-5, 603;
seed-buoyancy, 533;
capacities of dispersal, 258

Locusts as seed-dispersers, 509

Loranthus, 377, 381, =383=

Lotus corniculatus, 536

Luffa, =426=, 472, 529, 563, 571

Lumnitzera, 43, 69, 108, 435, 437, 529, 551, 563

Luzula campestris;
in Hawaii, 272, 286-8;
in Tahiti, 290, 292;
mucosity of the seeds and their dispersal by birds, 286, 417, 568;
seed-buoyancy, 89, 537

Lychnis diurna, 536

Lycopods, 220-30, 509, 517, 592, 593

Lycopsis arvensis, 537

Lycopus europæus, 28, 37, 86, 417-8, 537, 545, 568

Lyon, Prof., on the Hawaiian rainfall, 213

Lysimachia;
in Hawaii, 272, 283-5

Lysimachia thyrsiflora, 537

Lysimachia vulgaris, 536

Lythrum;
in Hawaii, 362

Lythrum salicaria, 417-8, 536, 603

Maba, 371-3, 532-3

Macaranga, 381, =384=, 436, 532

Macgregor, Sir W., 215

Machala plains (Ecuador), 484-5, 495

M’Lachlan, Mr., on dragon-flies in ancient drift, 510

Magellan Straits, its shore plants, 477

Maiden, Mr., on the plants of Pitcairn Island, 345, 355, 418, 562

Malayan era of the Pacific floras, 308-9, 330-2, 333, 353, 357, 359,
519

Malva rotundifolia and sylvestris, 536

Malvaceæ, the effect of sea-water on the seeds, 544

Man in the Pacific, subject to the laws of distribution, 325, 411-2,
427

Manchineel: _see_ Hippomane

Mandarin orange;
precocious germination of its seeds, 472

Mangle chico, 445, 448-9, 498, 501

Mangle grande, 445, 448-9, 498, 501

Mangrove formation, 9, =43-4=, 47, 50, 53-5, 77, 132, =483-7=
(Ecuador), =551=, 597

Marquesas, 208, 529

Marsilea, 408

Martins, Prof., on the effects of sea-water immersion on seeds, 24,
538-9, 542-4, 546

Mascarene islands;
association of several species of Pandanus with extinct Columbæ, 152,
157, 169, 517;
linked to the Pacific islands by Afzelia bijuga, 172;
by Acacia heterophylla, 200;
by Naias marina, 368;
by Potamogeton, 369;
by Eugenia, 351;
by Sophora, 148;
by Ochrosia, 153, 580;
by Pandanus, 157-9

Mat-names and plant-names in Polynesia, 324, 328

Matricaria chamomilla, 536, 568

Matricaria inodora;
inland form, 536, 543;
maritime form, 33-4, 109, 116, 536, 540, =543=

Matricaria maritima: _see_ under M. inodora

Maui, 207-8

Mauna Kea, 207-8, 210-4, 238, 509, 587

Mauna Loa, 207-8, 210-4, 238, 509, 587

Mauna Loa;
meteorological observations on summit, 582

Mauritius: _see_ Mascarene islands

Maxwell, Dr., on the Hawaiian rainfall, 213

Medanos, 482

Medicago;
fruits frequent in beach-drift, 429, =431=, 479;
buoyancy of fruits, 431, 536

Medinilla, 376

Mediterranean beach-drift, 430

Megapodes, as probable seed-dispersers, 160, 169, 200, 392

Melastoma, 376, 381, =382=, 532, 603

Melastoma denticulatum, =382=, 532, 548

Melia, 376

Meliaceæ, 376

Melica nutans, 538

Melicope, 263

Meliphagidæ (Honey-eaters), as seed-dispersers, 321, 331, 377, 388

Melochia, 376

Memecylon, 376

Mentha aquatica, 28, 537, 545

Menyanthes trifoliata, 537

Meryta, 381

Mesembryanthemum, 131, 478

Mesozoic continent in the Western Pacific, 303-6, 503, 509, 514, 519

Metrosideros, 259, 333, =341=, 357, 361, 510, 533, 549, 553, 554, 604

Meyer, Dr., 388

Mez, C., on the genus Embelia, 363

Mezoneuron, =146=, 363, 533

Micromelum, 393, 532

Miers, Mr. J., on the seed-structure of Barringtonia, 575

Millett, Mr., on the seed-buoyancy of Convolvulus soldanella, 91, 543,
566

Milner, Sir W., on seeds in petrels, 581

Mimosa pudica, 488

Mimusops kauki, 374

Miquel, on Scirpodendron, 406;
on Carapa, 564

Moa, its crop-stones, 159

Mœnchia erecta, 536

Mollendo (Peru), 482, 598, 599

Momordica charantia, 532

Montia fontana, 536

Moreno Bay (Chile), its beach-drift, 480

Morinda;
relation between the shore and inland species, 18, 134, =135=;
inland species, 42, 45, 135, 532, 534

Morinda citrifolia;
distribution, 18, 52, 552, 563;
station, 551;
inland extension, 49, 121, 547, 548, 553, 569;
buoyancy of pyrenes, 18, 107, 112, 121, 123, 124, 529, 531, 534, 569;
their occurrence in river-drift, 79, 435

Moseley, Prof., 242, 276, 286, 403, 434, 579

Mountain bananas, 412, =414=, 427

Mountain climates of the Pacific islands, 210-1, 214-5, 218, 582

Mountain ferns of the Pacific islands, 225-9, 593

Mountain floras of the Pacific islands, 268-306, 518-9

Mountain-shadows, 586

Mucosity of seeds, 102, 277, 417, 567

Mucuna (genus);
abortive germination of seeds in sea-water, 76, 79-82, 202;
long flotation experiments, 80, 81, 531;
cause of seed-buoyancy, 106, 109, 111;
dispersal by currents, 80, 81, 430;
seeds in river-drift, 435, 488;
in beach-drift, 430, 437-8, 489, 499

Mucuna gigantea, 44, 81, 94, 109, 115, 529, 531, 552, 554, 562, 563

Mucuna urens, 80, 106, 111, 123, 430 499, 531, 533, 534, 562, 563

Mueller, Dr. K., 224

Mueller, Baron von, 394

Musa, 412, 414

Musa ensete, 414, 436

Mussænda frondosa, 425, 532, 548

Mutations, 383

Myoporum, 271, 272, 274-5, 343, 533, 553-4

Myosotis, 37, 537

Myriophyllum, 37, 536

Myristica, 391, =402=, 532

Myrmecodia, =402=, 472, 532

Nadeaud, Dr. J., on Tahitian plants, 216, 312, 426, 578

Naias, 362, 364, =367=, 409, 520, 537, 556;
_see_ Additions and Corrections

Nama, 362

Narthecium, 537

Nasturtium, 37, 536

Nathorst, Prof., 511

Natural Selection, 105, 117, 119-29

Nature-Study, 101

Nelitris, 381, =382=, 532, 548

Nepeta cataria, 28, 537, 567, 568

Nepeta glechoma, 28, 537, 567, 568

Nephelium, 161, 532

Neraudia, 263

Nertera depressa, 270, 272, 285, =286=, 288, 290, 292-4, 296-7, 305

Nesopanax, 265

Nestor meridionalis, its vegetable food, 321, 343

New Caledonia, 298, 301-2, 341, 395-6, 420

New Zealand birds and seed-dispersal, 296, 321, 337

New Zealand flora;
from the standpoint of dispersal, 507-8;
its bearing on the continental theory, 508-9, 512, 514

New Zealand plants in the Pacific islands, 271-2, 287, 290, 295, 297,
305, 315, 336-8, 341, 357, 366, 503-4, 507-9

Nicobar-pigeon, as a seed-disperser, 8, 159

Nipa fruticans, 66, 68, 108, 472, 550

Nolana, 131, 431, 477, 479, 480, =596=

Norman, J. M., on Scandinavian beach-drift, 430, 432, 539, 543

Norwegian beach-drift: _see_ Scandinavia

Nothocestrum, 262-3

Nototrichium, 262-3

Nova Zembla;
plant-dispersal, 511

Nuphar luteum, 37, 89, 513, 535

Nymphæa alba, 37, 89, 513, 535

Oahu, 207-8

Oak: _see_ Quercus

Ochrosia;
relation of coast and inland species, 134, =151=;
dispersal, 152-4;
Schumann’s enumeration of the species, 580

Ochrosia parviflora, 49, 108, =152-4=, 530, 563, 580

Ochrosia sandwicensis, 153, 580

Ocimum basilicum, 568

Œnanthe crocata, 28, 536

Œnanthe phellandrium, 536

Oldenlandia, 605

Olea, 364, 508, 533

Oliver, Prof., 171, 178, 251

Oncocarpus, 265-6;
_see_ Additions and Corrections

Ononis arvensis, 536

Ophiorrhiza, 382, =383=, 532

Orchids;
suggested dispersal of their seeds by insects, 509

Oregon drift on the Hawaiian coasts, 58, 72, 430, =557=

Oreobolus, 271, 272, 275

Oreodoxa, 489

Osmanthus, 364

Osteomeles, 353, =354=, 554

Owen Stanley range;
climate of the summit, 215

Oxalis acetosella, 536

Oxalis corniculata, 349, 416, 427, 536, 604

Oxyria digyna, 512

Pachyrrhizus, 412, 413, 548

Paita (Payta), 482, 599

Palms;
specific differentiation in Borneo, 504

Panama isthmus;
its shore plants and seed-drift, 180, 498

Panax, 263

Pandanus, =155=, 508, 517;
relation between coast and inland species, 134, =155-60=, 166, 169,
517;
mode of dispersal, 157-60, 169;
insular distribution, 156-7, 160, 169, 580

Pandanus odoratissimus;
distribution, 52, 53, 156, 552, 563;
station, 551;
inland extension, 41, 42, 361, 548-9, 553-6, 560;
dispersal by currents, 53, 158;
buoyancy of fruits, 109, 530, 552;
their occurrence in river-drift, 435, and in beach-drift, 437, 557-9;
aboriginal food-plant, 413, 427

Pangerango, Mount;
rate of decrease of temperature with elevation, 210

Panicum, 272, 284

Papaver, seed-buoyancy, 535

Papaya, 472

Paphia, 265

Parinarium, 108, 435, 529, 563

Parrots, as seed-dispersers, 321, 388, 420, 602

Partridges, as seed-dispersers, 284, 356, 367, 416

Pastinaca sativa, 28, 536

Peach-stones in beach-drift, 429, 431, 479

Pea-nuts in beach-drift, 479

Peale, Mr., 355

Pebble-swallowing by birds, 8, 159

Pedicularis palustris, 537

Pelagodendron, 265

Pelea, 263

Pemphis acidula;
station, 42, 43, 551;
distribution, 54, 68, 563;
seed-buoyancy, 108, 114, 529

Penzig, Prof., on the beach-drift and flora of Krakatoa, 180, 189, 206,
571

Peperomia, 334, =348=, 357, 417, 509

Peplis portula, 536

Perkins, Mr., on birds and seeds in Hawaii, 151, 259, 275, 321, 329,
343, 348, 364, 595;
on the biological connection between the birds and the arborescent
Lobeliaceæ, 255, 504, 603;
on the Hawaiian fauna, 505

Perrottetia, 362

Peru;
littoral flora, 474-6, 481-2;
on the coast climate and the Humboldt current, 490-4, 500, 598-600;
corals, 496, 601

Pes-capræ formation, 550

Petrels and seed-dispersal, 242, 511, 581

Peucedanum, in Hawaii, 362, 363

Peucedanum palustre, 536

Phaethon (Tropic-bird) and plant-dispersal, 241

Phaseolus truxillensis, 605

Philippi, on the shells of Chile, 496

Phyllanthus, 309, =325=, 331, 532

Phyllostegia, 262, 263, 371, 533, 594

Physalis angulata, 605

Phytelephas, 489

Phytolacca, 362, 364

Pigeons, 356, 416, 506;
_see_ Fruit-pigeons

Pilea, 362

Pilger, R., on Podocarpus and Dacrydium, 301-2

Pimia, 265, 266

Pinguicula lusitanica, 537

Pinus sylvestris, 537

Piper, 532, 568

Pipturus, 356

Pisonia, 61, 333, =346=, 357, 568

Pistia, 435, 486-9

Pitcairn Island, 64, 345, 355, 418, 562

Pittosporum, 308, =309=, 313, 509, 532

Platydesma, 263

Plantago;
on tropical mountains, 269, 270;
Hawaiian species, 271-2;
mode of dispersal, 276;
seed-mucosity, 276, 568;
seed-buoyancy, 537, 540

Plantago coronopus, 33

Plantago lanceolata, 276, 537, 568

Plantago major, 276, 537, 568

Plantago maritima, 34, 478, 537, 541, 568

Plantago media, 537

Plectronia, =355=, 533

Plectrophanes nivalis (Snow-bunting), as seed-disperser, 510, 511, 605

Pleiosmilax, 371-3, 532

Plerandra, 265

Plum-stones in beach-drift, 429, 431, 479

Poa;
Hawaiian species, 272, 275

Poa aquatica, 538

Poa fluitans, 538

Podocarpus, 294, 297-8, =301=, 306, 508, 603

Polycarpon tetraphyllum, 33

Polygala vulgaris, 536

Polygonum;
dispersal by birds, 356;
buoyancy of fruits, 537, 543

Polygonum amphibium, 537

Polygonum aviculare, 356, 537

Polygonum convolvulus, 356, 537

Polygonum glabrum, 354, 356, 435, 487-8

Polygonum hydropiper, 37, 537

Polygonum lapathifolium, 537

Polygonum maritimum, 35, 477, 537, 543

Polygonum persicaria, 356, 537

Polygonum viviparum, 512

Polymorphous species, discussed, 33-4, 353, 357-8, 373, 381, 391, 519,
520, 522;
independent of insular conditions, 363, 368, 520;
noted under Elæocarpus, 335;
Dodonæa, 339;
Metrosideros, 341;
Alyxia, 345;
Alphitonia, 346;
Pisonia, 346;
Wikstrœmia, 348;
Sicyos, 363, 365;
Naias, 368;
Eurya, 372;
Maba, 372;
Grewia, 382;
Nelitris, 382;
Melastoma, 382;
Loranthus, 383;
Geniostoma, 384;
Macaranga, 384;
Tabernæmontana, 385;
Bischoffia, 386;
Micromelum, 393;
Limnanthemum, 396;
Weinmannia, 291;
Vaccinium, 280-1

Polynesians;
their differentiation, 325, 411, 427

Polynesian food-plants, 412-4, 427

Polynesian weeds, 415, 427, 604

Polynesian plant-names, 66, 324, 328, 341, 345, 387, 398, 414, 419,
421, 424, 425, 441, 578

Polypodium, 593

Pongamia glabra, 54, 106, 202, 529, 551, 563, 581

Pontederia, 435, 486-9

Porphyrio (Purple Water-hen), as concerned in plant-dispersal, 296,
305, 321, 331

Portulaca, 532, 552, 553-5

Potamogeton, 5, 30, 38, =369=, 513, 537

Potentilla;
buoyancy and station, 27

Potentilla comarum, 27, 536

Potentilla tormentilla, 27, 536

Premna;
the genus in the Pacific, =560=;
buoyancy and station, 19, 134, 139;
cause of buoyancy of the fruits of the coast species, 107, 112, 123,
124, 530, 532, 561, 569;
inland extension of the coast species, 42, 547, 548, 561, 569;
distribution of the species, 561, 563;
occurrence of the fruits in river-drift, 112, 435, 561;
modes of dispersal, 561

Prioria copaifera, in Panama beach and river-drift, 499

Pritchardia, 124, 308, 309, =326=, 532, 533-4, 554-5

Prosopis dulcis, 557

Prunella vulgaris, 28, 417, 537, 568

Pseudomorus, 371

Psittacirostra, 321, 505

Psychotria, 308-9, =314=, 331, 391, 532, 603

Pteris aquilina, 225, 548

Pteropidæ: _see_ Bats

Pterotropia, 262-3, 595

Ptilotus, 263

Ptychosperma, 389, 532

Puerto Bolivar (Ecuador);
sojourn of the author, 476, 484-8, 494

Pumice in beach-drift, 429, 558, 601

Puna Island (Ecuador), 489, 494

Puna coast (Hawaii), 553

Quercus, 61, 90, 126, 429, 431, 537, 538, =571=

Radiola millegrana, 417, 418

Raiatea, 250, 252, 257

Raillardella, 237

Raillardia, 236-8, 240, 243-4

Ranunculaceæ, 535, 544

Ranunculus;
on tropical mountains, 269, 272;
Hawaiian species, 272-3;
dispersal by birds, 277, 511;
buoyancy of fruits, 535

Ranunculus aquatilis, 37, 535

Ranunculus repens, 37, 86, 535, 545

Ranunculus sceleratus, 37, 535, 544, 545

Ranunculus;
other British species, 535

Raphanus, 33, 478, 536, 540, 596

Rarotonga;
its flora, 48, 177, 291, 295, 309, 317, 320, 323, 336, 374, 419, 425,
551, 578;
altitude, 208;
rainfall, 216;
absence of mangroves, 50;
ferns, 221-2;
endemic species of flowering plants, 232;
the age of arborescent Compositæ and Lobeliaceæ represented by
Fitchia, 237-8;
and by Sclerotheca, 250, 252, 256-7;
scanty representation of Tahitian mountain plants, 293;
a connection with the Kermadec islands, 295

Rauwolfia, 362

Reinecke, Dr., on the Samoan flora, 19, 232, 266, 291, 317, 387, 577,
579;
on Elatostema, 405

Remya, 236-7, 243-4

Rendle, Mr., on Naias, 368

Reseda luteola, 536

“Revue Scientifique,” 506

Rewa River (Fiji);
seed-drift, 76, 78, =435=;
temperature, 78, 564

Reynoldsia, 309, =310=, 371

Rhaphidophora, 404, 532

Rhinanthus crista galli, 537, 545

Rhizophora;
general discussion, =440=, 520-1;
dispersal by currents, 48, 77, 94, =458=;
the “Selala” or seedless form in Fiji, 443;
its representative in Ecuador, 449, 487, 521;
the genus in Ecuador, 483-7;
in Panama, 498-9;
absence from Hawaii, 54-5, and Tahiti, 47;
germination and growth of seedling, =451=, =468=, 575;
river-drift, 435, 460, 499;
beach-drift, 437-8, 460, 499;
polyembryony, 449;
distribution, 54-5, 69, 520

Rhizophora mangle;
in Fiji, 43, 441, 520;
in Ecuador, 484-7;
in Panama, 498-9;
discussed in detail in Chapter XXX

Rhizophora mucronata, 43, 68;
discussed in detail in Chapter XXX

Rhus, 353, 354

Rhynchospora, 272, 283-5

Richella, 265

Ricinus communis, 142, 533, 558

Ridley, Mr. H. N., on the plants of Fernando Noronha and of the Malay
peninsula, 8, 144, 162, 319, 340, 386, 603

River seed-drift;
Thames, 37, 85, 91, 430;
Rewa (Fiji), 76, 91, =435-6=;
Guayaquil or Guayas River (Ecuador), 77, 91, 435, =488-9=;
Panama isthmus, 499;
germination in river-drift, 76-8, 84, 85-6, 435, 488-9

River temperature, 78, 564

Robinson, Mr. E. Kay, on the dispersal of Aster tripolium, 605

Rodriguez Island, 157, 351;
_see_ Mascarene Islands

Rœmeria hybrida, 535

Rollandia, 252, 255, 258

Rosa arvensis, 536

Roxburgh, W., on the seed-structure of Barringtonia, 575

Rubiaceæ;
appearance of the order in the Pacific islands, 261, 262

Rubus, in Hawaii, 269, 272-4, 285, 533, 604

Rumex;
in Hawaii, 366;
modes of dispersal, 367;
fruits in Thames drift, 37, 86, 430;
buoyancy of the fruits of British species, 537, 545

Ruppia maritima, 371, 372, =374=, 398, 482, 537, 555

Russell, Prof., 493

Ruwenzori, Mount, 241, 251

Sachs, Prof., 100

Sadleria, 593

Sage-brush, 279

Sagina procumbens, 536

Sagittaria sagittifolia, 38, 537

Sagus vitiensis, 413

St. Helena, 8, 258

Salicornia;
mode of dispersal, 482, 489, 541, 545

Salicornia herbacea, 34, 35, 537, 541, 545

Salicornia peruviana in Ecuador, 77, 484-6, 489

Salicornia;
other species in Chile and Peru, 478, 482

Saline deposits of North Chile, suggested origin, 485

Salsola kali, 10, 34-6, 429, 431, 477, 479, 537, =541=

Salvia verbenaca, 28, 537, 567-8

Sambucus nigra, 536

Salvinia, 488

Samoa;
altitude, 208-9;
few peculiar genera, 266;
proportion of peculiar species of flowering plants, 232;
littoral flora, 46;
mountain flora, 290, 297, 305;
peculiar species of
Pandanus, 156, 580,
Elatostema, 405,
Eugenia, 349.
Amongst other genera possessing peculiar species are
Gardenia, 311,
Psychotria, 315,
Cyrtandra, 317,
Macaranga, 384,
Ficus, 387

Samolus valerandi, 34, 537, 541

Samolus, in Chile, 478

Sanicula, in Hawaii, 4, 269, 270, 272, =273=, 274-5

San Lorenzo Island (Peru) in the coast-clouds, 492;
decaying shells, 497

Santa Elena Point (Ecuador), its vegetation and climate, 490, 494, 597

Santa Rosa River (Ecuador), the vegetation of its banks, 486

Santalum, 272, 283-4, 285

Sapindus, 309, =325=, 332

Sapota, 373, 532

Sapotaceæ, 372-4

Sararanga, 156

Saxifraga, British species, 536

Saxifragaceæ, in Hawaii, 263

Scævola;
relation between littoral and inland species, 18, 134, =135=;
Hawaiian inland species, 18, 135, 533;
Fijian inland species, 18, 532, 576;
distribution, 67, 71, 564;
modes of dispersal, 135, 564;
buoyancy of fruits, 18, 135, 531-3

Scævola chamissoniana, Hawaii, 18, 533

Scævola gaudichaudii, Hawaii, 18, 533

Scævola floribunda, Fiji, 18, 532, 576

Scævola koenigii;
distribution, 56, 135, 552, 563;
station at coast, 18, 43, 551, 553-6, 560;
extension inland, 42, 59, 121, 547, 553, 560, 569, 579;
modes of dispersal, 57, 71, 552;
buoyancy of fruits, 108, 114, 121, 122, 529, 531, 569;
their occurrence in beach-drift, 437, 559;
in river-drift, 79;
variety, 579;
synonymy, 71, 564

Scævola lobelia, 71

Scævola sericea, 579

Scandinavia;
Sernander’s “Dispersal-biology,” 24;
Atlantic or Gulf-stream drift, 180,189, =430=, 570, 601

Schenck, Dr., 512

Scheuchzeria palustris, 537

Schiedea, 262-3

Schimper, Prof.;
relation between littoral and inland species of a genus and between
the littoral and inland floras, 17, 19, 130-4, 534;
grouping of Malayan littoral plants, 43, 550;
distribution and dispersal of tropical littoral plants, 62, 69;
the littoral Leguminosæ, 201, 581;
structure of the buoyant seeds and fruits of tropical plants and the
question of adaptation, 104, 112-5, 119-29, 569;
on Rhizophora, 441, 446, 453-4, 459;
the essential climatic conditions for mangrove-growth, 470;
the xerophytes and hygrophytes, 32, 40;
epiphytic habit of Vaccinium, 281;
miscellanea, 423, 561, 571, 577, 602;
letters to the author, 43, 121, 440-1, 454;
his indebtedness to Prof. Schimper’s work, 63

Schizæa, 593

Schizostege, 593

Schmidt, J., on Lathyrus maritimus, 543

Schönland, S., on the Hawaiian Lobeliaceæ, 255

Schumann, K., on Ochrosia, 153, 580;
on Musa, 414, 436

“Science Gossip,” 23, 277, 369, 398, 513, 567

Scilla nutans, 537

Scirpodendron costatum, 47, 108, =405=, 435, 530, 551, 552, 563

Scirpus;
buoyancy of the fruits of seven British species, 537;
dispersal by ducks implied in connection with the Cyperaceæ, 513

Scirpus maritimus, 34, 92, 109, 116, 537, 541

Scirpus palustris, 90, 537

Scleria;
dispersal by purple water-hens, 296

Sclerotheca, 252, 256-8

Scott, J., on Indian parrots, 420

Scott-Elliot, Mr., 568

Scrophularia aquatica, 89, 537

Scrophularia nodosa, 537

Scutellaria galericulata, 28, 37, 537, 545

Scyphiphora, 109

Sea, Fijian tree, 436, 602

Sea-aster: _see_ Aster tripolium

Sea-birds, as seed-dispersers, 188, 241, 242, 347, 506, =510-1=, 514,
605

Sea-eagle, as a seed-disperser, 286

Sea-thrifts: _see_ Armeria

Seed-mucosity: _see_ Mucosity

Seed-stage, regarded as an adaptation, 11, 85-7, 468, 473, 521

Seed-structure;
anomalies connected with a lost viviparous habit, 79, 132, 470, 473,
521, 575

Seemann, Dr., on the Fijian flora, 172, 177, 231, 264, 421, 442, 549,
576, 577, 579, 592;
on the northern limit of mangroves, 54;
on Polynesian weeds, 415, 604;
on Laguncularia in the Panama isthmus, 498;
on willow-leaved plants, 603

Selala, the seedless Rhizophora of Fiji, 441-9, 465-6, 487, 520-1;
_see_ Rhizophora

Selliera radicans, 477

Senecio, in Hawaii, 362

Senecio aquaticus, 28, 536, 568

Senecio candidans, 477

Senecio palustris, 536

Senecio vulgaris, 536, 568

Serianthes, 108, 389, =424=, 529, 563

Sernander, Dr., on the dispersal-biology of Scandinavia, 24;
on the Gulf-stream drift, 180, 189, 430, 570, 601;
miscellanea, 280, 511, 538, 542-3, 571

Sesbania, 528, 555

Sesuvium zone of west coast of South America, 476, 481

Sesuvium portulacastrum;
in Fiji, 529;
in Tahiti, 370, 528;
in Tonga, 528;
in Hawaii, 375, 552, 554-5;
in Ecuador, 485;
in Peru, 482;
seed-buoyancy, 16, 529

Shaddock (Citrus decumana), 125-6, 532-3

Shadow of mountains, 586

Sibthorpia europæa, 417, 418

Sicilian beach-drift, 430

Sicyos, 362, 363, =365=

Sideroxylon, 371-4, 507-8

Sida, 417, 427, 533, 555, 604

Siegesbeckia orientalis, 605

Silene, in Hawaii, 272, 279-80

Silene cucubalus, 536

Silene maritima, 33, 36, 280, 511, 536, 540, 544

Simpson, Mr. M.;
precocious germination of the coco-nut, 472

Siphocampylus, 251

Sisyrinchium, 272, 273, 279, 533

Sium;
buoyancy of the fruits of British species, 536

Smith, Mr. Jared G., 211, 213

Smyrnium olusatrum, 536

Smythea pacifica;
distribution, 47, 264-5, 551, 562;
buoyancy of fruits, 106, 112, 529;
their occurrence in river-drift, 435

Snow-buntings (Plectrophanes nivalis), as seed-dispersers, 510-1, 605

Solanum aculeatissimum, 533

Solanum dulcamara, 537

Solanum nigrum (and var. oleraceum), 417, 537, 605

Solereder, H., 265

Solitaire, 8, 159

Solomon Islands, 351, 394, 400

Sonchus asper, 241, 605

Sonchus oleraceus, 536, 593

Sonneratia, 69, 108

Sophora;
general discussion, =147=;
relation between littoral and inland species, 19, 134, =147=, 165,
200;
its indication in New Zealand, 508

Sophora chrysophylla, 20, =147-51=, 271-2, 278, 533, 580

Sophora tetraptera, 64, =148-9=, 271, 431, 478-9, =580=

Sophora tomentosa;
distribution, 19, 54, 68, 147-8, 563;
station and habit, 201-2, 551, 581;
seed-buoyancy, 107, 113, 529, 531, 579;
seeds in beach-drift, 437-8

Soulamea, 265

South America;
observations on the littoral flora of the west coast, 474-501;
_see_ Chile, Peru, Ecuador

Sparganium, 5, 31, 430, 513, 537, 540, 545

Species, their development in the Pacific islands;
the views of Hillebrand, 226-7;
the effect of the greater elevation of the Hawaiian islands on the
endemism (production of new species) of the ferns, 227-8;
tables showing the development of new species and new genera of the
flowering plants of the Pacific islands, 232-4, 252, 255, 263,
265;
the polymorphism of genera, as indicated by their fecundity in
species, illustrated by Cyrtandra, 317;
by Elatostema, 317, 405;
and by Psychotria, 315, 391;
the polymorphism of species, as indicated by their great variability,
520, 522;
_see_ Polymorphous species;
the question whether the relative fecundity of two genera like
Psychotria and Coprosma is to be connected with difference in
antiquity or with difference in their geographical position, 315,
331;
the connection between endemism and the suspension of means of
dispersal by birds, 5, 7, 229;
endemism thus produced nearly as pronounced with certain genera like
Cyrtandra, Embelia, and Naias, in continental regions as in
oceanic islands, 229, 317-3, 331, 363-4, 368, 409, 520;
the process concerned in endemism favoured by the deterioration of
plants in their capacity for dispersal, 262-3, 337, 350, 365, 507,
594-5;
biological connection between plants and birds, 504-5;
differentiation of climate, bird, and plant, the bird being dependent
on the climate, and the plant on the bird, 378, 410, 506-7, 514,
521-2

Spergula arvensis, 536

Spergularia rubra and S. marina, 33, 36, 536, 540, 544-5

Sphacele, 362

Spiræa ulmaria, 37, 92, 536

Spitzbergen;
flora and plant dispersal, 511-2, 514

Spondias dulcis, 532

Spondias lutea, 124, 438, 489, 499

Spondias, unidentified Fijian species, 602

Spring-centres for the diffusion of aquatic plants, 396, 399

Stachys, 27, 28, 537

Stapf, Dr., on the flora of Kinabalu, 101, 296, 301

Statice, 34

Station and seed-buoyancy, 16, 24, 30, 515

Stellaria, 37, 89, 536

Stenogyne, 262-3, 594-5

Sterculia, 11, 375-6, =391=, 532-3

Sterculiaceæ, 10, 375

Straussia, 262-3, 364

Stromboli, its shadow, 586

Strongylodon, 80, =82=, 107, 113, 123, =200=, 436, 529, 531, 558,
562-3, =565=

Struthious birds, suggested as seed-dispersers, 152, 159-60

Stubbs, Dr.;
Agricultural Report on Hawaii, 212, 578

Stylocoryne, 532

Suæda;
concerning dispersal by birds, 511

Suæda fruticosa, 34, 480, 482, 537, 541

Suæda maritima, 34, 537, 541

Suess, Prof., on the shell-fauna of Chile, 496;
on the salts of ancient seas, 566

Sula: _see_ Boobies

Suriana maritima, 105, 528, 529, 563

Swamp-hens: _see_ Porphyrio

Symphytum officinale, 537

Tabernæmontana, 381, =385=, 532

Tacca;
relation between littoral and inland species, 19, 134, =138=

Tacca maculata, 19, 138, 532, 548

Tacca pinnatifida, =19=, =138=;
distribution, 138, 552, 563;
modes of dispersal, 138;
seed-buoyancy, 19, 108, 530, 531, 569;
station, 138;
growing inland, 42, 138, 547, 548, 569;
as a food-plant, 138, 412, 414, 427

Tahiti;
area and altitude, 207-8;
temperature, 209;
rainfall, 216, 218

Tahitian flora;
littoral plants, 14, =47=, 528, =529=, 551, 552;
ferns and lycopods, 220-230, 592;
endemic genera and endemic species of flowering plants, 231-5;
the age of Compositæ, 236-40, 245, 248;
and of Lobeliaceæ, 250-2, 256-8, 266;
the mountain plants, 269, 290-3, 305, 518-9
The age of Malayan plants;
the general dispersal of these plants in the Pacific, 307-58;
the local dispersal,
(_a_) genera common to Tahiti and Hawaii, but not found in
Fiji, 359, 370;
(_b_) genera found in Tahiti and Fiji, but not in Hawaii, 359,
380-8;
(_c_) residual genera (found only in Tahiti), 359, 378.
The absentees from Tahiti, 360, 388
American genera in Tahiti, 379, 380

Talasinga plains: _see_ Fiji

Tamus communis, 537

Tarawau tree of Fiji: _see_ Dracontomelon

Taraxacum, 536

Taro, 412, 415

Taubert, P., on Strongylodon, 566

Taviuni, 207-8

Taxus baccata, 537

Teesdalia, 567

Temperature: _see_ under Beach-temperature, Estuaries, Humboldt current

Tephrosia piscatoria;
in Fiji, 42, 45, 547;
in Hawaii, 56, 58, 59, 552, 553-6;
buoyancy of seed and pod, 529, 562;
modes of dispersal, 45, 150, 417, 562

Terminalia;
relation between coast and inland species, =17=, 120, 533;
buoyancy, 108, 114

Terminalia katappa;
distribution, 52, 54, 551, 552, 563;
buoyancy of fruits, 108, 114, 529;
their occurrence in beach-drift, 437, 559

Terminalia litorea or littoralis, 108, 529, 563

Tertiary submergence of the Western Pacific archipelagoes, 245, 247,
249, 260, 267, =304-5=, 306, 503, 518, 520

Tetramolopium, 236-7, 243-4

Tetraonidæ (Grouse-family), as seed-dispersers, 279, 282, 511, 514

Teucrium inflatum, 416, 605

Teucrium scorodonia, 28, 537

Thacombauia, 265

Thalictrum flavum, 535

Thames;
the vegetation of the banks, 37;
the seed-drift, 37-8, 85-6, 430, 539;
_see_ Additions and Corrections

Thauziès, M. A., 506

Thespesia populnea, 562;
in Hawaii, 52, 552, 554-6, 558-9;
in Fiji, 547, 550;
seed-buoyancy, 105, 529, 530-1;
beach-drift, 558-9

Thlaspi arvense, 536

Thomson, Mr. G. M., on the New Zealand flora, 508

Thomson, Dr. T., on the seed-structure of Barringtonia, 575

Thuret, M., on the buoyancy of seeds and fruits, 24, 63, 538, 542, 544,
546

Thymus, 28, 537, 568

Tibet;
its flora, 34, 238

Tillandsia, 484, 486

Tonga;
included in the Fijian area, 207;
the littoral plants, 46;
the proportion of endemic species, 232;
absence of peculiar genera whether of the Compositæ or Lobeliaceæ or
of any other order, 235, 266;
several peculiar species of Eugenia, 349;
a home for the Sapotaceæ, 374;
Pritchardia pacifica thrives, 327;
vegetation of the leeward plains, 550.
Amongst other genera possessing peculiar species, or in other
respects remarkable, are Podocarpus, 301;
Pittosporum, 309;
Freycinetia, 320;
Elæocarpus, 335;
&c.

Touchardia, 263, 594

Tournefortia argentea;
distribution, 43, 64, 563;
station, 42, 43, 551;
buoyancy of fruits, 108, 114, 530, 531;
their occurrence in beach-drift, 437

Tragopogon, fruit-buoyancy of British species, 536

Tree-Lobelias, 250-60, 266-7;
_see_ Lobeliaceæ

Treub, Dr., on the new flora of Krakatoa, 206, 221, 577

Trevesia, 309, =310=, 371

Tribulus cistoides, 56, 58, 365, 366, 552, 553-6

Trichospermum, 265, =392=, 532

Trifolium, 150, 536

Triglochin, 34, 36, 537, 541, 545

Trimenia, 265

Triplasandra, 262-3

Trisetum, 272, 275

Tristan da Cunha, 242, 276, 286, 364, 366

Triumfetta procumbens, 42, 43, 45, 529, 550

Triumfetta rhomboidea, 529

Tropic-bird (Phaethon), 241

Tropical beach-drift, 434-9;
_see_ Beach-drift

Tschudi, Dr., 493

Tupa, 251, 258

Tussilago farfara, 536, 546

Tussilago petasites, 536

Typha;
seed-buoyancy, 537

Ulex europæus, 536

Ulmus campestris, 537

Umbelliferæ;
station and fruit-buoyancy, 28, 537

Uncinia, 271-5

Urena lobata, 417, 427, 532, 604

Urera, 362

Urtica dioica, 537

Urticaceæ;
represented in the early flora of Hawaii, 261, 263

Vaccinium, 4, 5, 269, 270, 272, 274-5, =280-2=, 290-4, 297, 305, 343

Valerianella, 536

Vallesia, 154

Valparaiso;
beach-plants and beach-drift, 477-9

Vandellia crustacea, 605

Van Tieghem, on Brackenridgea, 570

Vanua Levu;
area, altitude, and rainfall, 207-8, 216

Varigny, Dr. H. de, 506

Vascular cryptogams in the Pacific islands, 222, 592;
_see_ also under Ferns and Lycopods

Veitchia, 401, 532

Verbena officinalis, 537

Veronica;
buoyancy of the fruits of British species, 537

Veronica beccabunga, 37, 537, 568

Vicia;
in the Hawaiian mountain-flora, 272, 278

Vicia faba;
dispersed by pigeons, 150, 417

Vicia sativa;
seed-buoyancy, 536

Victoria Institute, 66

Vigna;
relation between littoral and inland species, 134, 139

Vigna lutea;
in Hawaii, 56-7, 139, 552, 554, 558-9, 560;
in Fiji, 550;
distribution, 68, 563;
seeds dispersed by currents, 57;
seed-buoyancy, 56, 106, 529;
seeds in beach-drift, 437-8, 558-9;
seeds in river-drift, 435, 489 (species in last case unidentified)

Viola;
in Hawaiian mountain-flora, 253, 272;
modes of dispersal, 277;
seed-buoyancy, 533, 536;
seed-mucosity, 278, 567-8

Viscum, 355, 358, 377, 536

Vitex;
relation between littoral and inland species, 134, 137

Vitex trifolia;
station and distribution, 50, 551-2, 563-4;
growing inland, 42, 547-8, 560;
dispersal by currents, 56, 57, 564;
buoyancy of fruits, 108, 122, 530;
their occurrence in river-drift, 435, and in beach-drift, 559;
dispersal by birds, 57, 122, 564;
_see_ Additions and Corrections
Variety, unifoliolata, 108, 547-8, 552, 556, 559, 560

Vitex agnus castus, 109, 431

Viti Levu, 207-8

Vivipary (germination on the plant), 78, 84-7, 132, 191, =468=, =521=,
564, 574-6;
a lost habit with many plants and only indicated by anomalies in
seed-structure, 79, 132, 470, 473, 521, 575

Vries, Prof. H. de, 573

Walker, Mr. F. P., 490, 495, 597

Wallace, Mr. A. R., 7, 242, 247, 259, 400

Wallis Island, 349, 551

Walsh, Canon, on the Cordyline of the Maoris, 420

Waltheria americana, 375, 416, 427, 533, 553-4, 604

Waltheria pyrolæfolia, 375

Warburg, Dr., 54;
on Pandanus, 155-8, 580;
on Freycinetia, 319-22, 510;
on the Samoan species of Ficus, 387

Warming, E., on Rhizophora, 450, 453

Webster, Mr. H., on Ecuador, 495

Wedelia, 108, 116, 529, 551, 563

Weed, Prof., 277, 364

Weeds of Polynesia, 415, 427, 604

Weinmannia, =290-1=, 292-4, 297, 305

Whymper, Mr., on insects at great altitudes, 510, 583

Wichmann, A., on the submergence of the Western Pacific islands, 304

Wiglesworth, Mr., on Polynesian birds, 66, 296

Wikstrœmia, 45, 333, =348=, 357, 530

Wilkes, Commodore;
meteorological observations on the summit of Mauna Loa, 210, 583,
585, =586=

Wilkesia, 236-7, 240, 243-4

Willow-leaved river-side plants, 395, 603

Wilson, Mr. S. B., on the vegetable food of Hawaiian birds, 151, 321

Winds;
in plant-dispersal, 226, 259, 511;
on the summit of Mauna Loa, 583;
_see_ Drying-winds

Wolff, Dr., on Ecuador, 476, 486, 494-5

Wolffia, found by the author in Fiji, 408

Wollastonia, 109

Woodford, Mr. C. M., 66

Xerophytes, 32, 39, 201, 515

Ximenia americana, 107, 113, 115, 122, 128, 529, 563

Yams, 412-4

Zannichellia palustris, 537

Zippelius, on Scirpodendron, 406

RICHARD CLAY AND SONS, LIMITED
BREAD STREET HILL, E.C., AND
BUNGAY, SUFFOLK.

Transcriber’s Note

Some corrections have been made to the original text. In particular,
punctuation has been normalized and ditto marks have been replaced by
the text they represent. Scale bars have been added to the illustrations
where a scale is indicated. Corrections listed in Additions and
Corrections have been made in cases where words or phrases are to be
substituted or removed. The page numbers in the Additions and
Corrections have been corrected and the order has been adjusted
accordingly.

Further corrections are as follows:

p. 48 seseedlings will float uninjured -> seedlings will float uninjured
p. 467 in the case of Bruguiera rheedi -> in the case of Bruguiera
rheedii

*** End of this LibraryBlog Digital Book "Observations of a Naturalist in the Pacific Between 1896 and 1899, v. 2 - Plant-Dispersal" ***

Home