Home
  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
all Classics books content using ISYS

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.
Copyright Status: Not copyrighted in the United States. If you live elsewhere check the laws of your country before downloading this ebook. See comments about copyright issues at end of book.

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

This book is indexed by ISYS Web Indexing system to allow the reader find any word or number within the document.



                           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

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.


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

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.


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

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.


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_)

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.


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

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.


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

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.


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

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.


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

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.


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

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.


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

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.


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

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.


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_)

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.


                               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

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.


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

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.


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.—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.


                         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 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.


              _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 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.


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.

 +-----------+-----+-------------+-------------+--------------------------------------+
 |Genus.     | N s | Distribution| Distribution| Height of  |    Nature of Station.   |
 |           | o p | of genus.   | in the      | plant.     +-------------+-----------+
 |           | . e |             | group.      |            | Elevation.  | Station.  |
 |           |   c |             |             |            +-------------+-----------+
 |           | o i |             |             |            |             |           |
 |           | f e |             |             |            |             |           |
 |           |   s |             |             |            |             |           |
 |           |   . |             |             |            |             |           |
 +-----------+-----+-------------+-------------+------------+-------------+-----------+
 |Brighamia  | 1   | Endemic.    | Molokai,    | 5 to 12    | Islands not | Steep     |
 |           |     |             | Niihau.     | feet.      | exceeding   | palis or  |
 |           |     |             |             |            | 3,500 feet. | mountain  |
 |           |     |             |             |            |             | gaps.     |
 |           |     |             |             |            |             |           |
 |Lobelia    | 5   | Non-endemic.| General.    | 4 to 6     | 2,000 to    | Bridges,  |
 |           |     |             |             | feet.      | 6,000 feet. | gulches   |
 |           |     |             |             |            |             | and woods.|
 |           |     |             |             |            |             |           |
 |Clermontia | 11  | Endemic.    | General.    | Usually 10 | 2,000 to    | Open      |
 |           |     |             |             | to         | 6,000 feet. | woods.    |
 |           |     |             |             | 20 feet.[2]|             |           |
 |           |     |             |             |            |             |           |
 |Rollandia  | 6   | Endemic.    | Oahu.       | Usually 4  | Higher parts| Woods.    |
 |           |     |             |             | to 6 feet, | of Oahu,    |           |
 |           |     |             |             | one species| which is    |           |
 |           |     |             |             | 10 to      | 4,000 feet  |           |
 |           |     |             |             | 15 feet.   | high.       |           |
 |           |     |             |             |            |             |           |
 |Delissea   | 7   | Endemic.    | General.    | 5          | 1,000 to    | Woods and |
 |           |     |             |             | to 10 feet.| 5,000 feet. | gulches.  |
 |           |     |             |             |            |             |           |
 |Cyanea     | 28  | Endemic.    | General.    | Usually    | 1,000 to    | Woods,    |
 |           |     |             |             | 6 to       | 5,000 feet. | ravines,  |
 |           |     |             |             | 15 feet.[3]|             | gulches.  |
 +-----------+-----+-------------+-------------+------------+-------------+-----------+


                  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.

 +--------------+------------------+---------+-----------+-----------+-------------------+
 |    Genus.    |      Order.      |Number of| Character.|   Fruit.  |  Affinities or    |
 |              |                  |species. |           |           | other localities. |
 +--------------+------------------+---------+-----------+-----------+-------------------+
 |Richella      |Anonaceæ.         |    1    |Tree.      |Baccate(?).|Indian in type (C).|
 |              |                  |         |           |           |                   |
 |Trimenia      |Ternstrœmiaceæ.   |    1    |Tree.      |Unknown.   |                   |
 |              |                  |         |           |           |                   |
 |Pimia         |Sterculiaceæ.     |    1    |Tree.      |Small      |Related to         |
 |              |                  |         |           |spinose    | Australian genera |
 |              |                  |         |           |capsule.   | (S).              |
 |              |                  |         |           |           |                   |
 |Græffea       |Tiliaceæ.         |    1    |Tree.      |Unknown.   |Near Trichospermum,|
 |              |                  |         |           |           | a Fijian and      |
 |              |                  |         |           |           | Malayan genus (S).|
 |              |                  |         |           |           |                   |
 |Thacombauia   |Humiriaceæ.       |    1    |Shrub.     |Drupe.     |Order mainly South |
 |              |                  |         |           |           | American.         |
 |              |                  |         |           |           |                   |
 |Amarouria     |Simarubeæ.        |    1    |Tree.      |Dry drupe. |Near Soulamea, a   |
 |              |                  |         |           |           | Malayan genus (S).|
 |              |                  |         |           |           |                   |
 |*Smythea      |Rhamneæ.          |    1    |Straggling |Capsule.   |Also in Burma, New |
 |              |                  |         |shrub.     |           | Guinea, and Malaya|
 |              |                  |         |           |           | (IK), (Sc).       |
 |              |                  |         |           |           |                   |
 |*Oncocarpus   |Anacardiaceæ.     |   2(H)  |Tree.      |Drupe.     |Also in New Guinea |
 |              |                  |         |           |           | (IK).             |
 |              |                  |         |           |           |                   |
 |*Haplopetalon |Rhizophoreæ.      |    2    |Shrub.     |Unknown.   |Also in New        |
 |              |                  |         |           |           | Caledonia (IK).   |
 |              |                  |         |           |           |                   |
 |*Nesopanax    |Plerandreæ.       |    1    |Tree.      |Drupe.     |=Plerandra (IK).   |
 |              |                  |         |           |           |                   |
 |Bakeria       |Plerandreæ.       |    1    |Tree.      |Drupe.     |                   |
 |              |                  |         |           |           |                   |
 |Pelagodendron |Rubiaceæ.         |    1    |Shrub.     |Berry.     |                   |
 |              |                  |         |           |           |                   |
 |*Paphia       |Ericaceæ.         |    1    |Shrub.     |Berry.     |=Agapetes, a       |
 |              |                  |         |           |           | Malayan genus     |
 |              |                  |         |           |           | (IK).             |
 |              |                  |         |           |           |                   |
 |*Carruthersia |Apocyneæ.         |   2(H)  |Climber.   |Berry.     |Also in            |
 |              |                  |         |           |           | Philippines (IK). |
 |              |                  |         |           |           |                   |
 |*Couthovia    |Loganiaceæ.       |    2    |Tree.      |Drupe.     |Also in Kaiser     |
 |              |                  |         |           |           | Wilhelmsland,     |
 |              |                  |         |           |           | New Guinea (So).  |
 |              |                  |         |           |           |                   |
 |Canthiopsis   |Loganiaceæ.       |    1    |Shrub.     |Drupe.     |                   |
 +--------------+------------------+---------+-----------+-----------+-------------------+

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 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.


           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.—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.


    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_)

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.


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 vill