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 | HTML | PDF ]

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: Earth Features and Their Meaning - An Introduction to Geology for the Student and the General Reader
Author: Hobbs, William Herbert
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 "Earth Features and Their Meaning - An Introduction to Geology for the Student and the General Reader" ***

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

Internet Archive (https://archive.org)

      includes the more than 500 original illustrations.
      Images of the original pages are available through
      Internet Archive. See

Transcriber’s note:

      Text enclosed by underscores is in italics (_italics_).

      Text enclosed by equal signs is in bold face (=bold=).

      A carat character is used to denote superscription. A
      single character following the carat is superscripted
      (example: b^1). Multiple superscripted characters are
      enclosed by curly brackets (example: ^{1-2}).

      Subscript numbers in chemical formulas have been
      rendered as _n or _{n}.


      *      *      *      *      *      *

[Illustration: LOGO]

The Macmillan Company
New York · Boston · Chicago
Dallas · San Francisco

Macmillan & Co., Limited
London · Bombay · Calcutta

The Macmillan Co. Of Canada, Ltd.

      *      *      *      *      *      *

¦ PLATE 1.                                                    ¦
¦                                                             ¦
¦ [Illustration: Mount Balfour and the Balfour Glacier in the ¦
¦ Selkirks.]                                                  ¦


An Introduction to Geology for the Student and the General Reader



Professor of Geology in the University of Michigan
Author of “Earthquakes. an Introduction to Seismic Geology”;
“Characteristics of Existing Glaciers”; etc.

New York
The Macmillan Company
All rights reserved

Copyright, 1912,
By The Macmillan Company.

Norwood Press
J. S. Cushing Co.—Berwick & Smith Co.
Norwood, Mass., U.S.A.

              TO THE MEMORY




THE series of readings contained in the present volume give in somewhat
expanded form the substance of a course of illustrated lectures
which has now for several years been delivered each semester at the
University of Michigan. The keynote of the course may be found in
the dominant characteristics of the different earth features and the
geological processes which have been betrayed in the shaping of them.
Such a geological examination of landscape is replete with fascinating
revelations, and it lends to the study of Nature a deep meaning which
cannot but enhance the enjoyment of her varied aspects.

That there is a real place for such a cultural study of geology within
the University is believed to be shown by the increasing number of
students who have elected the work. Even more than in former years the
American travels afar by car or steamship, and the earth’s surface
features in all their manifold diversity are thus one after the other
unrolled before him. The thousands who each year cross the Atlantic
to roam over European countries may by historical, literary, or
artistic studies prepare themselves to derive an exquisite pleasure
as they visit places identified with past achievement of one form or
another. Yet the Channel coast, the gorge of the Rhine, the glaciers of
Switzerland, and the wild scenery of Norway or Scotland have each their
fascinating story to tell of a history far more remote and varied. To
read this history, the runic characters in which it is written must
first of all be mastered; for in every landscape there are strong
individual lines of character such as the pen artist would skillfully
extract for an outline sketch. Such _character profiles_ are often many
times repeated in each landscape, and in them we have a key to the
historical record.

An object of the present readings has thus been to enable the
student to himself pick out in each landscape these more significant
lines and so read directly from Nature. In the landscapes which
have been represented, the aim has been to draw as far as possible
upon localities well known to travelers and likely to be visited,
either because of their historical interest or their purely scenic
attractions. It should thus be possible for a tourist in America or
Europe to pursue his landscape studies whenever he sets out upon his
travels. The better to aid him in this endeavor, some suggestions
concerning the itinerary of journeys have been supplied in an appendix.

Regarded as a textbook of geology, the present work offers some
departures from existing examples. Though it has been customary to
combine in a single text historical with dynamical and structural
geology, a tendency has already become apparent to treat the historical
division apart from the others. Again, a desire to treat the science
of geology comprehensively has led some authors into including so
many subjects as to render their texts unnecessarily encyclopedic
and correspondingly uninteresting to the general reader. It is the
author’s belief that there is a real need for a book which may be read
intelligently by the general public, and it must be recognized that
the beginner in the subject cannot cover the entire field by a single
course of readings. The present work has, therefore, been prepared with
a view to selecting for study those dominant geological processes which
are best illustrated by features in northern North America and Europe.
It is this desire to illustrate the readings by travels afield, which
accounts for the prominence given to the subject of glaciation; for the
larger number of colleges and universities in both America and Europe
are surrounded by the heavy accumulations that have resulted from
former glaciations.

Emphasis has also been placed upon the dependence of the dominant
geological processes of any region upon existing climatic conditions,
a fact to which too little attention has generally been given. This
explains the rather full treatment of desert regions, of which,
in our own country particularly, much may be illustrated upon the
transcontinental railway journeys.

More than in most texts the attempt has here been made to teach
directly through the eye with the efficient aid of apt illustrations
intimately interwoven with the text. For such success as has been
reached in this endeavor, the author is greatly indebted to two
students of the University of Michigan,—Mr. James H. Meier, who has
prepared the line drawings of landscapes, and Mr. Hugh M. Pierce, who
has draughted the diagrams. Though credit has in most cases been given
where illustrations have been made from another’s photographs, yet
especial mention should here be made of the debt to Dr. H. W. Fairbanks
of Berkeley, California, whose beautiful and instructive photographs
are reproduced upon many a page.

As given at the University of Michigan, the lectures reflected in the
present volume are supplemented by excursions and by so much laboratory
practice as is necessary to become familiar with the more common
minerals and rocks, and to read intelligently the usual topographical
and geological maps. In the appendices the means for carrying out such
studies, in part with newly devised apparatus, have been indicated.

The scope of the book precludes the possibility of furnishing the
reader with the sources for the body of fact and theory which is
presented, although much may be inferred from the names which appear
beneath the illustrations, and more definite knowledge will be found in
the references to literature supplied at the ends of chapters. A large
amount of original and unpublished material is for a similar reason
unlabeled, and it has been left for the professional geologist to
detect these new strands which have been drawn into the web.


  October 25, 1911.





  The sources of the history—Subdivisions of geology—The
  study of earth features and their significance—Tabular
  recapitulation—Geological processes not universal—Change, and not
  stability, the order of nature—Observational geology _versus_
  speculative philosophy—The scientific attitude and temper—The
  value of the hypothesis—Heading references                           1



  The lithosphere and its envelopes—The evolution of ideas
  concerning the earth’s figure—The oblateness of the earth—The
  arrangement of oceans and continents—The figure toward
  which the earth is tending—Astronomical _versus_ geodetic
  observations—Changes of figure during contraction of a spherical
  body—The earlier figures of the earth—The continents and
  oceans at the close of the Paleozoic era—The flooded portions
  of the present continents—The floors of the hydrosphere and
  atmosphere—Reading references                                        8



  The rigid quality of our planet—Probable composition of the
  earth’s core—The earth a magnet—The chemical constitution of
  the earth’s surface shell—The essential nature of crystals—The
  lithosphere a complex of interlocking crystals—Some properties of
  natural crystals, minerals—The alterations of minerals—Reading
  references                                                          20



  The processes by which rocks are formed—The marks of origin—The
  metamorphic rocks—Characteristic textures of the igneous
  rocks—The classification of rocks—Subdivisions of the sedimentary
  rocks—The different deposits of ocean, lake, and river—Special
  marks of littoral deposits—The order of deposition during a
  transgression of the sea—The basins of deposition of earlier
  ages—The deposits of the deep sea—Reading references                30



  The zones of fracture and flow—Experiments which illustrate
  the fracture and flow of solid bodies—The arches and troughs
  of the folded strata—The elements of folds—The shapes of rock
  folds—The overthrust fold—Restoration of mutilated folds—The
  geological map and section—Measurement of the thickness of
  formations—The detection of plunging folds—The meaning of an
  unconformity—Reading references                                     40



  The system of the fractures—The space intervals of joints—The
  displacements upon joints: faults—Methods of detecting faults—The
  base of the geological map—The field map and the areal geological
  map—Laboratory models for study of geological maps—The method of
  preparing the map—Fold _vs._ fault topography—Reading references    55



  Nature of earthquake shocks—Seaquakes and seismic sea waves—The
  grander and the lesser earth movements—Changes in the earth’s
  surface during earthquakes: faults and fissures—The measure
  of displacement—Contraction of the earth’s surface during
  earthquakes—The plan of an earthquake fault—The block movements
  of the disturbed district—The earth blocks adjusted during the
  Alaskan earthquake of 1899                                          67


  SEAQUAKES (_concluded_)

  Experimental demonstration of earth movements—Derangement of
  water flow by earth movement—Sand or mud cones and craterlets—The
  earth’s zones of heavy earthquake—The special lines of heavy
  shock—Seismotectonic lines—The heavy shocks above loose
  foundations—Construction in earthquake regions—Reading references   81



  Prevalent misconceptions about volcanoes—Early views concerning
  volcanic mountains—The birth of volcanoes—Active and extinct
  volcanoes—The earth’s volcano belts—Arrangement of volcanic
  vents along fissures, and especially at their intersections—The
  so-called fissure eruptions—The composition and the properties of
  lava—The three main types of volcanic mountain—The lava dome—The
  basaltic lava domes of Hawaii—Lava movements within the caldron
  of Kilauea—The draining of the lava caldrons—The outflow of the
  lava floods                                                         94



  The mechanics of crater explosions—Grander volcanic eruptions
  of cinder cones—The eruption of Volcano in 1888—The eruption of
  Taal volcano on January 30, 1911—The materials and the structure
  of cinder cones—The profile lines of cinder cones—The composite
  cone—The caldera of composite cones—The eruption of Vesuvius
  in 1906—The sequence of events within the chimney—The spine of
  Pelé—The aftermath of mud flows—The dissection of volcanoes—The
  formation of lava reservoirs—Character profiles—Reading
  references                                                         115



  The two contrasted processes of weathering—The rôle of the
  percolating water—Mechanical results of decomposition:
  spheroidal weathering—Exfoliation or scaling—Dome structure
  in granite masses—The prying work of frost—Talus—Soil flow in
  the continued presence of thaw water—The splitting wedges of
  roots and trees—The rock mantle and its shield in the mat of
  vegetation—Reading references                                      149



  The intricate pattern of river etchings—The motive power of
  rivers—Old land and new land—The earlier aspects of rivers—The
  meshes of the river network—The upper and lower reaches
  of a river contrasted—The balance between degradation and
  aggradation—The accordance of tributary valleys—The grading of
  the flood plain—The cycles of stream meanders—The cut-off of the
  meander—Meander scars—River terraces—The delta of the river—The
  levee—The sections of delta deposits                               158



  The newly incised upland and its sharp salients—The stage of
  adolescence—The maturely dissected upland—The Hogarthian line of
  beauty—The final product of river sculpture: the peneplain—The
  river cross sections of successive stages—The entrenchment of
  meanders with renewed uplift—The valley of the rejuvenated
  river—The arrest of stream erosion by the more resistant
  rocks—The capture of one river by another—Water and wind
  gaps—Character profiles—Reading references                         169



  The descent within the unsaturated zone—The trunk channels
  of descending water—The caverns of limestones—Swallow holes
  and limestone sinks—The sinter deposits—The growth of
  stalactites—Formation of stalagmites—The Karst and its features—A
  desert from the destruction of forests—The ponore and the
  polje—The return of the water to the surface—Artesian wells—Hot
  springs and geysers—The deposition of siliceous sinter by plant
  growth—Reading references                                          180



  The law of the desert—The self-registering gauge of past
  climates—Some characteristics of the desert waste—Dry weathering:
  the red and brown desert varnish—The mechanical breakdown of the
  desert rocks—The natural sand blast—The dust carried out of the
  desert                                                             197



  The wandering dunes—The forms of dunes—The cloudburst in the
  desert—The zone of the dwindling river—Erosion in and about the
  desert—Characteristic features of the arid lands—The war of dune
  and oasis—The origin of the high plains which front the Rocky
  Mountains—Character profiles—Reading references                    209



  The weathering processes under control of the fracture system—The
  fracture control of the drainage lines—The repeating pattern in
  drainage networks—The dividing lines of the relief patterns:
  lineaments—The composite repeating patterns of the higher
  orders—Reading references                                          223



  The motion of a water wave—Free waves and breakers—Effect
  of the breaking wave upon a steep, rocky shore: the notched
  cliff—Coves, sea arches, and stacks—The cut rock terrace—The
  cut and built terrace on a steep shore of loose materials—The
  work of the shore current—The sand beach—The shingle beach—Bar,
  spit, and barrier—The land-tied island—A barrier series—Character
  profiles—Reading references                                        231



  The characters in which the record has been preserved—Even
  coast line the mark of uplift—A ragged coast line the mark of
  subsidence—Slow uplift of the coasts; the coastal plain and
  cuesta—The sudden uplifts of the coast—The upraised cliff—The
  uplifted barrier beach—Coast terraces—The sunk or embayed
  coast—Submerged river channels—Records of an oscillation of
  movement—Simultaneous contrary movements upon a coast—The
  contrasted islands of San Clemente and Santa Catalina—The Blue
  Grotto of Capri—Character profiles—Reading references              245



  Conditions essential to glaciation—The snow-line—Importance
  of mountain barriers in initiating glaciers—Sensitiveness of
  glaciers to temperature changes—The cycle of glaciation—The
  advancing hemicycle—Continental and mountain glaciers
  contrasted—The nourishment of glaciers—The upper and lower cloud
  zones of the atmosphere                                            261



  The inland ice of Greenland—The mountain rampart and its
  portals—The marginal rock islands—Rock fragments which travel
  with the ice—The grinding mill beneath the ice—The lifting
  of the grinding tools and their incorporation within the
  ice—Melting upon the glacier margins in Greenland—The marginal
  moraines—The outwash plain or apron—The continental glacier
  of Antarctica—Nourishment of continental glaciers—The glacier
  broom—Field and pack ice—The drift of the pack—The Antarctic
  shelf ice—Icebergs and snowbergs and the manner of their
  birth—Reading references                                           271



  Earlier cycles of glaciation—Contrast of the glaciated and
  nonglaciated regions—The “driftless area”—Characteristics of
  the glaciated regions—The glacier gravings—Younger records over
  older: the glacier palimpsest—The dispersion of the drift—The
  diamonds of the drift—Tabulated comparison of the glaciated and
  nonglaciated regions—Unassorted and assorted drift—Features into
  which the drift is molded—Marginal or “kettle” moraines—Outwash
  plains—Pitted plains and interlobate moraines—Eskers—Drumlins—The
  shelf ice of the ice age—Character profiles                        297



  Interference of glaciers with drainage—Temporary lakes due to ice
  blocking—The “parallel roads” of the Scottish glens—The glacial
  Lake Agassiz—Episodes of the glacial lake history within the St.
  Lawrence Valley—The crescentic lakes of the earlier stages—The
  early Lake Maumee—The later Lake Maumee—Lakes Arkona and
  Whittlesey—Lake Warren—Lakes Iroquois and Algonquin—The Nipissing
  Great Lakes—Summary of lake stages—Permanent changes of drainage
  effected by the glacier—Glacial Lake Ojibway in the Hudson’s Bay
  drainage basin—Reading references                                  320



  The response of the earth’s shell to its ice mantle—The
  abandoned strands as they appear to-day—The records of uplift
  about Mackinac Island—The present inclinations of the uplifted
  strands—The hinge lines of uptilt—Future consequences of the
  continued uptilt within the lake region—Gilbert’s prophecy of a
  future outlet of the Great Lakes to the Mississippi—Geological
  evidences of continued uplift—Drowning of southwestern shores of
  Lakes Superior and Erie—Reading references                         340



  Features in and about the Niagara gorge—The drilling of the
  gorge—The present rate of recession—Future extinction of the
  American Fall—The captured Canadian Fall at Wintergreen Flats—The
  Whirlpool Basin excavated from the St. David’s gorge—The shaping
  of the Lewiston Escarpment—Episodes of Niagara’s history and
  their correlation with those of the glacial lakes—Time measures
  of the Niagara clock—The horologe of late glacial time in
  Scandinavia—Reading references                                     352



  Contrasted sculpturing of continental and mountain glaciers—Wind
  distribution of the snow which falls in mountains—The niches
  which form on snowdrift sites—The augmented snowdrift moves
  down the valley: birth of the glacier—The excavation of the
  glacial amphitheater or cirque—Life history of the cirque—Grooved
  and fretted uplands—The features carved above the glacier—The
  features shaped beneath the glacier—The cascade stairway in
  glacier-carved valleys—The character profiles which result from
  sculpture by mountain glaciers—The sculpture accomplished by ice
  caps—The Norwegian tind or beehive mountain—Reading references     367



  Transition from the ice cap to the mountain glacier—The piedmont
  glacier—The expanded-foot glacier—The dendritic glacier—The
  radiating glacier—The horseshoe glacier—The inherited-basin
  glacier—Summary of types of mountain glacier—Reading references    383



  The glacier flow—Crevasses and séracs—Bodies given up by the
  _Glacier des Bossons_—The moraines—Selective melting upon
  the glacier surface—Glacier drainage—Deposits within the
  vacated valley—Marks of the earlier occupation of mountains by
  glaciers—Reading references                                        390



  Fresh water and saline lakes—Newland lakes—Basin-range
  lakes—Rift-valley lakes—Earthquake lakes—Crater lakes—Coulée
  lakes—Morainal lakes—Pit lakes—Glint or colk lakes—Ice-dam
  lakes—Glacier-lobe lakes—Rock-basin lakes—Valley moraine
  lakes—Landslide lakes—Border lakes—Ox-bow lakes—Saucer
  lakes—Crescentic levee lakes—Raft lakes—Side-delta lakes—Delta
  lakes—Barrier lakes—Dune lakes—Sink lakes—Karst lakes:
  _poljen_—Playa lakes—Salines—Alluvial-dam lakes—Résumé—Reading
  references                                                         401



  Lakes as settling basins—Drawing off of water by erosion of
  outlet—The pulling in of headlands and the cutting off of
  bays—Lake extinction by peat growth—Extinction of lakes in desert
  regions—The rôle of lakes in the economy of nature—Ice ramparts
  on lake shores—Reading references                                  426



  A mountain defined—The festoons of mountain arcs—Theories of
  origin of the mountain arcs—The Atlantic and Pacific coasts
  contrasted—The block type of mountain—Mountains of outflow
  or upheap—Domed mountains of uplift; laccolites—Mountains
  carved from plateaus—The climatic conditions of the mountain
  sculpture—The effect of the resistant stratum—The mark of the
  rift in the eroded mountains—Reading references                    435


  A. The quick determination of the common minerals                  449

  B. Short descriptions of some common rocks                         462

  C. The preparation of topographical maps                           467

  D. Laboratory models for study in the interpretation of
     geological maps                                                 472

  E. Suggested itineraries for pilgrimages to study earth features   475

  INDEX 489



  1. Mount Balfour and the Balfour Glacier in the Selkirks

                                                             FACING PAGE

  2. A. Layers compressed in experiments and showing the effect of
        a competent layer in the process of folding                   44
     B. Experimental production of a series of parallel thrusts
        within closely folded strata                                  44
     C. Apparatus to illustrate shearing action within the
        overturned limb of a fold                                     44

  3. A. An earthquake fault opened in Formosa in 1906 with vertical
        and lateral displacements combined                            72
     B. Earthquake faults opened in Alaska in 1889 on which
        vertical slices of the earth’s shell have undergone
        individual adjustments                                        72

  4. A. Experimental tank to illustrate the earth movements which
        are manifested in earthquakes. The sections of the earth’s
        shell are here represented before adjustment has taken
        place                                                         82
     B. The same apparatus after a sudden adjustment                  82
     C. Model to illustrate a block displacement in rocks which are
        intersected by master joints                                  82

  5. A. Once wooded region in China now reduced to desert through
        deforestation                                                156
     B. “Bad Lands” in the Colorado Desert                           156

  6. A. Barren Karst landscape near the famous Adelsberg grottoes    188
     B. Surface of a limestone ledge where joints have been
        widened through solution                                     188

  7. A. Ranges of dunes upon the margin of the Colorado Desert       210
     B. Sand dunes encroaching upon the oasis of Oued Souf, Algeria  210

  8. A. The granite needles of Harney Peak in the Black Hills of
        South Dakota                                                 216
     B. Castellated erosion chimneys in El Cobra Cañon, New Mexico   216

  9. Map of the High Plains at the eastern front of the Rocky
     Mountains                                                       220

  10. A. View in Spitzbergen to illustrate the disintegration of
      rock under the control of joints                               228

  B. Composite pattern of the joint structures within recent
     alluvial deposits of the Syrian Desert                          228

  11. A. Ripple markings within an ancient sandstone                 232
      B. Wave breaking as it approaches the shore                    232

  12. A. V-shaped cañon cut in an upland recently elevated from the
         sea, San Clemente Island, California                        256
      B. A “hogback” at the base of the Bighorn Mountains, Wyoming   256

  13. A. Precipitous front of the Bryant Glacier outlet of the
         Greenland inland ice                                        272
      B. Lateral stream beside the Benedict Glacier outlet,
         Greenland                                                   272

  14. View of the margin of the Antarctic continental glacier in
      Kaiser Wilhelm Land                                            282

  15. A. An Antarctic ice foot with boat party landing               290
      B. A near view of the front of the Great Ross Barrier,
         Antarctica                                                  290

  16. A. Incised topography within the “driftless area”              300
      B. Built-up topography within the glaciated region             300

  17. A. Soled glacial bowlders which show differently directed
         striæ upon the same facet                                   306
      B. Perched bowlder upon a striated ledge of different rock
         type, Bronx Park, New York                                  306
      C. Characteristic knob and basin surface of a moraine          306

  18. A. Fretted upland of the Alps seen from the summit of Mount
         Blanc                                                       372
      B. Model of the Malaspina Glacier and the fretted upland
         above it                                                    372

  19. A. Contour map of a grooved upland, Bighorn Mountains,
         Wyoming                                                     372
      B. Contour map of a fretted upland, Philipsburg Quadrangle,
         Montana                                                     372

  20. Map of the surface modeled by mountain glaciers in the Sierra
      Nevadas of California                                          376

  21. A. View of the Harvard Glacier, Alaska, showing the
         characteristic terraces                                     394
      B. The terminal moraine at the foot of a mountain glacier      394

  22. A. Model of the vicinity of Chicago, showing the position of
         the outlet of the former Lake Chicago                       400
      B. Map of Yosemite Falls and its earlier site near Eagle Peak  400

  23. A. View of the American Fall at Niagara, showing the
         accumulation of blocks beneath                              414
      B. Crystal Lake, a landslide lake in Colorado                  414

  24. A. Apparatus for exercise in the preparation of topographic
         maps                                                        468
      B. The same apparatus in use for testing the contours of a map 468
      C. Modeling apparatus in use                                   468


  FIG.                                                              PAGE

    1. Diagram to show the measure of the earth’s surface
       irregularities                                                 11

    2. Map to show the reciprocal relation of areas of land and sea   11

    3. The tetrahedral form toward which the earth is tending         12

    4. A truncated tetrahedron to show the reciprocal relation of
       projection and depression upon the surface                     13

    5. Approximations to earlier and present figures of the earth     15

    6. Diagrams for comparison of coasts upon an upright and upon an
       inverted tetrahedron                                           17

    7. The continents, including submerged portions                   18

    8. Diagram to indicate the altitude of different parts of the
       lithosphere surface                                            18

    9. Diagram to show how the terrestrial rocks grade into the
       meteorites                                                     22

   10. Comparison of a crystalline with an amorphous substance        24

   11. “Light figure” seen upon etched surface of calcite             25

   12. Battered sand grains which have developed crystal faces        26

   13. Unassimilated grains of quartz within a garnet crystal         28

   14. New minerals developed about the core of an augite crystal     28

   15. A common rim of new mineral developed by reaction where
       earlier minerals come into contact                             28

   16. Laminated structure of a sedimentary rock                      30

   17. Characteristic textures of igneous rocks                       33

   18. Diagram to show the order of sediments laid down during a
       transgression of the sea                                       37

   19. Fractures produced by compression of a block of molder’s wax   41

   20. Apparatus to illustrate the folding of strata                  41

   21. Diagrams of fold types                                         42

   22. Diagrams to illustrate crustal shortening                      42

   23. Anticlinal and synclinal folds                                 43

   24. Diagrams to illustrate the shapes of rock folds                44

   25. Secondary and tertiary flexures superimposed upon the
   primary ones                                                       44

   26. A bent stratum to illustrate tension and compression upon
   opposite sides                                                     45

   27. A geological section with truncated arches restored            47

   28. Diagram to illustrate the nature of strike and dip             47

   29. Diagram to show the use of T symbols for strike and dip
       observation                                                    48

   30. Diagram to show how the thickness of a formation is
       determined                                                     49

   31. A plunging anticline                                           50

   32. A plunging syncline                                            50

   33. An unconformity upon the coast of California                   51

   34. Series of diagrams to illustrate the episodes involved in
       the production of an angular unconformity                      52

   35. Types of deceptive or erosional unconformities                 53

   36. A set of master joints in shale                                55

   37. Diagram to show the manner of replacement of one set of
       joints by another                                              56

   38. Diagram to show the different combinations of joint series     56

   39. View of the shore in West Greenland                            57

   40. View in Iceland which shows joint intervals of more than one
       order                                                          57

   41. Faulted blocks of basalt near Woodbury, Connecticut            58

   42. A fault in previously disturbed strata                         59

   43. Diagram to show the effect of erosion upon a fault             60

   44. A fault plane exhibiting drag                                  60

   45. Map to show how a fault may be indicated by abrupt changes
       in strike and dip                                              61

   46. A series of parallel faults revealed by offsets                61

   47. Field map prepared from the laboratory table                   64

   48. Areal geological map based upon the field map                  64

   49. A portion of the ruins of Messina                              67

   50. Ruins of the Carnegie Palace of Peace at Cartaga, Costa Rica   68

   51. Overturned bowlders from Assam earthquake of 1897              69

   52. Post sunk into ground during Charleston earthquake             69

   53. Map showing localities where shocks have been reported at
       sea off Cape Mendocino, California                             70

   54. Effect of seismic water wave in Japan                          70

   55. A fault of vertical displacement                               71

   56. Escarpment produced by an earthquake fault in India            72

   57. A fault of lateral displacement                                72

   58. Fence parted and displaced by lateral displacement on fault
       during California earthquake                                   72

   59. Fault with vertical and lateral displacements combined         72

   60. Diagram to show how small faults may be masked at the
       earth’s surface                                                73

   61. “Mole hill” effect above buried earthquake fault               73

   62. Post-glacial earthquake faults                                 74

   63. Earthquake cracks in Colorado desert                           74

   64. Railway tracks broken or buckled at time of earthquake         75

   65. Railroad bridge in Japan damaged by earthquake                 75

   66. Diagrams to show contraction of earth’s crust during an
       earthquake                                                     76

   67. Map of the Chedrang fault of India                             76

   68. Displacements along earthquake fault in Alaska                 77

   69. Abrupt change in direction of throw upon an earthquake fault   77

   70. Map of faults in the Owens Valley, California, formed during
       earthquake of 1872                                             78

   71. Marquetry of the rock floor in the Tonopah district, Nevada    79

   72. Map of Alaskan coast to show adjustments of level during an
       earthquake                                                     79

   73. An Alaskan shore elevated seventeen feet during the
       earthquake of 1899                                             80

   74. Partially submerged forest from depression of shore in
       Alaska during earthquake                                       80

   75. Effect of settlement of the shore at Port Royal during
       earthquake of 1907                                             80

   76. Diagrams to illustrate the draining of lakes during
       earthquakes                                                    83

   77. Diagram to illustrate the derangements of water flow during
       an earthquake                                                  84

   78. Mud cones aligned upon an earthquake fissure in Servia         84

   79. Craterlet formed near Charleston, South Carolina, during the
       earthquake of 1886                                             85

   80. Cross section of a craterlet                                   85

   81. Map of the island of Ischia to show the concentration of
       earthquake shocks                                              87

   82. A line of earth fracture revealed in the plan of the relief    87

   83. Seismotectonic lines of the West Indies                        88

   84. Device to illustrate the different effects of earthquakes in
       firm rock and in loose materials                               88

   85. House wrecked in San Francisco earthquake                      90

   86. Building wrecked in California earthquake by roof and upper
       floor battering down the upper walls                           91

   87. Breached volcanic cone in New Zealand showing the bending
       down of the strata near the vent                               96

   88. View of the new Camiguin volcano formed in 1871 in the
       Philippines                                                    97

   89. Map to show the belts of active volcanoes                      98

   90. A portion of the “fire girdle” of the Pacific                  98

   91. Volcanic cones formed in 1783 above the Skaptár fissure in
       Iceland                                                        99

   92. Diagrams to illustrate the location of volcanic vents upon
       fissure lines                                                 100

   93. Outline map showing the arrangement of volcanic vents upon
       the island of Java                                            100

   94. Map showing the migration of volcanoes along a fissure        101

   95. Basaltic plateau of the northwestern United States due to
       fissure eruptions of lava                                     102

   96. Lava plains about the Snake River in Idaho                    102

   97. Characteristic profiles of lava volcanoes                     103

   98. A driblet cone                                                104

   99. Leffingwell Crater, a cinder cone in the Owens Valley,
       California                                                    104

  100. Map of Hawaii and its lava volcanoes                          106

  101. Section through Mauna Loa and Kilauea                         106

  102. Schematic diagram to illustrate the moving platform in the
       crater of Kilauea                                             107

  103. View of the open lava lake of Halemaumau                      108

  104. Map to show the manner of outflow of the lava from Kilauea
       in the eruption of 1840                                       109

  105. Lava of Matavanu flowing down to the sea during the eruption
       of 1906                                                       110

  106. Lava stream discharging into the sea from a lava tunnel       111

  107. Diagrammatic representation of the structure of lava
       volcanoes as a result of the draining of frozen lava streams  112

  108. Diagram to show the formation of mesas by outflow of lava in
       valleys and subsequent erosion                                112

  109. Surface of lava of the Pahoehoe type                          113

  110. Three successive views to show the growth of the island of
       Savaii, from lava outflow in 1906                             113

  111. View of the volcano of Stromboli showing the excentric
       position of the crater                                        116

  112. Diagrams to illustrate the eruptions within the crater of
       Stromboli                                                     117

  113. Map of Volcano in the Æolian Islands                          118

  114. “Bread-crust” lava projectile from the eruption of Volcano
       in 1888                                                       119

  115. “Cauliflower cloud” of steam and ash rising above the cinder
  cone of Volcano 120

  116. Eruption of Taal volcano in 1911 seen from a distance of six
       miles                                                         120

  117. The thick mud veneer upon the island of Taal (after a
       photograph by Deniston)                                       121

  118. A pear-shaped lava projectile                                 121

  119. Artificial production of a cinder cone                        122

  120. Diagram to show the contrast between a lava dome and a
       cinder cone                                                   123

  121. Mayon volcano on the island of Luzon, Philippine Islands      123

  122. A series of breached cinder cones due to migration of the
       eruption along a fissure                                      124

  123. The mouth upon the inner cone of Mount Vesuvius from which
       flowed the lava of 1872                                       124

  124. A row of parasitic cones raised above a fissure opened on
       the flanks of Etna in 1892                                    125

  125. View of Etna, showing the parasitic cones upon its flanks     125

  126. Sketch map of Etna to show the areas covered by lava and
       tuff respectively                                             126

  127. Panum crater showing the caldera                              126

  128. View of Mount Vesuvius before the eruption of 1906            127

  129. Sketches of the summit of the Vesuvian cone to bring out the
       changes in its outline                                        128

  130. Night view of Vesuvius from Naples before the outbreak of
       1906, showing a small lava stream descending the central cone 129

  131. Scoriaceous lava encroaching upon the tracks of the Vesuvian
       railway                                                       130

  132. Map of Vesuvius, showing the position of the lava mouths
       opened upon its flanks during the eruption of 1906            131

  133. The ash curtain over Vesuvius lifting and disclosing the
       outlines of the mountain                                      132

  134. The central cone of Vesuvius as it appeared after the
       eruption of 1906                                              132

  135. A sunken road upon Vesuvius filled with indrifted ash         133

  136. View of Vesuvius from the southwest during the waning stages
       of the eruption                                               133

  137. The main lava stream advancing upon Boscotrecase              133

  138. A pine snapped off by the lava and carried forward upon its
       surface                                                       133

  139. Lava front pushing over and running around a wall in its
       path                                                          134

  140. One of the ruined villas in Boscotrecase                      134

  141. Three diagrams to illustrate the sequence of events during
       the cone-building and crater-producing periods                135

  142. The spine of Pelé rising above the chimney of the volcano
       after the eruption of 1902                                    136

  143. Successive outlines of the Pelé spine                         137

  144. Corrugated surface of the Vesuvian cone due to the mud flows
       which followed the eruption of 1906                           138

  145. View of the Kammerbühl near Eger in Bohemia                   139

  146. Volcanic plug exposed by natural dissection of a volcanic
       cone in Colorado                                              140

  147. A dike cutting beds of tuff in a partly dissected volcano of
       southwestern Colorado                                         140

  148. Map and general view of St. Paul’s rocks, a volcanic cone
       dissected by waves                                            141

  149. Dissection by explosion of Little Bandai-san in 1888          141

  150. The half-submerged volcano of Krakatoa before and after the
       eruption of 1883                                              142

  151. The cicatrice of the Banat                                    142

  152. Diagram to illustrate a probable cause of formation of lava
       reservoirs and the connection with volcanoes upon the surface 143

  153. Effect of relief of load upon rocks by arching of a
       competent formation                                           144

  154. Character profiles connected with volcanoes                   146

  155. Diagrams to show the effect of decomposition in producing
       spheroidal bowlders                                           150

  156. Spheroidal weathering of an igneous rock                      151

  157. Dome structure in granite mass                                152

  158. Talus slope beneath a cliff                                   153

  159. Striped ground from soil flow                                 154

  160. Pavement of horizontal surface due to soil flow               154

  161. Tree roots prying rock apart on fissure                       154

  162. Bowlder split by a growing tree                               155

  163. Rock mantle beneath soil and vegetable mat                    155

  164. Diagram to show the varying thickness of mantle rock upon
       the different portions of a hill surface                      156

  165. Gullies from earliest stage of a river’s life                 160

  166. Partially dissected upland                                    160

  167. Longitudinal sections of upper portion of a river valley      161

  168. Map and sections of a stream meander                          163

  169. Tree undermined on the outer bank of a meander                164

  170. Diagrams to show the successive positions of stream meanders  164

  171. An ox-bow lake in the flood plain of a river                  165

  172. Schematic representation of a series of river terraces        165

  173. “Bird-foot” delta of the Mississippi River                    167

  174. Diagrams to show the nature of delta deposits as exhibited
       in sections                                                   168

  175. Gorge of the River Rhine near St. Goars                       169

  176. Valley with rounded shoulders characteristic of the stage of
       adolescence                                                   170

  177. View of a maturely dissected upland                           170

  178. Hogarth’s line of beauty                                      171

  179. View of the oldland of New England, with Mount Monadnock
       rising in the distance                                        171

  180. Comparison of the cross sections of river valleys of
       different stages                                              172

  181. The Beavertail Bend of the Yakima River 173

  182. A rejuvenated river valley 174

  183. Plan of a river narrows 174

  184. Successive diagrams to illustrate the origin of “trellis
       drainage”                                                     175

  185. Sketch maps to show the earlier and present drainage near
       Harper’s Ferry                                                176

  186. Section to illustrate the history of Snickers Gap             177

  187. Character profiles of landscapes shaped by stream erosion in
       humid climates                                                177

  188. Diagram to show the seasonal range in the position of the
       water table                                                   180

  189. Diagram to show the effect of an impervious layer upon the
       descending water                                              181

  190. Sketch map to illustrate corrosion of limestone along two
       series of vertical joints                                     181

  191. Diagram to show the relation of limestone caverns to the
       river system of the district                                  182

  192. Plan of a portion of Mammoth Cave, Kentucky                   183

  193. Trees and shrubs growing upon the bottoms of limestone sinks  183

  194. Diagrams to show the manner of formation of stalactites and
       stalagmites                                                   185

  195. Sinter formations in the Luray caverns                        186

  196. Map of the dolines of the Karst region                        187

  197. Cross section of a doline formed by inbreak                   187

  198. Sharp Karren of the Ifenplatte                                188

  199. The Zirknitz seasonal lake                                    189

  200. Fissure springs arranged at intersections of rock fractures   190

  201. Schematic diagrams to illustrate the different types of
       artesian wells                                                191

  202. Cross section of Geysir, Iceland                              192

  203. Apparatus for simulating geyser action                        193

  204. Cone of siliceous sinter about the Lone Star Geyser           194

  205. Former shore lines in the Great Basin                         198

  206. Map of the former Lake Bonneville                             199

  207. Borax deposits in Death Valley, California                    201

  208. Hollowed forms of weathered granite in a desert of Central
       Asia                                                          201

  209. Hollow hewn blocks in a wall in the Wadi Guerraui             202

  210. Smooth granite domes shaped by exfoliation                    203

  211. Granite blocks rent by diffission                             204

  212. “Mushroom Rock” from a desert in Wyoming                      205

  213. Windkanten shaped by sand blast in the desert                 205

  214. The “stone lattice” of the desert                             206

  215. Shadow erosion in the desert                                  206

  216. Cliffs in loess with characteristic vertical jointing         207

  217. A cañon in loess worn by traffic and wind                     207

  218. Diagrams to illustrate the effects of obstructions in
       arresting wind-driven sand                                    209

  219. Sand accumulating on either side of a firm and impenetrable
       obstruction                                                   210

  220. Successive diagrams to illustrate the history of the town of
       Kunzen upon the Kurische Nehrung                              210

  221. View of desert barchans                                       211

  222. Diagrams to show the relationships of dunes to sand supply
       and wind direction                                            211

  223. Ideal section showing the rising mountain wall about a
       desert and the neighboring slope                              212

  224. Dry delta at the foot of a range upon the borders of a
       desert                                                        213

  225. Map of distributaries of streams which issue at the western
       base of the Sierra Nevadas                                    213

  226. A group of “demoiselles” in the “bad lands”                   214

  227. Amphitheater at the head of the Wadi Beni Sur                 215

  228. Mesa and outlier in the Leucite Hills of Wyoming              216

  229. Flat-bottomed basin separating dunes                          216

  230. Billowy surface of the salt crust on the central sink of the
       desert of Lop                                                 217

  231. Schematic diagram to show the zones of deposition in their
       order from the margin to the center of a desert               217

  232. Mounds upon the site of the buried city of Nippur             218

  233. Exhumed structures in the buried city of Nippur               218

  234. Section across the High Plains                                219

  235. Section across the lenticular threads of alluvial deposits
       of the High Plains                                            220

  236. Distributaries of the foot hills superimposed upon an
       earlier series                                                220

  237. Character profiles in the landscapes of arid lands            220

  238. Rain sculpturing under control by joints                      224

  239. Sagging of limestone above joints                             224

  240. Map of the joint-controlled Abisko Cañon in Northern Lapland  225

  241. Map of the gorge of the Zambesi River below Victoria Falls    225

  242. Controlled drainage network of the Shepaug River in
       Connecticut                                                   226

  243. A river network of repeating rectangular pattern              226

  244. Squared mountain masses which reveal a distribution of
       joints in block patterns of different orders                  228

  245. Island groups of the Lofoten Archipelago 229

  246. Diagrams to illustrate the composite profiles of the islands
       on the Norwegian coast                                        229

  247. Diagram to show the nature of the motions within a free
       water wave                                                    231

  248. Diagram to illustrate the transformation of a free wave into
       a breaker                                                     232

  249. Notched rock cliff and fallen blocks                          233

  250. A wave-cut chasm under control by joints                      233

  251. Grand Arch upon one of the Apostle Islands in Lake Superior   234

  252. Stack near the shore of Lake Superior                         234

  253. The Marble Islands, stacks in a lake of the southern Andes    235

  254. Squared stacks revealing the position of the joint planes on
       which they were carved                                        235

  255. Ideal section cut by waves upon a steep rocky shore           236

  256. Map showing the outlines of the island of Heligoland at
       different stages in its history                               236

  257. Ideal section carved by waves upon a steep shore of loose
       materials                                                     237

  258. Sloping cliff and boulder pavement at Scituate,
       Massachusetts                                                 237

  259. Map to show the nature of the shore current and the forms
       which are molded by it                                        238

  260. Crescent-shaped beach in the lee of a headland                239

  261. Cross section of a beach pebble                               239

  262. A storm beach on the northeast shore of Green Bay             240

  263. Spit of shingle on Au Train Island, Lake Superior             240

  264. Barrier beach in front of a lagoon                            241

  265. Cross section of a barrier beach with lagoon in its rear      242

  266. Cross section of a series of barriers and an outer bar        242

  267. A barrier series and an outer bar on Lake Mendota at
       Madison, Wisconsin                                            242

  268. Series of barriers at the western end of Lake Superior        243

  269. Character profiles resulting from wave action upon shores     243

  270. The even shore line of a raised coast                         246

  271. The ragged coast line produced by subsidence                  246

  272. Portion of the Atlantic coastal plain at the base of the
       oldland                                                       246

  273. Ideal form of cuestas and intermediate lowlands carved from
       a coastal plain                                               247

  274. Uplifted sea cave on the coast of California                  248

  275. Double-notched cliff near Cape Tiro, Celebes                  248

  276. Uplifted stacks on the coast of California                    249

  277. Uplifted shingle beach across the entrance to a former bay
       upon the coast of California                                  250

  278. Raised beach terraces near Elie, Fife, Scotland               250

  279. Uplifted sea cliffs and terraces on the Alaskan coast         250

  280. Diagrams to show how excessive sinking upon the sea floor
       will cause the shore to migrate landward                      251

  281. A drowned river mouth or estuary upon a coastal plain         251

  282. Archipelago of steep rocky islets due to submergence          252

  283. The submerged Hudsonian channel which continues the Hudson
       River across the continental shelf                            252

  284. Marine clay deposits near the mouths of the Maine rivers
       which preserve a record of earlier subsidence and later
       elevation                                                     253

  285. View of the three standing columns of the Temple of Jupiter
       Serapis, at Pozzuoli                                          254

  286. Three successive views to set forth the recent oscillations
       of level on the northern shore of the Bay of Naples           255

  287. Relief map of San Clemente Island, California                 256

  288. Relief map of Santa Catalina Island, California               257

  289. Cross section of the Blue Grotto, on the island of Capri      258

  290. Character profiles of coast elevation and subsidence          259

  291. Map showing the distribution of existing glaciers and the
       two important wind poles of the earth                         263

  292. An Alaskan glacier spreading out at the foot of the range
       which nourishes it                                            264

  293. Surface of a glacier whose upper layers spread with but
       slight restraint from retaining walls                         265

  294. Section through a mountain glacier                            267

  295. Profile across the largest of the Icelandic ice caps          267

  296. Ideal section across a continental glacier                    267

  297. View of the Eyriks Jökull, an ice cap of Iceland              268

  298. The zones of the lower atmosphere as revealed by recent kite
       and balloon exploration                                       269

  299. Map of Greenland, showing the area of inland ice and the
       routes of explorers                                           271

  300. Profile in natural proportions across the southern end of
       the continental glacier of Greenland                          272

  301. Map of a glacier tongue with dimple above                     273

  302. Edge of the Greenland inland ice, showing the nunataks
       diminishing in size toward the interior                       274

  303. Moat surrounding a nunatak in Victoria Land                   274

  304. A glacier pavement of Permo-Carboniferous age in South
       Africa                                                        276

  305. Diagrams to illustrate the manner of formation of scape
       colks                                                         277

  306. Marginal moraine now forming at the edge of the continental
       glacier of Greenland                                          279

  307. Small lake between the ice front and a moraine which it has
       recently built                                                279

  308. View of a drained lake bottom between the ice front and an
       abandoned moraine                                             280

  309. Diagrams to show the manner of formation and the structure
       of an outwash plain and fosse                                 280

  310. Map of the ice masses of Victoria Land, Antarctica            282

  311. Sections across the inland ice and the shelf ice of
       Antarctica                                                    283

  312. Diagram to show the nature of the fixed glacial anticyclone
       above continental glaciers                                    284

  313. Snow deltas about the margins of a glacier tongue in
       Greenland                                                     285

  314. View of the sea ice of the Arctic region                      286

  315. Map of the north polar regions, showing the area of drift
       ice and the tracks of the _Jeannette_ and the _Fram_          288

  316. The shelf ice of Coats Land with surrounding pack ice         290

  317. Tidewater cliff on a glacier tongue from which icebergs are
       born                                                          290

  318. A Greenlandic iceberg after a long journey in warm latitudes  291

  319. Diagram showing one way in which northern icebergs are born
       from the glacier tongue                                       291

  320. A northern iceberg surrounded by sea ice                      292

  321. Tabular Antarctic iceberg separating from the shelf ice       293

  322. Map of the globe, showing the areas covered by continental
       glaciers during the “ice age”                                 297

  323. Glaciated granite bowlder weathered out of a moraine of
       Permo-Carboniferous age, South Australia                      298

  324. Map to show the glaciated and nonglaciated regions of North
       America                                                       298

  325. Map of the glaciated and nonglaciated areas of northern
       Europe                                                        299

  326. An unstable erosion remnant characteristic of the “driftless
       area”                                                         300

  327. Diagram showing the manner in which a continental glacier
       obliterates existing valleys                                  301

  328. Lake and marsh district in northern Wisconsin                 302

  329. Cross section in natural proportion of the latest North
       American continental glacier                                  303

  330. Diagram showing the earlier and the later glacier records
       together upon the same limestone surface                      304

  331. Map to show the outcroppings of peculiar rock types in the
       region of the Great Lakes, and some localities where “drift
       copper” has been collected                                    305

  332. Map of the “bowlder train” from Iron Hill, Rhode Island 306

  333. Shapes and approximate natural sizes of some of the diamonds
       from the Great Lakes region                                   307

  334. Glacial map of a portion of the Great Lakes region            308

  335. Section in coarse till                                        310

  336. Sketch map of portions of Michigan, Ohio, and Indiana,
       showing the distribution of moraines                          312

  337. Map of the vicinity of Devil’s Lake, Wisconsin, partly
       covered by the continental glacier                            313

  338. Moraine with outwash apron in front                           313

  339. Fosse between an outwash plain and a moraine                  314

  340. View along an esker in southern Maine                         315

  341. Outline map of moraines and eskers in Finland                 315

  342. Sketch maps showing the relationships of drumlins and eskers  316

  343. View of a drumlin, showing an opening in the till             317

  344. Outline map of the front of the Green Bay lobe to show the
       relationships of drumlins, moraines, outwash plains, and
       ground moraine                                                317

  345. Character profiles referable to continental glacier           318

  346. View of the flood plain of the ancient Illinois River near
       Peoria                                                        320

  347. Broadly terraced valleys which mark the floods that once
       issued from the continental glacier of North America          321

  348. Border drainage about the retreating ice front south of Lake
       Erie                                                          321

  349. The “parallel roads” of Glen Roy in the Scottish Highlands    322

  350. Map of Glen Roy and neighboring valleys of the Scottish
       Highlands                                                     322

  351. Three successive diagrams to set forth the late glacial lake
       history of the Scottish glens                                 324

  352. Harvesting time on the fertile floor of the glacial Lake
       Agassiz                                                       325

  353. Map of Lake Agassiz                                           325

  354. Map showing some of the beaches of Lake Agassiz and its
       outlet                                                        326

  355. Narrows of the Warren River where it passed between jaws of
       granite and gneiss                                            327

  356. Map of the valley of the Warren River near Minneapolis        327

  357. Portion of the Herman beach on the shore of the former Lake
       Agassiz                                                       328

  358. Map of the continental glacier of North America when it
       covered the entire St. Lawrence basin                         329

  359. Outline map of the early Lake Maumee                          330

  360. Map to show the first stages of the ice-dammed lakes within
       the St. Lawrence basin                                        330

  361. Outline map of the later Lake Maumee and its outlet           332

  362. Outline map of lakes Whittlesey and Saginaw                   333

  363. Map of the glacial Lake Warren                                333

  364. Map of the glacial Lake Algonquin                             334

  365. Outline map of the Nipissing Great Lakes                      335

  366. Probable preglacial drainage of the upper Ohio region         337

  367. Diagrams to illustrate the episodes in the recent history of
       a Connecticut river                                           338

  368. The notched rock headland of Boyer Bluff on Lake Michigan     341

  369. View of Mackinac Island from the direction of St. Ignace      342

  370. The “Sugar Loaf”, a stack of Lake Algonquin upon Mackinac
       Island                                                        342

  371. Beach ridges in series on Mackinac Island                     343

  372. Notched stack of the Nipissing Great Lakes at St. Ignace      343

  373. Series of diagrams to illustrate the evolution of ideas
       concerning the uplift of the lake region since the Ice Age    344

  374. Map of the Great Lakes region to show the isobases and hinge
       lines of uptilt                                               345

  375. Series of diagrams to indicate the nature of the recovery of
       the crust by uplift when unloaded of an ice mantle            346

  376. Portion of the Inner Sandusky Bay, for comparison of the
       shore line of 1820 with that of to-day                        350

  377. Ideal cross section of the Niagara Gorge to show the
       marginal terrace                                              353

  378. View of the bed of the Niagara River above the cataract
       where water has been drained off                              353

  379. View of the Falls of St. Anthony in 1851                      354

  380. Ideal section to show the nature of the drilling process
       beneath the cataract                                          355

  381. Plan and section of the gorge, showing how the depth is
       proportional to the width                                     355

  382. Comparative views of the Canadian Falls in 1827 and 1895      356

  383. Map to show the recession of the Canadian Fall                357

  384. Comparison of the present with the future falls               358

  385. Bird’s-eye view of the captured Canadian Fall at Wintergreen
       Flats                                                         358

  386. Map of the Whirlpool Basin                                    360

  387. Map of the cuestas which have played so important a part in
       fixing the boundaries of the lake basins                      361

  388. Bird’s-eye view of the cuestas south of Lakes Ontario and
       Erie                                                          362

  389. Sketch map of the greater portion of the Niagara Gorge to
       illustrate Niagara history                                    363

  390. Snowdrift hollowing its bed by nivation                       368

  391. Amphitheater formed upon a drift site in northern Lapland     369

  392. The marginal crevasse on the highest margin of a glacier      370

  393. Niches and cirques in the Bighorn Mountains of Wyoming        371

  394. Subordinate cirques in the amphitheater on the west face of
       the Wannehorn                                                 371

  395. “Biscuit cutting” effect of glacial sculpture in the Uinta
       Mountains of Wyoming                                          372

  396. Diagram to show the cause of the hyperbolic curve of cols     372

  397. A col in the Selkirks                                         373

  398. Diagrams to illustrate the formation of comb ridges, cols,
       and horns                                                     374

  399. The U-shaped Kern Valley in the Sierra Nevadas of California  375

  400. Glaciated valley wall, showing the sharp line which
       separates the abraded from the undermined rock surface        375

  401. View of the Vale of Chamonix from the séracs of the _Glacier
       des Bossons_                                                  376

  402. Map of an area near the continental divide in Colorado        377

  403. Gorge of the Albula River in the Engadine cut through a rock
       bar                                                           378

  404. Idealistic sketch, showing glaciated and nonglaciated side
       valleys                                                       378

  405. Character profiles sculptured by mountain glaciers            379

  406. Flat dome shaped under the margin of a Norwegian ice cap      379

  407. Two views which illustrate successive stages in the shaping
       of tinds                                                      380

  408. Schematic diagram to bring out the relationships of the
       various types of mountain glaciers                            383

  409. Map of the Malaspina Glacier of Alaska                        384

  410. Map of the Baltoro Glacier of the Himalayas                   385

  411. View of the Triest Glacier, a hanging glacieret               385

  412. Map of the Harriman Fjord Glacier of Alaska                   386

  413. Map of the Rotmoos Glacier, a radiating glacier of
       Switzerland                                                   386

  414. Outline map of the Asulkan Glacier in the Selkirks, a
       horseshoe glacier                                             387

  415. Outline map of the Illecillewaet Glacier of the Selkirks, an
       inherited-basin glacier                                       388

  416. Diagram to illustrate the surface flow of glaciers            390

  417. Diagram to show the transformation of crevasses into séracs   391

  418. View of the _Glacier des Bossons_, showing the position of
       accidents to Alpinists                                        392

  419. Lines of flow upon the surface of the _Hintereisferner_
       Glacier in the Alps                                           393

  420. Lateral and medial moraines of the _Mer de Glace_ and its
       tributaries                                                   393

  421. Ideal cross section of a mountain glacier                     394

  422. Diagrams to illustrate the melting effects upon glacier ice
       of rock fragments of different sizes                          394

  423. Small glacier table upon the Great Aletsch Glacier            395

  424. Effects of differential melting and subsequent refreezing
       upon a glacier surface                                        396

  425. Dirt cone with its casing in part removed                     396

  426. Schematic diagram to show the manner of formation of glacier
       cornices                                                      397

  427. Superglacial stream upon the Great Aletsch Glacier            398

  428. Ideal form of the surface left on the site of a piedmont
       glacier apron                                                 399

  429. Map of the site of the earlier piedmont glacier of the Upper
       Rhine                                                         399

  430. Diagram and map to bring out the characteristics of newland
       lakes                                                         402

  431. View of the Warner Lakes, Oregon                              402

  432. Schematic diagram to illustrate the characteristics of
       basin-range lakes                                             403

  433. Schematic diagram of rift-valley lakes and the valley of the
       Jordan                                                        403

  434. Map of the rift-valley lakes of East Central Africa           404

  435. Earthquake lakes formed in 1811 in the flood plain of the
       Lower Mississippi                                             404

  436. View of a crater lake in Costa Rica                           405

  437. Diagrams to illustrate the characteristics of crater lakes    406

  438. View of Snag Lake, a coulée lake in California                406

  439. Diagrams to illustrate the characteristics of morainal lakes  407

  440. Diagram to show the manner of formation of pit lakes          408

  441. Diagrams to illustrate the characteristics of pit lakes       408

  442. Diagram to show the manner of formation of glint lakes        409

  443. Map of a series of glint lakes on the boundary of Sweden and
       Norway                                                        409

  444. Map of ice-dam lakes near the Norwegian boundary of Sweden    410

  445. Wave-cut terrace of a former ice-dam lake in Sweden           410

  446. View of the Márjelen Lake from the summit of the Eggishorn    411

  447. Diagrams to illustrate the arrangement and the characters of
       rock-basin lakes                                              412

  448. Convict Lake, a valley-moraine lake of California             413

  449. Lake basins produced by successive slides from the steep
       walls of a glaciated mountain valley                          414

  450. Lake Garda, a border lake upon the site of a piedmont apron   414

  451. Diagrams to bring out the characteristics of ox-bow lakes     415

  452. Diagrammatic section to illustrate the formation of
       saucer-like basins between the levees of streams on a flood
       plain                                                         415

  453. Saucer lakes upon the bed of the former river Warren          416

  454. Levee lakes developed in series within meanders in a delta
       plain                                                         417

  455. Raft lakes along the banks of the Red River in Arkansas and
       Louisiana                                                     418

  456. Map of the Swiss lakes Thun and Brienz                        419

  457. Delta lakes formed at the mouth of the Mississippi            419

  458. Delta lakes at the margin of the Nile delta                   420

  459. Diagrams to illustrate the characteristics of barrier lakes   420

  460. Dune lakes on the coast of France                             421

  461. Sink lakes in Florida, with a schematic diagram to
       illustrate the manner of their formation                      421

  462. Map of the Arve and the Upper Rhone                           426

  463. View of the Arve and the Rhone at their junction              427

  464. A village in Switzerland built upon a strath at the head of
       Lake Poschiavo                                                428

  465. View of the floating bog and surrounding zones of vegetation
       in a small glacial lake                                       429

  466. Diagram to show how small lakes are transformed into peat
       bogs                                                          430

  467. Map to show the anomalous position of the delta in Lake St.
       Clair                                                         431

  468. A bowlder wall upon the shore of a small lake                 432

  469. Diagrams to show the effect of ice shove in producing ice
       ramparts upon the shores of lakes                             433

  470. Various forms of ice ramparts                                 433

  471. Map of Lake Mendota, showing the position of the ridge which
       forms from ice expansion and the ice ramparts upon the shores 434

  472. The great multiple mountain arc of Sewestan, British India    436

  473. Diagrams to illustrate the theories of origin of mountain
       arcs                                                          437

  474. Festoons of mountain arcs about the borders of the Pacific
       Ocean                                                         438

  475. The interrupted Armorican Mountains common to western Europe
       and eastern North America                                     438

  476. A zone of diverse displacement in the western United States   439

  477. Section of an East African block mountain                     439

  478. Tilted crust blocks in the Queantoweap valley                 440

  479. View of the laccolite of the Carriso Mountain                 441

  480. Map of laccolitic mountains                                   441

  481. Ideal sections of laccolite and bysmalite                     442

  482. The gabled façade largely developed in desert landscapes      443

  483. Balloon view of the Mythen in Switzerland                     444

  484. The battlement type of erosion mountain                       445

  485. Symmetrically formed low islands repeated in ranks upon
       Temagami Lake, Ontario                                        445

  486. Forms of crystals of a number of minerals                     454

  487. Forms of crystals of a number of minerals                     457

  488. A student’s contour map                                       469

  489. Models to represent outcrops of rock                          472

  490. Special laboratory table set with a problem in geological
       mapping which is solved in Figs. 47 and 48                    472

  491. Three field maps to be used as suggestions in arranging
       laboratory table for problems in the preparation of areal
       geological maps                                               473

  492. Sketch map of Western Scotland and the Inner Hebrides to
       show location of some points of special geological interest   481

  493. Outline map of a geological pilgrimage across the continent
       of Europe                                                     483


 Am. Geol.: American Geologist.

 Am. Jour. Sci.: American Journal of Science, New Haven.

 Ann. de Géogr.: Annales de Géographie, Paris.

 Ann. Rept. Geol. and Geogr. Surv. Ter.: Annual Report of the
 Geological and Geographical Survey of the Territories (Hayden),

 Ann. Rept. Geol. and Nat. Hist. Surv. Minn.: Annual Report of the
 Geological and Natural History Survey of Minnesota, Minneapolis.

 Ann. Rept. Mich. Geol. Surv.: Annual Report of the Michigan Geological
 Survey, Lansing.

 Ann. Rept. U. S. Geol. Surv.: Annual Report of the United States
 Geological Survey, Washington.

 Bull. Am. Geogr. Soc.: Bulletin of the American Geographical Society,
 New York.

 Bull. Earthq. Inv. Com. Japan: Bulletin of the Earthquake
 Investigation Committee of Japan, Tokyo.

 Bull. Geogr. Soc. Philadelphia: Bulletin of the Geographical Society
 of Philadelphia.

 Bull. Geol. Soc. Am.: Bulletin of the Geological Society of America.

 Bull. Mus. Comp. Zoöl.: Bulletin of the Museum of Comparative Zoölogy,
 Harvard College, Cambridge.

 Bull. N. Y. State Mus.: Bulletin of the New York State Museum, Albany.

 Bull. Soc. Belge d’Astronomie: Bulletin de la Société Belge
 d’Astronomie, Brussels.

 Bull. Soc. Belge Géol.: Bulletin de la Société Belge de Géologie,

 Bull. Soc. Sc. Nat. Neuchâtel: Bulletin de la Société des Sciences
 Naturelles de Neuchâtel.

 Bull. Univ. Calif. Dept. Geol.: Bulletin of the University of
 California, Department of Geology, Berkeley.

 Bull. U. S. Geol. Surv.: Bulletin of the United States Geological
 Survey, Washington.

 Bull. Wis. Geol. and Nat. Hist. Surv.: Bulletin of the Wisconsin
 Geological and Natural History Survey, Madison.

 C. R. Cong. Géol. Intern.: Comptes Rendus de la Congrès Géologique

 Dept. of Mines, Geol. Surv. Branch, Canada: Department of Mines,
 Geological Survey Branch, Canada.

 Geogr. Abh.: Geographische Abhandlungen.

 Geogr. Jour.: Geographical Journal, London.

 Geol. Folio U. S. Geol. Surv.: Geological Folio of the United States
 Geological Survey.

 Geol. Mag.: Geological Magazine, London (sections designated by

 Jour. Am. Geogr. Soc.: Journal of the American Geographical Society,
 New York.

 Jour. Coll. Sci. Imp. Univ. Tokyo: Journal of the College of Science
 of the Imperial University, Tokyo, Japan.

 Jour. Geol.: Journal of Geology, Chicago.

 Jour. Sch. Geogr.: Journal of School Geography.

 Livret Guide Cong. Géol. Intern.: Livret Guide Congrès Géologique

 Mem. Geol. Surv. India: Memoirs of the Geological Survey of India,

 Mitt. Geogr. Ges. Hamb.: Mitteilungen der Geographische Gesellschaft,

 Mon. U. S. Geol. Surv.: Monograph of the United States Geological
 Survey, Washington.

 Nat. Geogr. Mag.: National Geographic Magazine, Washington.

 Nat. Geogr. Mon.: National Geographic Monographs, American Book
 Company, New York.

 Naturw. Wochenschr.: Naturwissenschaftliche Wochenschrift.

 Pet. Mitt.: Petermanns Mittheilungen aus Justus Perthes’
 Geographischer Anstalt, Gotha.

 Pet. Mitt., Ergänzungsh. or Erg.: Petermanns Mittheilungen, Gotha
 (Ergänzungsheft or Supplementary Paper).

 Phil. Jour. Sci.: Philippine Journal of Science, Manila.

 Phil. Trans.: Philosophical Transactions of the Royal Society, London.

 Proc. Am. Acad. Arts and Sci.: Proceedings of the American Academy of
 Arts and Sciences.

 Proc. Am. Assoc. Adv. Sci.: Proceedings of the American Association
 for the Advancement of Science.

 Proc. Am. Phil. Soc.: Proceedings of the American Philosophical
 Society, Philadelphia.

 Proc. Bost. Soc. Nat. Hist.: Proceedings of the Boston Society of
 Natural History, Boston.

 Proc. Ind. Acad. Sci.: Proceedings of the Indiana Academy of Science.

 Proc. Linn. Soc. New South Wales: Proceedings of the Linnean Society
 of New South Wales.

 Proc. Ohio State Acad. Sci.: Proceedings of the Ohio State Academy of

 Prof. Pap. U. S. Geol. Surv.: Professional Paper of the United States
 Geological Survey, Washington.

 Pub. Carneg. Inst.: Publication of the Carnegie Institution of

 Pub. Mich. Geol. and Biol. Surv.: Publication of the Michigan
 Geological and Biological Survey, Lansing.

 Quart. Jour. Geol. Soc. Lond.: Quarterly Journal of the Geological
 Society, London.

 Rept. Brit. Assoc. Adv. Sci.: Report of the British Association for
 the Advancement of Science.

 Rept. Geol. Surv. Mich.: Report of the Geological Survey of Michigan,

 Rept. Mich. Acad. Sci.: Report of the Michigan Academy of Science,

 Rept. Nat. Conserv. Com.: Report of the National Conservation
 Commission, Washington.

 Rept. Smithson. Inst.: Report of the Smithsonian Institution,

 Sci. Bull. Brooklyn Inst. Arts and Sci.: Science Bulletin of the
 Brooklyn Institute of Arts and Sciences.

 Scot. Geogr. Mag.: Scottish Geographic Magazine, Edinburgh.

 Smith. Cont. to Knowl.: Smithsonian Contributions to Knowledge,

 Tech. Quart.: Technology Quarterly of the Massachusetts Institute of
 Technology, Boston.

 Trans. Am. Inst. Min. Eng.: Transactions of the American Institute of
 Mining Engineers, New York.

 Trans. Roy. Dublin Soc.: Transactions of the Royal Dublin Society.

 Trans. Seis. Soc. Japan: Transactions of the Seismological Society of
 Japan, Tokyo.

 Trans. Wis. Acad. Sci.: Transactions of the Wisconsin Academy of
 Sciences, Arts, and Letters, Madison.

 U. S. Geogr. and Geol. Surv. Rocky Mt. Region: United States
 Geographical and Geological Survey of the Rocky Mountain Region
 (Powell), Washington.

 Zeit. d. Gesell. f. Erdk. z. Berlin: Zeitschrift der Gesellschaft für
 Erdkunde zu Berlin.

 Zeit. f. Gletscherk: Zeitschrift für Gletscherkunde, Berlin.




=The sources of the history.=—The science which deals with the
chapters of earth history that antedate the earliest human writings
is geology. The pages of the record are the layers of rock which make
up the outer shell of our world. Here as in old manuscripts pages are
sometimes found to be missing, and on others the writing is largely
effaced so as to be indistinct or even illegible. An intelligent
interpretation of this record requires a knowledge of the materials
and the structure of the earth, as well as a proper conception of the
agencies which have caused change and so developed the history. These
agencies in operation are physical and chemical processes, and so the
sciences of physics and chemistry are fundamental in any extended study
of geology. Not only is geology, so to speak, founded upon chemistry
and physics, but its field overlaps that of many other important
sciences. The earliest earth history has to do with the form, size, and
physical condition of a minor planet in the solar system. The earliest
portion of the story belongs therefore to astronomy, and no sharp line
can be drawn to separate this chapter from those later ones which are
more clearly within the domain of geology.

=Subdivisions of geology.=—The terms “cosmic geology” and “astronomic
geology” have sometimes been used to cover the astronomy of the earth
planet. The later earth history develops, among other things, the
varied forms of animal and vegetable life which have had a definite
order of appearance. Their study is to a large extent zoölogy and
botany, though here considered from an essentially different viewpoint.
This subdivision of our science is called paleontological geology or
paleontology, which in common usage includes the plant as well as the
animal world, or what is sometimes called paleobotany. In order to fix
the order of events in geological history, these biological studies
are necessary, for the pages of the record have many of them been
misplaced as a result of the vicissitudes of earth history, and the
remains of life in the rock layers supply a pagination from which it is
possible to correctly rearrange the misplaced pages. As compiled into a
consecutive history of the earth since life appeared upon it, we have
the division of historical geology; though this differs but little from
stratigraphical geology, the emphasis in the case of the former being
placed on the history itself and in the latter upon the arrangement of
events—the pagination of the record.

So far as they are known to us, the materials of which the earth is
composed are minerals grouped into various characteristic aggregates
known as rocks. Here the science is founded upon mineralogy as well as
chemistry, and a study of the rock materials of the earth is designated
petrographical geology or petrography. The various rocks which enter
into the composition of the earth’s outer shell—the only portion known
to us from direct observation—are built into it in an architecture
which, when carefully studied, discloses important events in the
earth’s history. The division of the science which is concerned with
earth architecture is geotectonic or structural geology.

=The study of earth features and their significance.=—The features
upon the surface of the earth have all their deep significance, and
if properly understood, a flood of light is thrown, not only upon
present conditions, but upon many chapters of the earth’s earlier
history. Here the relation of our study to topography and geography is
very close, so that the lines of separation are but ill defined. The
terms “physiographical geology”, “physiography”, and “geomorphology”
are concerned with the configuration of the earth’s surface—its
physiognomy—and with the genesis of its individual surface features.
It is this genetical side of physiography which separates it from
topography and lends it an absorbing interest, though it causes it
to largely overlap the division of dynamical geology or the study of
geological processes. In fact, the difference between dynamical geology
and physiography is largely one of emphasis, the stress being laid
upon the processes in the former and upon the resultant features in the

Under dynamical geology are included important subdivisions, such
as seismic geology, or the study of earthquakes, and vulcanology,
or the study of volcanoes. Another large subject, glacial geology,
belongs within the broad frontier common to both dynamical geology
and physiography. A relatively new subdivision of geological science
is orientational geology, which is concerned with the trend of earth
features, and is closely related both to physiography and to dynamical
and structural geology.

=Tabular recapitulation.=—In a slightly different arrangement from the
above order of mention, the subdivisions of geology are as follows:—

_Subdivisions of Geology_

  _Petrographical Geology._              Materials of the earth.
  _Geotectonic Geology._                 Architecture of the earth’s
                                            outer shell.
  _Dynamical Geology._                   Earth processes.
    Seismic Geology—earthquakes.
      Vulcanology—volcanoes. Glacial
      Geology—glaciers, etc.

  _Physiographical Geology._             Earth physiognomy and its

  _Orientational Geology._               The arrangement and the trend
                                           of earth features.

In one way or another all of the above subdivisions of geology are in
some way concerned in the genesis of earth physiognomy, and they must
therefore be given consideration in a work which is devoted to a study
of the meaning of earth features. The compiled record of the rocks is,
however, something quite apart and without pertinence to the present
work. As already indicated its subdivisions are:—

  _Astronomic Geology._          Planetary history of the earth.
  _Statigraphic Geology._        The pagination of earth records.
  _Historical Geology._          The compiled record and its
  _Paleontological Geology._     The evolution of life upon the earth.

In every attempt at systematic arrangement difficulties are
encountered, usually because no one consideration can be used
throughout as the basis of classification. Such terms as “economic
geology” and “mining geology” have either a pedagogical or a commercial
significance, and so would hardly fit into the system which we have

=Geological processes not universal.=—It is inevitable that the
geology of regions which are easily accessible for study should
have absorbed the larger measure of attention; but it should not be
forgotten that geology is concerned with the history of the entire
world, and that perspective will be lost and erroneous conclusions
drawn if local conditions are kept too often before the eyes. To
illustrate by a single instance, the best studied regions of the globe
are those in which fairly abundant precipitation in the form of rain
has fitted the land for easy conditions of life, and has thus permitted
the development of a high civilization. In degree, and to some extent
also in kind, geologic processes are markedly different within those
widely extended regions which, because either arid or cold, have been
but ill fitted for human habitation. Yet in the historical development
of the earth, those geologic processes which obtain in desert or
polar regions are none the less important because less often and less
carefully observed.

=Change, and not stability, the order of nature.=—Man is ever prone to
emphasize the importance of apparent facts to the disadvantage of those
less clearly revealed though equally potent. The ancient notion of the
_terra firma_, the safe and solid ground, arose because of its contrast
with the far more mobile bodies of water; but this illusion is quickly
dispelled with the sudden quaking of the ground. Experience has clearly
shown that, both upon and beneath the earth’s surface, chemical and
physical changes are going on, subject to but little interruption. “The
hills rock-ribbed and ancient as the sun” is a poetical metaphor; for
the Himalayas, the loftiest mountains upon the globe, were, to speak
in geological terms, raised from the sea but yesterday. Even to-day
they are pushing up their heads, only to be relentlessly planed down
through the action of the atmosphere, of ice, and of running water.
Even more than has generally been supposed, the earth suffers change.
Often within the space of a few seconds, to the accompaniment of a
heavy earthquake, many square miles of territory are bodily uplifted,
while neighboring areas may be relatively depressed. Thus change, and
not stability, is the order of nature.

=Observational geology _versus_ speculative philosophy.=—There appears
to be a more or less prevalent notion that the views which are held
by scientists in one generation are abandoned by those of the next;
and this is apt to lead to the belief that little is really known and
that much is largely guessed. Some ground there undoubtedly is for
such skepticism, though much of it may be accounted for by a general
failure among scientists, as well as others, to clearly differentiate
that which is essentially speculative from what is based broadly upon
observed facts. Even with extended observation, the possibility of
explaining the facts in more than one way is not excluded; but the
line is nevertheless a broad one which separates this entire field
of observation from what is essentially speculative philosophy. To
illustrate: the mechanics of the action which goes on within volcanic
craters is now fairly well understood as a result of many and extended
observations, and it is little likely that future generations of
geologists will discredit the main conclusions which have been reached.
The cause of the rise of the lava to the earth’s surface is, on the
other hand, much less clearly demonstrated, and the views which are
held express rather the differing opinions than any clear deductions
from observation. Again, and similarly, the physical history of the
great continental glaciers of the so-called “ice age” is far more
thoroughly known than that of any existing glacier of the same type;
but the cause of the climatic changes which brought on the glaciation
is still largely a matter for speculation.

In the present work, the attempt will be, so far as possible, to give
an exposition of geologic processes and the earth features which result
from them, with hints only at those ultimate causes which lie hidden in
the background.

=The scientific attitude and temper.=—The student of science should
make it his aim, not only clearly to separate in his studies the
proximate from the ultimate causes of observed phenomena, but he should
keep his mind always open for reaching individual conclusions. No
doctrines should be accepted finally upon faith merely, but subject
rather to his own reasoning processes. This should not be interpreted
to mean that concerning matters of which he knows little or nothing
he should not pay respect to the recognized authorities; but his
acceptance of any theory should be subject to review so soon as his
own horizon has been sufficiently enlarged. False theories could hardly
have endured so long in the past, had not too great respect been given
to authorities, and individual reasoning processes been held too long
in subjection.

=The value of the hypothesis.=—Because all the facts necessary for a
full interpretation of observed phenomena are not at one’s hand, this
should not be made to stand in the way of provisional explanations. If
science is to advance, the use of hypothesis is absolutely essential;
but the particular hypothesis adopted should be regarded as temporary
and as indicating a line of observation or of experimentation which
is to be followed in testing it. Thus regarded with an open mind,
inadequate hypotheses are eventually found to be untenable, whereas
correct explanations of the facts by the same process are confirmed.
Most hypotheses of science are but partially correct, for we now “see
through a glass darkly”; but even so, if properly tested, the false
elements in the hypothesis are one after the other eliminated as the
embodied truth is confirmed and enlarged. Thus “working hypothesis”
passes into theory and becomes an integral part of science.


 The most comprehensive of general geological texts written in English
 is Chamberlin and Salisbury’s “Geology” in three volumes (Henry Holt,
 1904-1906), the first volume of which is devoted exclusively to
 geological processes and their results. An abridged one-volume edition
 of the work intended for use as a college text was issued in 1906
 (College Geology, Henry Holt). Other standard texts are:—

 SIR ARCHIBALD GEIKIE. Text-book of Geology, 4th ed. 2 vols. London,
 1902, pp. 1472.

 W. B. SCOTT. An Introduction to Geology. 2d ed. Macmillan, 1907, pp.

 J. D. DANA. Manual of Geology. New edition. American Book Company,
 1895, pp. 1087.

 JOSEPH LECONTE. Elements of Geology. (Revised by Fairchild.) Appleton,
 1905, pp. 667.

A very valuable guide to the recent literature of dynamical and
structural geology is Branner’s “Syllabus of a Course of Lectures on
Elementary Geology” (Stanford University, 1908).

On the relation of geology to landscape, a number of interesting books
have been written:—

 JAMES GEIKIE. Earth Sculpture or the Origin of Land-Forms. New York
 and London, 1896, pp. 397.

 JOHN E. MARR. The Scientific Study of Scenery. Methuen, London, 1900,
 pp. 368.

 SIR A. GEIKIE. The Scenery of Scotland. 3d ed. Macmillan, London,
 1901, pp. 540.

 SIR JOHN LUBBOCK. The Scenery of Switzerland and the Causes to which
 it is Due. Macmillan, London, 1896, pp. 480.

 LORD AVEBURY. The Scenery of England. Macmillan, London, 1902, pp. 534.

 SIR A. GEIKIE. Landscape in History, and Other Essays. Macmillan,
 London, 1905, pp. 352.

 N. S. SHALER. Aspects of the Earth. Scribners, New York, 1889, pp. 344.

 G. DE LA NOE ET EMM. DE MARGERIE. Les Formes du Terrain, Service
 Géographique de l’Armée. Paris, 1888, pp. 205, pls. 48.

 W. M. DAVIS. Practical Exercises in Physical Geography, with
 Accompanying Atlas. Ginn and Co., Boston, 1908, pp. 148, pls. 45.

 JOHN MUIR. The Mountains of California. Unwin, London, 1894, pp. 381.

Upon the use and interpretation of topographic maps in illustration of
characteristic earth features, the following are recommended:—

 R. D. SALISBURY and W. W. ATWOOD. The Interpretation of Topographic
 Maps, Prof. Pap., 60 U.S. Geol. Surv., pp. 84, pls. 170.

 D. W. JOHNSON and F. E. MATTHES. The Relation of Geology to
 Topography, in Breed and Hosmer’s Principles and Practice of
 Surveying, vol. 2. Wiley, New York, 1908.

 GÉNÉRAL BERTHAUT. Topologie, Étude du Terrain, Service Géographique de
 l’Armée. Paris, 1909, 2 vols., pp. 330 and 674, pls. 265.

The United States Geological Survey issues free of charge a list of
100 topographic atlas sheets which illustrate the more important
physiographic types. In his “Traité de Géographie Physique”, Professor
E. de Martonne has given at the end of each chapter the important
foreign maps which illustrate the physiographic types there described.

“The Principles of Geology”, by Sir Charles Lyell, published first in
three volumes, appeared in the years 1830-1833, and may be said to
mark the beginning of modern geology. Later reduced to two volumes, an
eleventh edition of the work was issued in 1872 (Appleton) and may be
profitably read and studied to-day by all students of geology. Those
familiar with the German language will derive both pleasure and profit
from a perusal of Neumayr’s “Erdgeschichte” (2d ed. revised by Uhlig.
Leipzig and Vienna, 2 vols., 1895-1897), and especially the first
volume, “Allgemeine Geologie.” A recent French work to be recommended
is Haug’s “Traité de Géologie” (Paris, 1907).

Some texts of physical geography may well be consulted, especially Emm.
de Martonne’s “Traité de Géographie Physique.” Colin, Paris, 1909, pp.
910, pls. 48, and figs. 396.

       *       *       *       *       *

NOTE. An explanatory list of abbreviations used in the reading
references follows the List of Illustrations.



=The lithosphere and its envelopes.=—The stony part of the earth is
known as the _lithosphere_, of which only a thin surface shell is known
to us from direct observation. The relatively unknown central portion,
or “core”, is sometimes referred to as the centrosphere. Inclosing the
lithosphere is a water envelope, the _hydrosphere_, which comprises
the oceans and inland bodies of water, and has a mass 1/4540 that of
the lithosphere. If uniformly distributed, the hydrosphere would cover
the lithosphere to the depth of about two miles, instead of being
collected in basins as it now is. Though apparently not continuous, if
we take into account the zone of underground water upon the continents,
the hydrosphere may properly be considered as a continuous film about
the lithosphere. It is a fact of much significance that all the ocean
basins are connected, so that the levels are adjusted to furnish a
common record of deposits over the entire surface that is sea-covered.

Enveloping the hydrosphere is the gaseous envelope, the _atmosphere_,
with a mass 1/1200000 that of the lithosphere. The atmosphere is a
mixture of the gases oxygen and nitrogen in parts by volume of one of
the former to four of the latter, with a relatively small percentage
of carbon dioxide. Locally, and at special seasons, the atmosphere
may be charged with relatively large percentages of water vapor; and
we shall see that both the carbon dioxide and the vapor contents are
of the utmost importance in geological processes and in the influence
upon climate. Unlike the water which composes the hydrosphere, the
gases of the atmosphere are compressible. Forced down by the weight of
superincumbent gas, the layers of the atmosphere at the level of the
sea sustain a pressure of about fifteen pounds to the square inch; but
this pressure steadily decreases in ascending to higher levels. From
direct instrumental observation, the air has now been investigated to
a height of more than twelve miles from the earth’s surface.

=The evolution of ideas concerning the earth’s figure.=—The ideas
which in all ages have been promulgated concerning the figure of the
earth have been many and varied. Though among them are not wanting the
purely speculative and fantastic, it will be interesting to pass in
review such theories as have grown directly out of observation.

The ancient Hebrews and the Babylonians were dwellers of the desert,
and in the mountains which bounded their horizon they saw the confines
of the earth. Pushing at last westward beyond the mountains, they
found the Mediterranean, and thus arrived at the view that the earth
was a disk with a rim of mountains which was floated upon water. The
rare but violent rainfalls to which they were accustomed—the desert
cloudburst—further led them to the belief that the mountain rim was
continued upward in a dome or firmament of transparent crystal upon
which the heavenly bodies were hung and from which out of “windows
of heaven” the water “which is above the earth” was poured out upon
the earth’s surface. Fantastic as this theory may seem to-day, it
was founded upon observation, and it well illustrates the dangers of
reasoning from observation within too limited a field.

As soon as men began to sail the sea, it was noticed that the water
surface is convex, for the masts of ships were found to remain
visible long after their hulls had disappeared below the horizon. It
is difficult to say how soon the idea of the earth’s rotundity was
acquired, but it is certainly of great antiquity. The Dominican monk
Vincentius of Beauvais, in a work completed in 1244, declared that the
surfaces of the earth and the sea were both spherical. The poet Dante
made it clear that these surfaces were one, and in his famous address
upon “The Water and the Land”, which was delivered in Verona on the
20th of January, 1320, he added a statement that the continents rise
higher than the ocean. His explanation of this was that the continents
are pulled up by the attraction of the fixed stars after the manner
of attraction of magnets, thus giving an early hint of the force of

The earth’s rotundity may be said to have been first proven when
Magellan’s ships in 1521 had accomplished the circumnavigation of
the globe. Circumnavigation, soon after again carried out by Sir
Francis Drake, proved that the earth is a closed body bounded by
curving surfaces in part enveloped by the oceans and everywhere by the
atmosphere. The great discovery of Copernicus in 1530 that the earth,
like Venus, Mars, and the other planets, revolves about the sun as a
part of a system, left little room for doubt that the figure of the
earth was essentially that of a sphere.

=The oblateness of the earth.=—Every schoolboy is to-day familiar with
the fact that the earth departs from a perfect spherical figure by
being flattened at the ends of its axis of rotation. The polar diameter
is usually given as 1/299 shorter than the equatorial one. This
oblateness of the spheroid was proven by geodesists when they came to
compare the lengths of measured degrees of arc upon meridians in high
and in low latitudes.


FIG. 1.—Diagrams to afford a correct impression of the measure of the
inequalities upon the earth’s surface compared to the earth’s radius.
The shell represented in _b_ is 1/100 of the earth’s radius, and in _a_
this zone is magnified for comparison with surface inequalities.]

The oblateness of the geoid is well understood from accepted hypotheses
to be the result of the once more rapid rotation of the planet when its
materials were more plastic, and hence more responsive to deformation.
An elastic hoop rotating rapidly about an axis in its plane appears to
the eye as a solid, and becomes flattened at the ends of its axis in
proportion as the velocity of rotation is increased. Like the earth,
the other planets in the solar system are similarly oblate and by
amounts dependent on the relative velocities of rotation.

The departure of the geoid from the spherical surface, owing to its
oblateness, is so small that in the figures which we shall use for
illustration it would be less than the thickness of a line. Since it
is well recognized and not important in our present consideration, we
shall for the time being speak of the figure of the earth in terms of
departures from a standard spherical surface.

=The arrangement of oceans and continents.=—There are other departures
from a spherical surface than the oblateness just referred to,
and these departures, while not large, are believed to be full of
significance. Lest the reader should gain a wrong impression of their
magnitude, it may be well to introduce a diagram drawn to scale and
representing prominent elevations and depressions of the earth (Fig. 1).

Wrong impressions concerning the figure of the lithosphere are
sometimes gained because its depressions are obliterated by the oceans.
The oceans are, indeed, useful to us in showing where the depressions
are located, but the figure of the earth which we are considering
is the naked surface of the rock. In a broad way, the earth’s shape
will be given by the arrangement of the oceans and the continents. As
soon as we take up the study of this arrangement, we find that quite
significant facts of distribution are disclosed.


FIG. 2.—Map on Mercator’s projection to show the reciprocal relation
of the land and sea areas (after Gregory and Arldt).]

One of the most significant facts involved in the distribution
of land and sea, is a concentration of the land areas within the
northern and the seas within the southern hemisphere. The noteworthy
exception is the occurrence of the great and high Antarctic continent
centered near the earth’s south pole; and there are extensions of the
northern continent as narrowing land masses to the southward of the
equator. Hardly less significant than the existence of land and water
hemispheres is the reciprocal or antipodal distribution of land and sea
(Fig. 2). A third fact of significance is a dovetailing together of sea
and land along an east-and-west direction. While the seas are generally
A-shaped and narrow northward, the land masses are V-shaped and narrow
southward, _but this occurs mainly in the southern hemisphere_. Lastly,
there is some indication of a belt of sea dividing the land masses
into northern and southern portions along the course of a great circle
which makes a small angle with the earth’s equator. Thus the western
continent is nearly divided by a mediterranean sea,—the Caribbean,—and
the eastern is in part so divided by the separation of Europe from


FIG. 3.—The form toward which the figure of the earth is tending, a
tetrahedron with symmetrically truncated angles.]

=The figure toward which the earth is tending.=—Thus far in our
discussion of the earth’s figure we have been guided entirely by the
present distribution of land and water. There are, however, depressions
upon the surface of the land, in some cases extending below the level
of the sea, which are not to-day occupied by water. By far the most
notable of these is the great Caspian Depression, which with its
extension divides central and eastern Asia upon the east from Africa
and Europe upon the west. This depression was quite recently occupied
by the sea, and when added to the present ocean basins to indicate
depressions of the lithosphere, it shows that the earth’s figure
departs from the standard spheroid _in the direction_ of the form
represented in Fig. 3. This form approximates to a tetrahedron, a
figure bounded by four equal triangular faces, here with symmetrically
truncated angles. Of all regular figures with plane surfaces the
tetrahedron has the smallest volume for a given surface, and it
presents moreover a reciprocal relation of projection to depression.
Every line passing through its center thus finds the surface nearer
than the average distance upon one side and correspondingly farther
upon the other (Fig. 4).

=Astronomical _versus_ geodetic observations.=—Confirmation of the
conclusions arrived at from the arrangement of oceans and continents
has been secured in other fields. It was pointed out that the earth’s
oblateness was proven by comparison of the measured degrees of latitude
upon the earth’s surface in lower and higher latitudes, the degree
being found longer as the pole is approached. Any variation from the
spherical surface must obviously increase the size of the measured
degree of latitude in proportion to the departure from the standard
form, and so the tetrahedral figure with one of its angles at the
south pole will require that the degrees of latitude be longer in the
southern than they are in the northern hemisphere. This has been found
by measurement to be the case, and the result is further confirmed
by pendulum studies upon the distribution of the earth’s attraction
or gravity. If less of the mass of the earth is concentrated in the
southern hemisphere, its attraction as measured in vibrations of the
pendulum should be correspondingly smaller.


FIG. 4.—A truncated tetrahedron, showing how the depression upon one
side of the center is balanced by the opposite projection.]

Other confirmations of the tetrahedral figure of the earth have been
derived from a comparison of astronomical data, which assume the earth
to be a perfect spheroid, with geodetic measurements, which are based
upon direct measurements. Thus the arc measured in an east-and-west
direction across Europe revealed a different curvature near the angle
of the tetrahedral figure from what was found farther to the eastward.

=Changes of figure during contraction of a spherical body.=—If
we inquire why the earth in cooling should tend to approach the
tetrahedral figure, an answer is easily found. When formed, the earth
appears to have been a but slightly oblate spheroid, or practically
a sphere—the shape which of all incloses the most space for a given
surface. Cooled and solidified at the surface to the temperature of
the surrounding air, and the core still hot and continuing to lose
heat, the core must continue to contract though the outer shell is
no longer able to do so. The superficial area being thus maintained
constant while the volume continues to diminish, the figure must change
from the initial one of greatest bulk to others of smaller volume,
and ultimately, if the process should continue indefinitely, to the
tetrahedron, which of all regular figures has the minimum volume for a
given surface.

That a contracting sphere does indeed pass through such a series of
changes has been shown by the behavior of contracting soap bubbles and
of rubber balloons, as well as by experiments upon the exhaustion of
air contained in hollow metal spheres of only moderate strength. In
all these instances, the ultimate form produced indicates an indenting
of four sides of the sphere which have the positions of the faces
of a tetrahedron. The late Professor Prinz of Brussels secured some
extremely interesting results in which he obtained intermediate forms
with six angles, but unfortunately these studies were not prepared for
publication at the time of his death.

The earth’s departure from the spheroid in the direction of the
modified tetrahedron is, as we have seen, no hypothesis, but observed
fact revealed in (1) the concentration of the land about a central
ocean in the northern hemisphere; in (2) the antipodal relation of the
land to the water areas, and in (3) the threefold subdivision of the
surface into north and south belts by the two greater oceans and the
Caspian Depression.

=The earlier figures of the earth.=—The manner in which continent and
ocean are dovetailed into each other in an east-and-west direction
has been generally adduced as additional evidence for the tetrahedral
figure as above described. Closer examination shows that instead of
being in harmony with this figure, it indicates a departure from it,
and, as we shall see, a significant departure which undoubtedly has its
origin in the earlier history of the planet. The mediterranean seas of
both the eastern and the western hemispheres likewise interfere with
the perfection of the tetrahedral figure and require an explanation.

Let us then examine in outline the past history of the world with
reference especially to the evolution of the continents and to the
times and the manners of surface change. It is now well known that
there have been three major periods of great deformation of the earth’s
shell. The first of these of which we have record came at the end of
the first great era of geologic history, the so-called Eozoic era;
a second great transformation came at the close of the second or
Paleozoic era; and a third began at the end of the next or Mesozoic
era, an adjustment which is apparently continuing to-day. Each of
these great surface deformations was accompanied by great volcanic
eruptions of which we have the evidence in the lavas remaining for our
inspection, and each was followed by the formation of great glaciers
which spread over large areas of the existing continents.

Before the earliest of these great changes, the earth appears to have
approximated in its figure somewhat closely to the ideal spheroid, for
it was everywhere enveloped in the hydrosphere as a universal ocean.
Toward the close of this period came the adjustments which brought
the lithosphere to protrude through the hydrosphere in shield-like
continents whose distribution, as shown by the rocks of this period, is
of great significance. Within the northern hemisphere rose three land
shields spaced at nearly equal intervals and at nearly equal distances
from the northern pole. One of these was centered where now is Hudson
Bay, another about the present Baltic Sea, and the relics of the third
are found in northeastern Siberia. These earliest continents have been
referred to as the Laurentian, Baltic, and Angara shields. Within
the southern hemisphere shields appear to have developed in somewhat
similar grouping, namely, in South America, in South Africa, and in
Australia (Figs. 3 and 5).


FIG. 5.—Approximations to earlier and present figures of the earth.]

These coigns or angles of a form into which the earlier spheroid of the
earth was being transformed have persisted through the greater part
of subsequent geologic time, and have been enlarged by the growth of
sediments about them as well as by the later elevation and wrinkling
of these deposits into marginal mountain ranges.

=The continents and oceans which arose at the close of the Paleozoic
era.=—At the close of the second great era in the recorded history
of the earth, the now somewhat enlarged continents were profoundly
altered during a series of convulsive movements within the surface
shell of the lithosphere. When these convulsions were over, there was
a new disposition of land and sea, but one quite different from the
present arrangement. Instead of being extended in north-south belts, as
they are at present, the continents stretched out in broad east-west
zones, one in the northern and the other in the southern hemisphere.
To the broad southern continent of which so little now remains, the
name “Gondwana Land” has been given, and to the sea which divided the
northern from the southern continent the name “Ocean of Tethys.” The
northern continent stretched across the site of the present Atlantic
Ocean as the “North Atlantis”, its northern shore to the westward
being somewhat farther south than the present northern coast of North
America, since life forms migrated in the northern ocean from the site
of Behring Sea to that of the present North Atlantic.

This arrangement of land and water during the middle period of the
earth’s recorded history, when considered with reference both to its
earlier and to its later evolution, may perhaps be best accounted
for by the assumption that the lithosphere was then shaped like Fig.
5 (middle). In this figure two truncated tetrahedrons are joined in
a common plane of contact which may be described as the twin plane.
This medial depression upon the lithosphere was occupied by the
intercontinental sea, the Ocean of Tethys.

Near the close of this second great era of the earth’s continental
history, crustal convulsions, which were perhaps the most remarkable in
the entire record, resulted in the almost complete disappearance of the
southern continent and a concentration of the land within the northern
hemisphere as a somewhat interrupted belt surrounding a central polar
ocean (Figs. 3 and 5).

Upon the assumption of twin tetrahedrons in the intermediate era of
continental evolution, both the Ocean of Tethys of that time and its
present remnants, the Caribbean and Mediterranean seas, are accounted
for. The V-shaped continent extensions and the A-shaped oceans of the
southern hemisphere (Fig. 2) may likewise be considered as relics of
the now largely submerged tetrahedron of the southern hemisphere, since
this had its apex to the northward (Fig. 6).


FIG. 6.—Diagrams for comparison of shore lines upon tetrahedrons which
have an angle, the first at the south and the second at the north.]

Thus we see that the lithosphere can scarcely be regarded as a perfect
spheroid, since in the course of geologic ages it has undergone
successive departures from this original form. In its present state it
has been described as tetrahedral, though we must keep in mind that
the sharp angles of that figure are deeply truncated. The soundings
first by Nansen and more recently by Peary in the Arctic basin, far to
the north of the continental border, showed that this depression is
characterized by profound depths, and so have afforded confirmation
of the tetrahedral figure. To match this depression at the northern
extremity of the earth’s axis, a high continent reaching to elevations
in excess of 10,000 feet has been penetrated by Sir Ernest Shackleton
at the opposite extremity of this polar diameter. Considering its size
and its elevation, the Antarctic continent with its glacier mantle is
the largest protuberance upon the surface of the lithosphere.

In our study of the departures of the earth from the standard
spheroidal surface, we might even go a step farther and show how
the tetrahedron, which best represents the symmetry of the present
figure, is somewhat deformed by a flattening perpendicular to the
Pacific Ocean. To draw attention to this flattening of the earth, it
has sometimes been described as “potato-shaped”, since the outline
perpendicular to this face is imperfectly heart-shaped or like a
flattened “peg top.”


FIG. 7.—The continents with submerged portions added (after Gilbert).]

=The flooded portions of the present continents.=—We are accustomed
to think of the continents as ending at the shores of the oceans.
If, however, we are to regard them as platforms which rise from the
ocean depressions, their margins should be considerably extended, for
a submerged shelf now practically surrounds all the continents to a
nearly uniform depth of 100 fathoms or 600 feet. The oceans thus more
than fill their basins and may be thought of as spilling over upon the
continents. In Fig. 7, the submerged portions of the continents have
been joined to those usually represented, thus adding about 10,000,000
square miles to their area, and giving them one third, instead of one
fourth, of the lithosphere surface.


FIG. 8.—Diagram to indicate the altitude of different parts of the
lithosphere surface.]

=The floors of the hydrosphere and atmosphere.=—The highest altitudes
upon the continents and the profoundest deeps of the ocean are each
removed about 30,000 feet, or nearly 6 miles, from the level of the
sea. In comparison with the entire surface of the lithosphere, these
extremes of elevation represent such small areas as to be almost
inappreciable. Only about 1/80 of the lithosphere surface rises more
than 6000 feet above sea level, and about the same proportion lies
deeper than 18,000 feet below the same datum plane (Fig. 8). Almost
the entire area of the lithosphere is included either in the so-called
continental plateau or platform, in the oceanic platform, or in the
slope which separates the two. The continental platform includes the
continental shelf above referred to, and represents about one third of
the entire area of the planet. This platform has a range of elevation
from 6000 feet above to 600 feet below sea level and has an average
altitude of about 2300 feet. The oceanic platform slopes more steeply,
ranges in depth from 12,000 to 18,000 feet below sea level, and
comprises about one half the lithosphere surface. The remaining portion
of the surface, something less than one eighth of all, is included in
the steep slopes between the two platforms, between 600 and 12,000
feet below sea. The two platforms and the slope between them must
not, however, be thought of as continuous features upon the surface,
but merely as representing the average elevations of portions of the


 On the evolution of ideas concerning the earth’s figure:—

 SUESS. The Face of the Earth (Clarendon Press, 1906), vol. 2, Chapter

 V. ZITTEL. History of Geology and Paleontology (Walter Scott, London,
 1901), Chapters 1-2.

The departure of the spheroid toward the tetrahedron:—

 W. LOWTHIAN GREEN. Vestiges of the Molten Globe, Part 1. London, 1875.
 (Now a rare work, but it contains the original statement of the idea.)

 J. W. GREGORY. The Plan of the Earth and Its Causes, Geogr. Jour.,
 vol. 13, 1899, pp. 225-251 (the best general statement of the
 arguments for a tetrahedral form).

 W. PRINZ. L’échelle reduite des expériences géologiques, Bull. Soc.
 Belge d’Astronomie, 1899.

 B. K. EMERSON. The Tetrahedral Earth and Zone of the Intercontinental
 Seas, Bull. Geol. Soc. Am., vol. 11, 1911, pp. 61-106, pls. 9-14.

 M. P. RUDSKI. Physik der Erde (Tauchnitz, Leipzig, 1911), Chapters
 1-3 (the best discussion of the geoid from the purely mathematical
 standpoint, so far as the spheroid is concerned).

The earlier figures of the earth:—

 TH. ARLDT. Die Entwicklung der Kontinente und ihrer Lebewelt.
 Engelmann, Leipzig, 1907. (Contains a valuable series of map plates,
 showing the probable boundaries of the continents in the different
 geological periods).



=The rigid quality of our planet.=—For a long time it was supposed
that the solid earth constituted a crust only which was floated
upon a liquid interior. This notion was clearly an outgrowth of the
then generally accepted Laplacian hypothesis of the origin of the
universe, which assumed fluid interiors for the planets, the crust
being suggested by the winter crust of frozen water upon the surface of
our inland lakes. To-day the nebular hypothesis in the Laplacian form
is fast giving place to quite different conceptions, in which solid
particles, and not gaseous ones, are conceived to have built up the
lithosphere. The analogy with frozen water has likewise been abandoned
with the discovery that frozen rock, instead of floating, sinks in its
molten equivalent.

Yet even more cogent arguments have been brought forward to show that
whatever may be the state of aggregation within the earth’s core—and
it may be different from any now known to us—it nevertheless has many
of the properties recognized as belonging to solid and rigid bodies.
Provisionally, therefore, we may regard the earth’s core as rigid and
essentially solid. It was long ago pointed out by the late Lord Kelvin
that if our lithosphere were not more rigid than a ball of glass of the
same size, it would be constantly passing through periodic six-hourly
distortions of great amplitude in response to the varying attractions
of the moon. An equally striking argument emanating from the same high
authority is furnished by the well-known egg-spinning demonstration.
For illustration, Kelvin was accustomed to take two eggs, one boiled
and the other raw, and attempt to spin them upon their ends. For the
boiled, and essentially solid, egg this is easily accomplished, but
internal friction of the liquid contents of the raw egg quickly stops
any rotary motion which is imparted to it. Upon the same grounds it
is argued that had the earth’s interior possessed the properties of a
liquid, rotation must long since have ceased.

A stronger proof of earth rigidity than either of these has been lately
furnished by the instrumental study of earthquakes. With the delicate
apparatus which is now installed for the purpose, heavy earthquakes may
be sensed which have occurred anywhere upon the earth’s surface, the
earth movement sending its own message by the shortest route through
the core of the earth to the observing station. A heavy shock which
occurs in New Zealand is recorded in England, almost diametrically
opposite, in about twenty-one minutes after its occurrence. The laws
of wave propagation and their relation to the properties of the
transmitting medium are well known, and in order to explain such
extraordinary velocity it is necessary to assume that for such impulses
the earth’s interior is much more rigid than the finest tool steel.

=Probable composition of the earth’s core.=—In deriving views
concerning the nature of the earth’s interior we are greatly aided
by astronomical studies. The common origin long ago indicated for
the planets of the solar system and the sun has been confirmed by
the analysis of light with the aid of the spectroscope. It has thus
been found that the same chemical elements which we find in the earth
are present also in the sun and in the other stellar bodies. Again,
the group of planets of the solar system which are nearest to the
sun—Mercury, Venus, the Earth, and Mars—have each a high density, all
except Mars, the most distant, having specific gravities very closely
5½, that of Mars being about 4. This average specific gravity is also
that of the solid bodies, the so-called meteorites, which reach the
surface of our planet from the surrounding space. Yet though the earth
as a whole is thus found to have a specific gravity five and a half
times that of water, its surface shell has an average density of less
than half this value, or 2.7.

The study of meteorites has given us a possible clew to the nature of
the earth’s interior; for when both terrestrial and celestial rock
types are classified and placed in orderly arrangement, it is found
that the chemical elements which compose the two groups are identical,
and that these are united according to the same physical and chemical
laws. No new element has been discovered in the one group that has not
been found in the other, and though some compounds of these elements,
the minerals, occur in the earth’s crust that have not been found in
meteorites, and though some occur in meteorites which are not known
from the earth, yet of those which are common to both bodies there is
agreement, even to the minor details (Fig. 9). It is found, however,
that the commonest of the minerals in the earth’s shell are absent from
meteorites, as the commoner constituents of meteorites are wanting in
the earth’s crust. This observation would go far to show that we may in
the two cases be examining different portions of quite similar bodies;
and this view is strikingly confirmed when the rocks of the two groups
are arranged in the order of their densities (Fig. 9).


FIG. 9.—Diagram to show how terrestrial rocks grade into those of the
meteorites. 1, oxygen; 2, silicon; 3, aluminium; 4, alkali metals; 5,
alkaline earth metals; 6, iron, nickel, cobalt, etc.; _a_, granites and
rhyolites; _b_, syenites and trachytes; _c_, diorites and andesites;
_d_, gabbros and basalts; _e_, ultra-basic rocks; _f_, basic inclosures
in basalt, etc.; _g_, iron basalts of west Greenland; _h_, iron masses
of Ovifak, west Greenland; _a’-d’_, meteorites in order of density
(after Judd).]

In a broad way, density, structure, and chemical composition are all
similarly involved in the gradations illustrated by the diagram; and it
is significant that while there are terrestrial rocks not represented
by meteorites, the densest and the most unusual of the terrestrial
rocks are chemically almost identical with the less dense of the
celestial bodies.

=The earth a magnet.=—The denser, and likewise the more common, of
the meteorite rocks—the so-called meteoric irons—are composed almost
entirely of the elements iron, nickel, and cobalt. Such aggregates
are not known as yet from terrestrial sources, although transitional
types appear to exist upon the island of Disco off the west coast of
Greenland. If it were possible to explore the earth’s interior, would
such combinations of the iron minerals be encountered? Apart from the
surprising velocity of transmission of earthquake waves, the strongest
argument for an iron core to the lithosphere is found in the magnetic
property of the earth as a whole. The only magnetic elements known to
us are those of the heavy meteorites—iron, nickel, and cobalt,—and
the earth is, as we know, a great magnet whose northern pole in British
America and whose southern pole in Antarctica have at last been visited
by Amundsen and David, respectively. The specific gravity of iron is
7.15, and those of nickel and cobalt, which in the meteorites are
present in relatively small amounts, are 7.8 and 7.5, respectively.
Considering that the surface shell of the earth has a specific gravity
of 2.7, these values must be regarded as agreeing well with the
determined density of the earth (5.6) and the other planets of its
group (Mercury 5.7, Venus 5.4, Mars 4).

=The chemical constitution of the earth’s surface shell.=—The number
of the so-called chemical elements which enter into the earth’s
composition is more than eighty, but few of these figure as important
constituents of the portion known to us. Nearly one half of the mass
of this shell is oxygen, and more than a quarter is silicon. The
remaining quarter is largely made up of aluminium, iron, calcium,
magnesium, and the alkalies sodium and potassium, in the order named.
These eight constituent elements are thus the only ones which play any
important rôle in the composition of the earth’s surface shell. They
are not found there in the free condition, but combined in the definite
proportions characteristic of chemical compounds, and as such they are
known as _minerals_.

=The essential nature of crystals.=—A crystal we are accustomed to
think of as something transparent bounded by sharp edges and angles,
our ideas having been obtained largely from the gem minerals. This
outward symmetry of form is, however, but an expression of a power
which resides, so to speak, in the heart or soul of the crystal
individual—it has its own structural make-up, its individuality. No
more correct estimates of the comparison of crystal individualities
would be obtained by the study of outward forms alone of two minerals
than would be gained by a judgment of persons from the cut of their
clothing. Too often this outward dress tells only of the conditions by
which both men and crystals have been surrounded, and but little of the
power inherent in the individual. Many a battered mineral fragment with
little beauty to recommend it, when placed under suitable conditions
for its development, has grown into a marvel of beauty. Few minerals
are so mean that they have not within them this inherent power of
individuality which lifts them above the world of the amorphous and


FIG. 10.—Comparison of a crystalline with an amorphous substance when
expanded by heat and when attacked by acid.]

Just as the real nature of a person is first disclosed by his behavior
under trying circumstances, so of a crystal it is its conduct under
stress of one sort or another which brings out its real character. By
way of illustration let us prepare a sphere from the mineral quartz—it
matters not whether we destroy the beautiful outlines of the crystal or
employ a battered fragment—and then prepare a sphere of similar size
and shape from a noncrystalline or amorphous substance like glass. If
now these two spheres be introduced into a bath of oil and raised to a
higher temperature, the glass globe undergoes an enlargement without
change of its form; but the crystal ball reveals its individuality
by expanding into a spheroid in which each new dimension is nicely
adjusted to this more complex figure (Fig. 10).

We may, instead of submitting the two balls to the “trial by fire”,
allow each to be attacked by the powerful reagent, hydrofluoric acid.
The common glass under the attack of the acid remains as it was before,
a sphere, but with shrunken dimensions. The crystal, on the other hand,
is able to control the action of the solvent, and in so doing its
individuality is again revealed in a beautifully etched figure having
many curving outlines—it is as though the crystal had possessed a soul
which under this trial has been revealed. This glimpse into the nature
of the crystal, so as to reveal its structural beauty, is still more
surprising when the crystal is taken from the acid in the early stages
of the action and held close beneath the eye. Now the little etchings
upon the surface display each the individuality of the substance, and
joining with their neighbors they send out a beautifully symmetrical
and entirely characteristic picture (Fig. 11).


FIG. 11.—“Light figure” seen upon an etched surface of a crystal of
calcite (after Goldschmidt and Wright).]

=The lithosphere a complex of interlocking crystals.=—To the layman
the crystal is something rare and expensive, to be obtained from
a jeweler or to be seen displayed in the show cases of the great
museums. Yet the one most striking quality of the lithosphere which
separates it from the hydrosphere and the atmosphere is its crystalline
structure,—a structure belonging also to the meteorite, and with
little doubt to all the planets of the earth group. A snowflake caught
during its fall from the sky reveals all the delicate tracery of
crystal boundary; collected from a thick layer lying upon the ground,
it appears as an intricate aggregate of broken fragments more or less
firmly cemented together. And so it is of the lithosphere, for the
myriads of individuals are either the ruins of former crystals, or
they are grown together in such a manner that crystal facets had no
opportunity to develop.

Such mineral individuals as once possessed the crystal form may have
been broken and their surfaces ground away by mutual attrition under
the rhythmic beating of the waves upon a shore or in the continuous
rolling of pebbles on a stream bed, until as battered relics they
are piled away together in a bed of sand. Yet no amount of such rough
handling is sufficient to destroy the crystal individuality, and if
they are now surrounded with conditions which are suitable for their
growth, their individual nature again becomes revealed in new crystal
outlines. Many of our sandstones when turned in the bright sunlight
send out flashes of light to rival a bank of snow in early spring.
These bright flashes proceed from the facets of minute crystals formed
about each rounded grain of the sand, and if we examine them under a
lens, we may note the beauty of line formed with such exactness that
the most delicate instruments can detect no difference between the
similar angles of neighboring crystals (Fig. 12).


FIG. 12.—Battered sand grains which have taken on a new lease of life
and have developed a crystal form. _a_, a single grain grown into an
individual crystal; _b_, a parallel growth about a single grain; _c_,
growth of neighboring grains until they have mutually interfered and so
destroyed the crystal facets—the common condition within the mass of a
rock (after Irving and Van Hise).]

This individual nature of the crystal is believed to reside in a
symmetrical grouping of the chemical molecules of the substance into
larger and so-called “crystal molecules.” The crystal quality belongs
to the chemical elements and to their compounds in the solid condition,
but not to ordinary mixtures of them.

=Some properties of natural crystals, minerals.=—No two mineral
species appear in crystals of the same appearance, any more than two
animal species have been given the same form; and so minerals may be
recognized by the individual peculiarities of their crystals. Yet
for the reason that crystals have so generally been prevented from
developing or retaining their characteristic faces, in the vast number
of instances it is the behavior, and not the appearance, of the mineral
substance which is made use of for identification.

When a mineral is broken under the blow of a hammer, instead of
yielding an irregular fracture, like that of glass, it generally
tends to part along one or more directions so as to leave plane
surfaces. This property of _cleavage_ is strikingly illustrated
for a single direction in the mineral mica, for two directions in
feldspar, and for three directions in calcite or Iceland spar. Other
properties of minerals, such as hardness, specific gravity, luster,
color, fusibility, etc., are all made use of in rough determinations
of the minerals. Far more delicate methods depend upon the behavior
of minerals when observed in polarized light, and such behavior is
the basis of those branches of geological science known as optical
mineralogy and as microscopical petrography. An outline description of
some of the common minerals and the means for identifying them will be
found in appendix A.

=The alterations of minerals.=—By far the larger number of minerals
have been formed in the cooling and consequent consolidation of molten
rock material such as during a volcanic eruption reaches the earth’s
surface as lava. Beginning their growth at many points within the
viscous mass, the individual crystals eventually may grow together and
so prevent a development of their crystal faces.

Another class of minerals are deposited from solution in water within
the cavities and fissures of the rocks; and if this process ceases
before the cavities have been completely closed, the minerals are
found projecting from the walls in a beautiful lining of crystal—the
_Krystallkeller_ or “crystal cellar.” It is from such pockets or veins
within the rocks that the valuable ores are obtained, as are the
crystals which are displayed in our mineral cabinets.


FIG. 13.—Crystal of garnet developed in a schist with grains of quartz
included because not assimilated.]

There is, however, a third process by which minerals are formed,
and minerals of this class are produced within the solid rock as a
product of the alteration of preëxisting minerals. Under the enormous
pressures of the rocks deep below the earth’s surface, they are as
permeable to the percolating waters as is a sponge at the surface.
Under these conditions certain minerals are dissolved and their
material redeposited after traveling in the solution, or solution
and redeposition of mineral matter may go on together within the mass
of the same rock. One new mineral may have been produced from the
dissolved materials of a number of earlier species, or several new
minerals may be the result of the alteration of a preëxisting mineral
with a more complex chemical structure. Where the new mineral has been
formed “in place”, it has sometimes been able to utilize the materials
of all the minerals which before existed there, or it may have been
obliged to inclose within itself those earlier constituents which it
could not assimilate in its own structure (Fig. 13).


FIG. 14.—A crystal of augite within the mass of a rock altered in
part to form a rim of the minerals hornblende and magnetite. Note the
original outline of the augite crystal.]

At other times a crystal which is imbedded in rock has been attacked
upon its surface by the percolating solutions, and the dissolved
materials have been deposited in place as a crown of new minerals
which steadily widens its zone until the center is reached and the
original crystal has been entirely transformed (Fig. 14). It is
sometimes possible to say that the action by which these changes have
been brought about has involved a nice adjustment of supply of the
chemical constituents necessary to the formation of the new mineral or
minerals. In rocks which are aggregates of several mineral species, a
newly formed mineral may appear only at the common margin of certain of
these species, thus showing that they supply those chemical elements
which were necessary to the formation of the new substance (Fig. 15).
Thus it is seen that below the earth’s surface chemical reactions are
constantly going on, and the earlier rocks are thus locally being
transformed into others of a different mineral constitution.


FIG. 15.—A new mineral (hornblende) forming as an intermediate
“reaction rim” between the mineral having irregular fractures (olivine)
and the dusty white mineral (lime-soda feldspar).]

Near the earth’s surface the carbon dioxide and the moisture which are
present in the atmosphere are constantly changing the exposed portions
of the lithosphere into carbonates, hydrates, and oxides. These
compounds are more soluble than are the minerals out of which they were
formed, and they are also more bulky and so tend to crack off from
the parent mass on which they were formed. As we are to see, for both
of these reasons the surface rocks of the lithosphere succumb to this
attack from the atmosphere.

In connection with those wrinklings of the surface shell of the
lithosphere from which mountains result, the underlying rocks are
subjected to great strains, and even where no visible partings are
produced, the rocks are deformed so that individual minerals may be
bent into crescent-shaped or S-shaped forms, or they are parted into
one or more fragments which remain imbedded within the rock.


 Theories of origin of the earth:—

 THOMSON and TAIT. Natural Philosophy. 2d ed. Cambridge, 1883, pp. 422.

 T. C. CHAMBERLIN. Chamberlin and Salisbury’s Geology, vol. 2, pp. 1-81.

Rigidity of the earth:—

 LORD KELVIN. The Internal Condition of the Earth as to Temperature,
 Fluidity, and Rigidity, Popular Lectures and Addresses, vol. 2, pp.
 299-318; Review of evidence regarding the physical condition of the
 earth, _ibid._, pp. 238-272.

 HOBBS. Earthquakes (Appleton, New York, 1907), Chapters xvi and xvii.

Composition of the earth’s core and shell:—

 O. C. FARRINGTON. The Preterrestrial History of Meteorites, Jour.
 Geol., vol. 9, 1901, pp. 623-236.

 E. S. DANA. Minerals and How to Study Them (a book for beginners in
 mineralogy). Wiley, New York, 1895.

On the nature of crystals:—

 VICTOR GOLDSCHMIDT. Ueber das Wesen der Krystalle, Ostwalds Annalen
 der Naturphilosophie, vol. 9, 1909-1910, pp. 120-139, 368-419.



=The processes by which rocks are formed.=—Rocks may be formed in
any one of several ways. When a portion of the molten lithosphere,
so-called _magma_, cools and consolidates, the product is _igneous_
rock. Either igneous or other rock may become disintegrated at the
earth’s surface, and after more or less extended travel, either in the
air, in water, or in ice, be laid down as a sediment. Such sediments,
whether cemented into a coherent mass or not, are described as
_sedimentary_ or _clastic_ rocks. If the fluid from which they were
deposited was the atmosphere, they are known as _subaërial_ or _eolian_
sediments; but if water, they are known as _subaqueous_ deposits. Still
another class are ice-deposited and are known as _glacial_ deposits.


FIG. 16.—Laminated structure of sedimentary rock, Western Kansas
(after a photograph by E. S. Tucker).]

But, as we have learned, rocks may undergo transformations through
mineral alteration, in which case they are known as _metamorphic_
rocks. When these changes consist chiefly in the production of
more soluble minerals at the surface, accompanied by thorough
disintegration, due to the direct attack of the atmosphere, the
resulting rocks are called _residual_ rocks.

=The marks of origin.=—Each of the three great classes of rocks,
the igneous, sedimentary, and metamorphic, is characterized by both
coarser and finer structures, in the examination of which they may be
identified. The igneous rocks having been produced from magmas, which
are essentially homogeneous, are usually without definite directional
structures due to an arrangement of their constituents, and are said
to have a _massive_ structure. Sedimentary rocks, upon the other hand,
have been formed by an assorting process, the larger and heavier
fragments having been laid down when there was high velocity of either
wind or water current, and the smaller and lighter fragments during
intermediate periods. They are therefore more or less banded, and are
said to have a _bedded_ or _laminated_ structure (Fig. 16).

Again, igneous rocks, being due to a process of crystallization, are
composed of mineral individuals which are bounded either by crystal
planes or by irregular surfaces along which neighboring crystals have
interfered with each other; but in either case the grains possess
sharply angular boundaries. Quite different has been the result of
the attrition between grains in the transportation and deposition of
sediments, for it is characteristic of the sedimentary rocks that their
constituent grains are well rounded. Eolian sediments have usually more
perfectly rounded grains than subaqueous deposits.

Glacial deposits, if laid down directly by the ice, are unstratified,
relatively coarse, and contain pebbles which are both faceted
and striated. Such deposits are described as till or tillite. If
glacier-derived material is taken up by the streams of thaw water and
is by them redeposited, the sediments are assorted or stratified, and
they are described as _fluvio-glacial_ deposits.

=The metamorphic rocks.=—Both the coarser structures and the finer
textures of the metamorphic rocks are intermediate between those of
the igneous and the sedimentary classes. A metamorphosed sedimentary
rock, in proportion to its alteration, loses the perfect lamination
and the rounded grain which were its distinguishing characters; while
an igneous rock takes on in the process an imperfect banding, and
the sharp angles of its constituent grains become rounded off by a
sort of peripheral crushing or granulation. Metamorphic rocks are
therefore characterized by an imperfectly banded structure described
as _schistosity_ or _gneiss banding_, and the constituent grains may
be either angular or rounded. If the metamorphism has not been too
intense or too long continued, it is generally possible to determine,
particularly with the aid of the polarizing microscope, whether
the original rock from which it was derived was of igneous or of
sedimentary origin. There are, however, many examples which have defied
a reliable verdict concerning their origin.

=Characteristic textures of the igneous rocks.=—In addition to the
massiveness of their general aspect and the angular boundaries of
their constituents, there are many additional textures which are
characteristic of the igneous rocks. While those that have consolidated
below the earth’s surface, the _intrusive_ rocks, are notably compact,
the magmas which arrive at the surface of the lithosphere before their
consolidation reveal special structures dependent either upon the
expansion of steam and other gases within them, or upon the conditions
of flow over the earth’s surface. Magmas which thus reach the surface
of the earth are described as _lavas_, and the rocks produced by their
consolidation are _extrusive_ or _volcanic_ rocks. The steam included
in the lava expands into bubbles or vesicles which may be large or
small, few or many. According to the number and the size of these
cavities, the rock is said to have a _vesicular_, _scoriaceous_, or
_pumiceous_ texture.

Most lavas, when they arrive at the earth’s surface, contain crystals
which are more or less disseminated throughout the molten mass. The
tourist who visits Mount Vesuvius at the time of a light eruption
may thrust his staff into the stream of lava and extract a portion
of the viscous substance in which are seen beautiful white crystals
of the mineral leucite, each bounded by twenty-four crystal faces.
It is clear that these crystals must have developed by a slow growth
within the magma while it was still below the surface, and when the
inclosing lava has consolidated, these earlier crystals lie scattered
within a _groundmass_ of glassy or minutely crystalline material.
This scattering of crystals belonging to an earlier generation within
a groundmass due to later consolidation is thus an indication of
interruption in the process of crystallization, and the texture which
results is described as _porphyritic_ (Fig. 17 _b_). Should the lava
arrive at the surface before any crystals have been generated and
consolidate rapidly as a rock glass, its texture is described as
_glassy_ (Fig. 17 _c_).

When the crystals of the earlier generation are numerous and
needle-like in form, as is very often the case, they arrange themselves
“end on” during the rock flow, so that when consolidation has occurred,
the rock has a kind of puckered lamination which is the characteristic
of the _fluxion_ or _flow_ texture. This texture has sometimes been
confused with the lamination of the sedimentary rocks, so that wrong
conclusions have been reached regarding origin. At other times the same
needle-like crystals within the lava have grouped themselves radially
to form rounded nodules called spherulites. Such nodules give to the
rock a _spherulitic_ texture, which is nowhere better displayed than
in the beautiful glassy lavas of Obsidian Cliff in the Yellowstone
National Park.


FIG. 17.—Characteristic textures of igneous rocks. _a_, granitic
texture characteristic of the deep-seated intrusive rocks; _b_,
porphyritic texture characteristic of the extrusive and of the
near-surface intrusive rocks; _c_, glassy texture of an extrusive rock.]

Those intrusive rocks which consolidate deep below the earth’s surface,
part with their heat but slowly, and so the process of crystallization
is continued without interruption. Starting from many centers,
the crystals continue to grow until they mutually intersect in an
interlocking complex known as the _granitic_ texture (Fig. 17 _a_).

=Classification of rocks.=—In tabular form rocks may thus be
classified as follows:—

                             {_Intrusive._ Granitic or porphyritic
                             {    texture.
  _Igneous._ Massive and     {_Extrusive._ Glassy or porphyritic
    with sharply angular     {   texture; often also with vesicular,
    grains.                  {   scoriaceous, pumiceous, fluxion,
                             {   or spherulitic textures.

                             {_Subaërial._ Sands and loess.
                             {_Subaqueous._ (See below.)
  _Sedimentary._ Laminate    { _Glacial._ Coarse, unstratified
    and with rounded         {   deposits with faceted pebbles. Till
    grains.                  {   and tillite.
                             {_Fluvio-glacial._ Stratified sands
                             {   and gravels with “worked over”
                             {   glacial characters.

  _Metamorphic._ Schistose   {_Metamorphic proper._ Due to below
    and with grains either   {   surface changes.
    angular or rounded.      {_Residual._ Disintegrated at or near
                             {   surface.

=Subdivisions of the sedimentary rocks.=—While the eolian sediments
are all the product of a purely mechanical process of lifting,
transportation, and deposition of rock particles, this is not always
the case with the subaqueous sediments, since water has the power of
dissolving mineral substance, as it has also of furnishing a home
for animal and vegetable life. Deposited materials which have been
in solution in water are described as _chemical_ deposits, and those
which have played a part in the life process as _organic_ deposits. The
organic deposits from vegetable sources are peat and the coals, while
limestones and marls are the chief depositories of the remains of the
animal life of the water. The tabular classification of the sediments
is as follows:—

_Classification of Sediments._

                    { _Subaqueous_               Conglomerate, sandstone
                    {   Deposited by water.        and shale.
                    { _Subaërial_ or _Eolian_    Sandstone and loess.
  _Mechanical_      {   Deposited by wind.
                    { _Glacial_                  Till and tillite.
                    {   Deposited by ice.
                    { _Fluvio-glacial_           Sands and gravels.
                    {   Glacier-water deposits.

                    { Calcareous tufa            Deposited in springs
                    {                              and rivers.
  _Chemical_        { Oölitic limestone          Deposited at the
                    {                              mouths of rivers
                    {                              between high and
                    {                              low tide.

                    { Formed of plant remains.   Peats and coals.
  _Organic_         { Formed of animal remains.  Limestones and
                    {                            marls.

Winds are under favorable conditions capable of transporting both dust
and sand, but not the larger rock fragments. The dust deposits are
found accumulating outside the borders of deserts as the so-called
_loess_ (Fig. 216), though the sand is never carried beyond the desert
border, near which it collects in wide belts of ridges described as
dunes. When this sand has been cemented into a coherent mass, it is
known as eolian sandstone. A section of the appendix (B) is devoted to
an outline description of some of the commoner rock types.

=The different deposits of ocean, lake, and river.=—Of the
subaqueous sediments, there are three distinct types resulting: (1)
from sedimentation in rivers, the _fluviatile_ deposits; (2) from
sedimentation in lakes, the _lacustrine_ deposits; and (3) from
sedimentation in the ocean, _marine_ deposits. Again, the widest
range of character is displayed by the deposits which are laid down
in the different parts of the course of a stream. Near the source of
a river, coarse river gravels may be found; in the middle course the
finer silts; and in the mouth or delta region, where the deposits
enter the sea or a lake, there is found an assortment of silts and
clays. Except within the delta region, where the area of deposition
begins to broaden, the deposits of rivers are stretched out in long and
relatively narrow zones, and are so distinguished from the far more
important lacustrine and marine deposits.

Lakes and oceans have this in common that both are bodies of standing
as contrasted with flowing water; and both are subject to the
periodical rhythmic motions and alongshore currents due to the waves
raised by the wind. About their margins, the deposits of lake and ocean
are thus in large part wrested by the waves from the neighboring land.
Their distribution is always such that the coarsest materials are laid
down nearest to the shore, and the deposits become ever finer in the
direction of deeper water. Relatively far from shore may be found the
finest sands and muds or calcareous deposits, while near the shore are
sands, and, finally, along the beach, beds of beach pebbles or shingle.
When cemented into coherent rocks, these deposits become shales or
limestones, sandstones, and conglomerates, respectively.

As regards the limestones, their origin is involved in considerable
uncertainty. Some, like the shell limestone or coquina of the Florida
coast, are an aggregation of remains of mollusks which live near
the border of the sea. Other limestones are deposited directly from
carbonate of lime in solution in the water. A deposit of this nature is
forming in southern Florida, both as a flocculent calcareous mud and as
crystals of lime carbonate upon a limestone surface. Again, there is
the reef limestone which is built up of the stony parts of the coral
animal, and, lastly, the calcareous ooze of the deep-sea deposits.

The marine sediments which are derived from the continents, the
so-called _terrigenous_ deposits, are found only upon the continental
shelf and upon the continental slope just outside it. Of these
terrigenous deposits, it is customary to distinguish: (1) _littoral_
or alongshore deposits, which are laid down between high and low tide
levels; (2) _shoal water_ deposits, which are found between low-water
mark and the edge of the continental shelf; and (3) aktian or offshore
deposits, which are found upon the continental slope. The littoral and
shoal water deposits are mainly gravels and sands, while the offshore
deposits are principally muds or lime deposits.

=Special marks of littoral deposits.=—The marks of ripples are often
left in the sand of a beach, and may be preserved in the sandstone
which results from the cementation of such deposits (pl. 11 A). Very
similar markings are, however, quite characteristic of the surface
of wind-blown sand. For the reason that deposits are subject to many
vicissitudes in their subsequent history, so that they sometimes stand
at steep angles or are even overturned, it is important to observe the
curves of sand ripples so as to distinguish the upper from the lower

In the finer sands and muds of sheltered tidal flats may be preserved
the impressions from raindrops or of the feet of animals which have
wandered over the flat during an ebb tide. When the tide is at flood,
new material is laid down upon the surface and the impressions are
filled, but though hardened into rock, these surfaces are those upon
which the rock is easily parted, and so the impressions are preserved.
In the sandstones of the Connecticut valley there has been preserved
a quite remarkable record in the footprints of animals belonging to
extinct species, which at the time these deposits were laid down must
have been abundant upon the neighboring shores.

Between the tides muds may dry out and crack in intersecting lines
like the walls of a honeycomb, and when the cracks have been filled at
high tide, a structure is produced which may later be recognized and
is usually referred to as “mud-crack” structure. This structure is of
special service in distinguishing marine deposits from the subaërial or
continental deposits.

A variation in the direction of winds of successive storms may be
responsible for the piling up of the beach sand in a peculiar “plunge
and flow” or “cross-bedded” structure, a structure which is extremely
common in littoral deposits, though simulated in rocks of eolian origin.

=The order of deposition during a transgression of the sea.=—Many
shore lines of the continents are almost constantly migrating either
landward or seaward. When the shore line advances over the land, the
coast is sinking, and marine deposits will be formed directly above
what was recently the “dry land.” Such an invasion of the land by
the sea, due to a subsidence of the coast, is called a transgression
of the sea, or simply a _transgression_. Though at any moment the
littoral, shoal water, and offshore deposits are each being laid down
in a particular zone, it is evident that each must advance in turn in
the direction of the shore and so be deposited above the zones nearer
shore. Thus there comes to be a definite series of continuous beds,
one above the other, provided only that the process is continued (Fig.
18). At the very bottom of this series there will usually be found a
thin bed of pebbly beach materials, which later will harden into the
so-called _basal conglomerate_. If the size of the pebbles is such as
to make possible an identification, it will generally be found that
these represent the ruins of the rock over which the sea has advanced
upon the land.

[Illustration: FIG. 18.—Diagram to show the order of the sediments
laid down during a transgression of the sea.]

Next in order above the basal conglomerate, will follow the coarser and
then the finer sands, upon which in turn will be laid down the offshore
sediments—the muds and the lime deposits. Later, when cemented
together, these become in order, coarser and finer sandstones, shales,
and limestones. The order of superposition, reading from the bottom to
the top, thus gives the order of decreasing age of the formations.

A subsequent uplift of the coast will be accompanied by a recession
of the sea, and when later dissected by nature for our inspection,
the order of superposition and the individual character of each of
the deposits may be studied at leisure. From such studies it has been
found that along with the inorganic deposits there are often found the
remains of life in the hard parts of such invertebrate animals as the
mollusks and the crustacea. These so-called _fossils_ represent animals
which were gradually developed from simpler to more and more complex
forms; and they thus serve the purpose of successive page numbers in
arranging the order of disturbed strata, at the same time that they
supply the most secure foundation upon which rests the great doctrine
of evolution.

=The basins of earlier ages.=—It was the great Viennese geologist,
Professor Suess, who first pointed out that in mountain regions there
are found the thickest and the most complete series of the marine
deposits; whereas outside these provinces the formations are separated
by wide gaps representing periods when no deposits were laid down
because the sea had retired from the region. The completeness of the
series of deposits in the mountain districts can only be interpreted to
mean that where these but lately formed mountains rise to-day, were for
long preceding ages the basins for deposition of terrigenous sediments.
It would seem that the lithosphere in its adjustment had selected these
earlier sea basins with their heavy layers of sediment for zones of
special uplift.

=The deposits of the deep sea.=—Outside the continental slope, whose
base marks the limit of the terrigenous deposits, lies the deeper sea,
for the most part a series of broad plains, but varied by more profound
steep-walled basins, the so-called “deeps” of the ocean. As shown by
the dredgings of the _Challenger_ expedition and others of more recent
date, the deposits upon the ocean floor are of a wholly different
character from those which are derived from the continents. Except in
the great deeps, or between depths of five hundred and fifteen hundred
fathoms, these deposits are the so-called “ooze”, composed of the
calcareous or chitinous parts of algæ and of minute animal organisms.
The pelagic or surface waters of the ocean are, as it were, a great
meadow of these plant forms, upon which the minute crustacea, such
as globigerina, foraminifera, and the pteropods, feed in countless
myriads. The hard parts of both plant and animal organisms descend to
the bottom and there form the ooze in which are sometimes found the ear
bones of whales and the teeth of sharks.

In the deeps of the ocean, none of these vegetable or animal deposits
are being laid down, but only the so-called “red clay”, which is
believed to represent decomposed volcanic material deposited by the
winds as fine dust on the surface of the ocean, or the product of
submarine volcanic eruption. From the absence of the ooze in these
profound depths, the conclusion is forced upon us that the hard parts
of the minute organisms are dissolved while falling through three or
four miles of the ocean water.


 J. S. DILLER. The Educational Series of Rock Specimens collected and
 distributed by the United States Geological Survey, Bull. 150 U. S.
 Geol. Surv., 1898, pp. 1-400.

 L. V. PIRSSON. Rocks and Rock Minerals. Wiley, New York, 1908.

 SIR JOHN MURRAY. Deep-sea Deposits, Reports of the _Challenger_
 expedition, Chapter iii.

 L. W. COLLET. Les dépôts marins. Doin, Paris, 1907 (Encyclopédie



=The zones of fracture and flow.=—It is easy to think of the
atmosphere and the hydrosphere as each sustaining at any point the load
of the superincumbent material. At the sea level the weight of air
upon each square inch of surface is about fifteen pounds, whereas upon
the floor of the hydrosphere in the more profound deeps the load upon
the square inch must be measured in tons. Near the lithosphere surface
the rocks support by their strength the load of rock above them, but
at greater depths they are unable to do this, for the load bears upon
each portion of the rock with a pressure equivalent to the weight
of a rock column which extends upward to the surface. The average
specific gravity of rock is 2.7, and it is thus easy to calculate the
length of the inch square column which has a weight equivalent to the
crushing strength of any given rock. At the depth represented by the
length of such a column, rocks cannot yield to pressure by fracture,
for the opening of a crack implies that the rock upon either side is
strong enough to prevent the walls from closing. At this depth, rock
must therefore yield to pressure not by fracture, as it would at the
surface, but by flow after the manner of a liquid; and so the zone
below this critical level is referred to as the _zone of flow_.


FIG. 19.—Two intersecting parallel series of fractures produced upon
each free surface of a prismatic block of stiff molders’ wax when
broken by compression from the ends (after Daubrée and Tresca).]

In contrast, the near-surface zone is called the _zone of fracture_.
But different rocks possess different strengths, and these are subject
to modifications from other conditions, such, for example, as the
proximity of an uncooled magma. The zone of flow is therefore joined
to the zone of fracture, not upon a definite surface, but in an
intermediate zone described as the _zone of fracture and flow_.

=Experiments which illustrate the fracture and flow of solid
bodies.=—A prismatic block prepared from stiff molders’ wax, if
crushed between the jaws of a testing machine, yields a system of
intersecting fractures which are perpendicular to the free surfaces
of the block and take two directions each inclined by half of a right
angle to the direction of compression (Fig. 19). This experiment
may illustrate the manner in which fractures are produced by the
compression within the zone of fracture of the lithosphere, as its core
continues to contract.

To reproduce the conditions within the zone of flow, it will be
necessary to load the lateral surfaces of the block instead of leaving
them unconstrained as in the above-described experiment. The experiment
is best devised as in Fig. 20. Here a series of layers having varying
degrees of rigidity is prepared from beeswax as a base, either
stiffened by admixture of varying proportions of plaster of Paris, or
weakened by the use of Venice turpentine. Such a series of layers may
represent rocks of as widely different characters as limestone and
shale. The load which is to represent superincumbent rock is supplied
in the experiment by a deep layer of shot.

[Illustration: FIG. 20.—Apparatus to illustrate the folding of strata
within the zone of flow (after Willis).]

When compression is applied to the layers from the ends, these normally
solid materials, instead of fracturing, are bent into a series of
folds. The stiffer, or more competent, layers are found to be less
contorted than are the weaker layers, particularly if the latter have
been protected under an arch of the more competent layer (pl. 2 A).

=The arches and troughs of the folded strata.=—Every series of folds
is made up of alternating arches and troughs. The arches of the strata
the geologist calls _anticlines_ or _anticlinal folds_, and the troughs
he calls _synclines_ or _synclinal folds_ (Fig. 21). When a stratum is
merely dropped in a bend to a lower level without producing a complete
arch or a complete trough, this half fold is termed a _monocline_.


FIG. 21.—Diagrams representing _a_, an anticline; _b_, a syncline; and
_c_, a monocline.]

Any flexuring of the strata implies a reduction of their surface area,
or, considering a single section, a shortening. If the arches and
troughs are low and broad, the deformation of the strata is slight,
the shortening is comparatively small, and the folds are described as
_open_ (Fig. 22 _b_). If they be relatively both high and narrow, the
deformation is considerable, a larger amount of crustal shortening has
gone on, and the folds are described as _close_ (Fig. 22 _c_). This
closing up of the folds may continue until their sides have practically
the same slope, in which case they are said to be _isoclinal_ (Fig. 22


FIG. 22.—A comparison of folds to express increasing degrees of
crustal shortening or progressive deformation within the zone of flow:
_a_, stratum before folding; _b_, open folds; _c_, close folds; _d_,
isoclinal folds.]

=The elements of folds.=—Folds must always be thought of as having
extension in each of the three dimensions of space (Fig. 23), and not
as properly included within a single plane like the cross sections
which we so often use in illustration. A fold may be conceived of as
divided into equal parts by a plane which passes along the middle of
either the arch or the trough, and is called the _axial plane_. The
line in which this plane intersects the arch or the trough is the
_axis_, which may be called the _crestline_ in an anticline, and the
_troughline_ in a syncline.

In the case of many open folds the axis is practically horizontal, but
in more complexly folded regions this is seldom true. The departure of
the axis from the horizontal is called the _pitch_, and folds of this
type are described as _pitching folds_ or _plunging folds_. The axis
is in reality in these cases thrown into a series of undulations or
“longitudinal folds”, and hence pitch will vary along the axis.

[Illustration: FIG. 23.—Anticlinal and synclinal folds in strata
(after Willis).]


FIG. 24.—Diagrams to illustrate the different shapes of rock folds.]

=The shapes of rock folds.=—By the axial plane each fold is divided
into two parts which are called its _limbs_, which may have either the
same or different average inclinations. To describe now the shapes
of rock folds and not the degree of compression of the district,
some additional terms are necessary. Anticlines or synclines whose
limbs have about the same inclinations are known as _upright_ or
_symmetrical folds_. The axial plane of the symmetrical fold is
vertical (Fig. 24). If this plane is inclined to the vertical, the
folds are _unsymmetrical_. So soon as the steeper of the two limbs has
passed the vertical position and inclines in the same direction as the
flatter limb, the fold is said to be _overturned_. The departure from
symmetry may go so far that the axial plane of the fold lies at a very
flat angle, and the fold is then said to be _recumbent_. The observant
traveler by train along any of the routes which enter the Alps may from
his car window find illustrations of most of these types of rock folds,
as he may also, though generally less easily, in passing through the
Appalachian Mountains.

[Illustration: FIG. 25.—Secondary and tertiary flexures superimposed
upon the primary ones.]

In regions which have been closely folded the larger flexures of
the strata may be found with folds of a smaller order of magnitude
superimposed upon them, and these in turn may show crumplings of still
lower orders. It has been found that the folds of the smaller orders
of magnitude possess the shapes of the larger flexures, and much is
therefore to be learned from their careful study (Fig. 25). It is also
quite generally discovered that parallel planes of ready parting,
which are described as _rock cleavage_, take their course parallel to
the axial plane within each minor fold. As was long ago shown by the
pioneer British geologists, these planes of cleavage are essentially
parallel and follow the fold axes throughout large areas.

│ PLATE 2.                                                               │
│                                                                        │
│ [Illustration: _A._ Layers compressed in experiments and showing the   │
│ effect of a competent layer in the process of folding (after Willis).] │
│                                                                        │
│ [Illustration: _B._ Experimental production of a series of parallel    │
│ thrusts within closely folded strata (after Willis).]                  │
│                                                                        │
│ [Illustration: _C._ Apparatus to illustrate shearing action within the │
│ overturned limb of a fold.]                                            │

=The overthrust fold.=—Whenever a stratum is bent, there is a tendency
for its particles to be separated upon the convex side of the bend,
at the same time that those upon the concave side are crowded closer
together—there is tension in the former case and compression in the
latter (Fig. 26). Within an unsymmetrical or an overturned fold, the
peculiar distortions in the different sections of the stratum are less
simple and are best illustrated by pl. 2 C. This apparatus shows two
similar piles of paper sheets, upon the edges of each of which a series
of circles has been drawn. When now one of the piles is bent into an
unsymmetrical fold, it is seen that through an accommodation by the
paper sheets sliding each over its neighbor large distortions of the
circles have occurred. In that steeper limb which with closer folding
will be overturned the circles have been drawn out into long and narrow
ellipses, and this indicates that those rock particles which before the
bending were included in the circle have been moved past each other in
the manner of the blades of a pair of shears. Such extreme “shearing”
action is thus localized in the underturned limb of the fold, and a
time must come with continuation of the compression when the fold
will rupture at this critical place along a plane parallel to the
longest axis of the ellipses or nearly parallel to the axial plane of
the anticline. Such structures probably occur in the zone of combined
fracture and flow, up into which the beds are forced in cases of close
compression. Relief thus being found upon this plane of fracture,
the upper portion of the fold will now ride over the lower, and the
displacement is described as a _thrust_ or _overthrust_.


FIG. 26.—A bent stratum to illustrate tension upon the convex and
compression upon the concave side (after Van Hise).]

In the long series of experiments conducted by Mr. Bailey Willis of the
United States Geological Survey, all the stages between the overturned
fold and the overthrust fold were reproduced. Where a series of folds
was closely compressed, a parallel series of thrusts developed (pl. 2
B), so that a series of slices cutting across neighboring strata was
slid in succession, each over the other, like the scales upon a fish
or the shingles upon a roof. Quite remarkable structures of this kind
have been discovered in rocks of such closely folded districts as the
Northwest Highlands of Scotland, where the overriding is measured in
miles. Near the thrust planes the rocks show a crushing of the grains,
and the planes themselves are sometimes corrugated and polished by the

=Restoration of mutilated folds.=—Since flexuring of the rocks takes
place within the zone of flow at a distance of several miles below the
earth’s surface, it is quite obvious that the results of the process
can be studied only after some thousands of feet of superincumbent
strata have been removed. We are a little later to see by what
processes this lowering of the surface is accomplished, but for the
present it may be sufficient to accept the fact, realizing that before
foldings in the strata can reach the surface, they must have passed
through the upper zone of fracture.

It might perhaps be supposed that the anticlines would appear as
the mountains upon the surface, and occasionally this is true; as,
for example, in the folded Jura Mountains of western Europe. More
generally, the mountains have a synclinal structure and the valleys an
anticlinal one; but as no general rule can be applied, it is necessary
to make a restoration of the truncated folds in each district before
their character can be known.

=The geological map and section.=—The earth’s surface is in most
regions in large part covered with soil or with other incoherent rock
material, so that over considerable areas the hard rocks are hidden
from view. Each locality at which the rock is found at the earth’s
surface “in place” is described as an _outcropping_ or _exposure_. In
a study of the region each such exposure must be examined to determine
the nature of the rock, especially for the purpose of correlation
with neighboring exposures, and, in addition, both the probable
direction in which it is continued along the surface—the _strike_—and
the inclination of its beds—the _dip_. If the outcroppings are
sufficiently numerous, and rock type, strike and dip, may all be
determined, the folds of the district may be restored with almost as
much accuracy as though their curves were everywhere exposed to view.
A cross section through the surface which represents the observed
outcrops with their inclinations and the assumed intermediate strata
in their probable attitudes is described as a _geological section_
(Fig. 27). A map upon which the data have been entered in their correct
locations, either with or without assumptions concerning the covered
areas, is known as a _geological map_.

[Illustration: FIG. 27.—A geological section based upon observations
at outcrops, but with the truncated arches restored.]

If the axes of folds are absolutely horizontal, and the surface of
the earth be represented as a plain, the lines of intersection of the
truncated strata with the ground, or with any horizontal surface, will
give the directions of continuation of the individual strata. This
strike direction is usually determined at each exposure by use of a
compass provided with a spirit level. When that edge of the leveled
compass which is parallel to the north-south line upon the dial is held
against the sloping rock stratum, the angle of strike is measured in
degrees by the compass needle. If the cardinal directions have been
placed in their correct positions upon the compass dial, the needle
will point to the northwest when the strike is northeast, and _vice
versa_ (Fig. 28 _a_). Upon the geologist’s compass it is therefore
customary to reverse the initials which represent the east and west
directions, in order that the correct strike may be read directly from
the dial (Fig. 28 _b_).


FIG. 28.—Diagram to illustrate the manner of determining the strike
of rock beds at an outcropping. _a_, a compass which has the cardinal
directions in their natural positions; _b_, a compass with the east
and west initials reversed upon the dial; _c_, home-made clinometer in
position to determine the dip.]

By the dip is meant the inclination of the stratum at any exposure, and
this must obviously be measured in a vertical plane along the steepest
line in the bedding plane. The dip angle is always referred to a
horizontal plane, and hence vertical beds have a dip of 90°. The device
for measuring this angle of dip, the _clinometer_, is merely a simple
pendulum which serves as an indicator and is centered at the corner of
a graduated quadrant. A home-made variety is easily constructed from a
square piece of board and an attached paper quadrant (Fig. 28 _c_), but
the geologist’s compass is always provided with a clinometer attachment
to the dial.


FIG. 29.—Diagram to show the use of T symbols to indicate the dip and
strike of outcroppings.]

Since the strike is the intersection of the bedding plane with
a horizontal surface, and the dip is the intersection with that
particular vertical plane which gives the steepest inclination, the
strike and dip are perpendicular to each other. To represent them upon
maps, it is more or less customary to use the so-called T symbols, the
top of the T giving the direction of the strike and the shank that of
the dip. If meridians are drawn upon the map, the direction or attitude
of the T can be found by the use of a simple protractor; and when
entered upon the map, the exact angle of the strike may be supplied by
a figure near the top of the T, and the dip angle by a figure at the
end of the shank. It is the custom, also, to make the length of the
shank inversely proportional to the steepness of the dip, so that in
a broad way the attitudes of the strata may be taken in at a glance
(Fig. 29). It is further of advantage to make the top of the T a double
line, so that some symbol or color may show the correlations of the
different exposures. To illustrate, in Fig. 29, the symbol marked _a_
represents an outcrop of limestone, the strike of which is 50° east
of north (N. 50° E.), and the dip of which is 45° southeast. In the
same figure _b_ represents a shale outcrop in horizontal beds, which
have in consequence a universal strike and a dip of 0°. An exposure of
limestone in vertical beds which strike N. 60° E. is shown at _c_, etc.


FIG. 30.—Diagram to show how the thickness of a formation may be
obtained from the angle of the dip and the width of the exposures.]

=Measurement of the thickness of formations.=—When formations still
lie in horizontal beds, we may sometimes learn their thickness
directly either from the depth of borings to the underlying rock, or
by measurements upon steep cañon walls. If the beds stand vertically,
the matter is exceedingly simple, for in this case the thickness is the
width of the outcrops of the formation between the beds which bound it
upon either side. In the general case, in which the beds are neither
horizontal nor vertical, the thickness must be obtained indirectly from
the width of the exposures and the angle of the dip. The factor by
which the exposure width must be multiplied is known as the sine of the
dip angle (Fig. 30), which is given with sufficient accuracy for most
purposes in the following table. It is obvious that in order to obtain
the full thickness of a formation it is necessary to measure from the
contact with the adjacent formation upon the one side to a similar
contact with the nearest formation upon the other.

_Natural Sines_

   0°   .00          35°   .57          70°   .94
   5°   .09          40°   .64          75°   .97
  10°   .17          45°   .71          80°   .98
  15°   .26          50°   .77          85°  1.00
  20°   .34          55°   .82          90°  1.00
  25°   .42          60°   .87
  30°   .50          65°   .91

[Illustration: FIG. 31.—Combined surface and sectional views of a
plunging anticline (after Willis).]

[Illustration: FIG. 32.—Combined surface and sectional views of a
plunging syncline (after Willis).]

=The detection of plunging folds.=—When the axis of a fold is
horizontal, its outcrops upon a plain will continue to have the same
strike until the formation comes to an end. Upon a generally level
surface, therefore, any regular progressive variation in the strike
direction is an indication that the folds have a plunging or pitching
character. Many serious mistakes of interpretation have been made
because of a failure to recognize this evidence of plunging folds. The
way in which the strikes are progressively modified will be made clear
by the diagrams of Figs. 31 and 32, the first representing a pitching
anticline and the second a pitching syncline. In both these reciprocal
cases the strikes of the beds undergo the same changes, and the dip
directions serve to distinguish which of the two structures is present
in a given case. There is, however, one further difference in that the
hard layers of the plunging anticline, where they disappear below
the surface in the axis, will present a domed surface sloping forward
like the back of a whale as it rises above the surface of the sea.
Plunging folds in series will thus appear in the topography as a series
of sharply zigzagging ranges at those localities where the harder
layers intersect the surface. Such features are encountered in eastern
Pennsylvania, where the hard formations of the Appalachian Mountain
system plunge northeastward under the later formations. The pitch of
the larger fold is often disclosed by that of the minor puckerings
superimposed upon it.


FIG. 33.—Unconformity between a lower and an upper series of beds upon
the coast of California. Note how the hard layer stands in relief upon
the connecting surface (after Fairbanks).]

=The meaning of an unconformity.=—The rock beds, which are deposited
one above the other during a transgression of the sea, are usually
parallel and thus represent a continuous process of deposition. Such
beds are said to be _conformable_. Where, upon the other hand, two
series of deposits which are not parallel to each other are separated
by a break, they are said to form _unconformable_ series, and the break
or surface of junction is an _unconformity_ (Fig. 33).

Here it is evident that the sediments which compose the lower series
of beds have been folded in the zone of flow, though the upper series
has evidently escaped this vicissitude. Furthermore, the surface which
delimits the lower series from the upper is somewhat irregular and
shows a hard layer standing in relief, as it would if it had opposed
greater resistance to the attacks of the atmosphere upon it.


FIG. 34.—Series of diagrams to illustrate in succession the episodes
involved in the historical development of an angular unconformity. The
vertical arrows indicate direction of movement of the land, and the
horizontal arrows the direction of shore migration.]

In reality, an unconformity between formations must be interpreted to
mean that the lower series is not only older than the upper, as shown
by the order of superposition, but that the time of its deposition was
separated from that of the upper by a hiatus in which important changes
took place in the lower series. The stages or episodes in the history
of the beds represented in Fig. 33 may be read as follows (see Fig. 34

(_a_) Deposition of the lower series during a transgression of the sea.

(_b_) Continued subsidence and burial of the lower series beneath
overlying sediments, and flexuring in the zone of flow.

(_c_) Elevation of the combined deposits to and far above sea level and
removal by erosion of vast thicknesses of the upper sediments.

(_d_) A new subsidence of the truncated lower series and deposition of
the upper series across its eroded surface.

(_e_) A new elevation of the double series to its present position
above sea level.


FIG. 35.—Types of deceptive or erosional unconformities.]

From this succession of episodes it is seen that a break of this
kind between two series of deposits involves a double oscillation of
subsidence followed by elevation—a large depression followed by a
large elevation, a smaller subsidence followed by elevation. The time
interval which must have been represented by these repeated operations
is so vast as at first to stagger the mind in contemplating it. When,
as in this instance, the dips of the lower series of beds differ from
those of the upper, we have to do with an _angular unconformity_.
It may be, however, that the lower series was not so far depressed
as to enter the zone of flow, and that its beds meet those of the
upper series with apparent conformity. Such an unconformity is often
extremely difficult to recognize, and it is described as a _deceptive_
or _erosional unconformity_.

With a deceptive unconformity the clew to its real nature is usually
some fact which indicates that the lower series of sediments had been
raised above the level of the sea before the upper series was deposited
upon it. This may be apparent either in the irregularity of the surface
on which the two series are joined, in some evidence of the action of
waves such as would be furnished by a basal conglomerate in the upper
series, or some indication of different resistance of different rocks
of the lower series to attacks of the atmosphere upon them (Figs. 33
and 35 _a-c_).

In most cases, at least, the lowest member of the upper series will
be a different type of rock from the uppermost member of the lower
series, hence the frequent occurrence of the discordant cross bedding
in sandstone should not deceive even the novice into the assumption of
an unconformity.


 The zones of fracture and flow:—

 C. R. VAN HISE. Principles of North American Precambrian Geology, 16th
 Ann. Rept. U.S. Geol. Surv., 1895, Pt. I, pp. 581-603.

 BAILEY WILLIS. Mechanics of Appalachian Structure, 13th Ann. Rept.
 U.S. Geol. Surv., 1893, Pt. II, pp. 217-253.

 A. DAUBRÉE. Études Synthétiques de Géologie Expérimentale. Paris,
 1879, pp. 306-328, pl. II.

 W. PRINZ. Quelques remarques générales à propos de l’essai de carte
 tectonique de la belgique, etc., Bull. Soc. Belge Geol., vol. 18,
 1904, p. 143, pl. V.

Analysis of folds:—

 VAN HISE and WILLIS as above; DE MARGERIE et HEIM; Les dislocations de
 l’écorce terrestre (in French and German languages). Zurich, 1888.

Geological maps:—

 WM. H. HOBBS. The Mapping of the Crystalline Schists, Jour. Geol.,
 vol. 10, 1902, pp. 780-792, 858-890.




FIG. 36.—A set of master joints developed in shale upon the shores of
Cayuga Lake near Ithaca, New York (after U. S. G. S.).]


FIG. 37.—Diagram to show how sets of master joints differing in
direction by half a right angle may abruptly replace each other.]


FIG. 38.—Diagram to show the different combinations of the series
composing two double sets of master joints, and in _a_, _a_, _a_
additional disorderly fractures.]

=The system of the fractures.=—In referring to experiments made upon
the fracture of solid blocks under compression (p. 41), it was shown
that two series of parallel fractures develop perpendicular to each
free surface of the block, and that these series are each of them
inclined by half of a right angle to the direction of compression, and
thus perpendicular to each other. The fragments into which a block
with one free surface would thus tend to be divided should be square
prisms perpendicular to the free surface. It would be interesting, if
it were practicable, to learn from experiment how these prisms would be
further fractured by a continuation of the compression. From mechanical
considerations involving the resolution of forces with reference to
the ready-formed fractures, it seems probable that the next series of
fractures to form would bisect the angles of the first double series
or set. Wherever rocks are found exposed in their original attitudes,
they are, in fact, seen to be intersected by two parallel series of
fractures which are perpendicular to the earth’s surface and to each
other and are described as _joints_. In many cases more than two series
of such fractures are found, yet even in these cases two more perfectly
developed series are prominent and almost exactly perpendicular to each
other as well as to the earth’s surface. This omnipresent double series
or _set_ of joints is the well-known set of _master joints_, and very
often it is found developed practically alone (Fig. 36). Over large
areas, the direction of the set of master joints may remain practically
constant, or this set may quite suddenly give place to a similar set
which is, however, turned through half a right angle from the first
(Fig. 37). Not infrequently two such sets of master joints are found
together bisecting each other’s angles within the same rocks, and to
them are sometimes added additional though less perfect series of
joint planes.

Studied throughout a considerable district, the various series which
make up these two sets of master joints may be seen locally developed
in different combinations as well as in association with additional
fissure planes which are not easily reduced to any simple law of
arrangement (Fig. 38 _a_, _a_, _a_). Only rarely are regular joint
series observed which do not stand perpendicular to the original
attitude of the rock beds. In a few localities, however, rectangular
joint sets have been discovered which divide the rock into prisms
parallel to the earth’s surface and with the joint series inclined
to it each by half a right angle. Where the rock beds have been much
disturbed, the complex of joints may be such as to defy all attempts
at orderly arrangement.

[Illustration: FIG. 39.—View on the shore at Holstensborg, West
Greenland, to show the subequal spacing of the joints (after Kornerup).]


FIG. 40.—View of an exposed hillside in Iceland upon which the snow
collected in crannies along the joints brings out to advantage both
the larger and the smaller intervals of the joint system (after

=The space intervals of joints.=—The same kind of subequal spacing
which characterizes the fractures near the surface of the block in
Daubrée’s experiment (Fig. 19, p. 41) is found simulated by the rock
joints (Fig. 39). Such unit intervals between fractures may be grouped
together into larger units which are separated by fractures of unusual
perfection. We may think of such larger space units as having the
smaller ones superimposed upon them (Fig. 40).

=The displacements upon joints—faults.=—In the vast majority of
cases, the joint fractures when carefully examined betray no evidence
of any appreciable movement of the two walls upon each other. Generally
the rock layers are seen to cross the joints without apparent
displacement. Joints are therefore planes of disjunction only, and not
planes of displacement.


FIG. 41.—Faulted blocks of basalt divided by joints near Woodbury,
Connecticut. To show the structure of the rock, some of the foliage has
been removed in preparing the sketch from a photograph.]

Within many districts, however, a displacement may be seen to have
occurred upon certain of the joint planes, and these are then described
as _faults_. Such displacements of necessity imply a differential
movement of sections or blocks of the earth’s crust, the so-called
_orographic blocks_, which are bounded by the joint planes and play
individual rôles in the movement. A simple case of such displacements
in rocks intersected by a single set of master joints is represented
in the model of plate 4 C. The most prominent fault represented by
this model runs lengthwise through the middle, and the displacement
which is measured upon it not only varies between wide limits, but is
marked by abrupt changes at the margins of the larger blocks. This
vertical displacement upon the fault is called its _throw_. Though not
illustrated by the model, horizontal displacements may likewise occur,
and these will be more fully discussed when the subject of earthquakes
is considered in the following chapter. An actual example of blocks
displaced by vertical adjustment is represented in Fig. 41, a simple
type of faulting which has taken place in rocks but slightly disturbed
from their original attitude, but intersected by a relatively simple
system of master joints. In those regions where the beds have been
folded and perhaps overthrust before their elevation into the zone
of fracture, and which are further intersected by disorderly fissure
planes, the results are far more complex. In such cases the planes of
individual displacement may not be vertical, though they are generally
steeper than 45°. For their description it is necessary to make use
of additional technical terms (Fig. 42). The inclination of a sloping
fault plane measured against the vertical is called the _hade_ of
the fault. The _total displacement_ is measured along the plane of
the fault from a point upon one limb to the point from which it was
separated in the other. The additional terms are made sufficiently
clear by the diagram.


FIG. 42.—A fault in previously disturbed strata. _AB_, displacement;
_AC_, throw; _BD_, stratigraphic throw; _BC_, heave; angle _CAB_, hade.]

=Methods of detecting faults.=—The first effect of a fault is
usually to produce a crack at the surface of the earth; and, provided
there is a vertical displacement or throw, an escarpment which rises
upon the upthrown side of the fault. In general it may be said that
escarpments which appear at the earth’s surface as plane surfaces
probably represent planes of fracture, though not necessarily planes
of faulting. In many cases the actual displacements lie buried under
loose rock débris near to and paralleling the escarpment, and in some
cases as a result of the erosional processes working upon alternately
hard and soft layers of rock, the escarpment may later appear upon the
downthrown side or limb of the fault (Fig. 43). As an illustration of a
fault escarpment, the façade of El Capitan and many other rock faces of
the Yosemite valley may be instanced.


FIG. 43.—Diagrams to show how an escarpment originally on the upthrown
side of the fault may, through erosion, appear upon the downthrown


FIG. 44.—A fault plane exhibiting “drag.” The opening is artificial
(after Scott).]

When we have further studied the erosional processes at the earth’s
surface, it will be appreciated that faults tend to quickly bury
themselves from sight, whereas fold structures will long remain in
evidence. Many faults will thus be overlooked, and too great weight
is likely to be ascribed to the folds in accounting for the existing
attitudes and positions of the rock masses. Faults must therefore be
sought out if mistakes of interpretation are to be avoided.

The most satisfactory evidence of a fault is the discovery of a
rock bed which may be easily identified, and which is actually seen
displaced on a plane of fracture which intersects it (Fig. 42, p. 59).
When such an easily recognizable layer is not to be found, the plane
of displacement may perhaps be discovered as a narrow zone composed of
angular fragments of the rock cemented together by minerals which form
out of solution in water. Such a fractured rock zone which follows a
plane of faulting is a _fault breccia_. If the fault breccia, or vein
rock, is much stronger than the rock on either side, it may eventually
stand in relief at the surface like a dike or wall. At other times the
displacement produces little fracture of the walls, but they slide over
each other in such a manner as to yield either a smoothly corrugated or
an evenly polished surface which is described as “slickensides.” It may
be, however, that during the movement either one or both of the walls
have “dragged”, and so are curled back in the immediate neighborhood of
the fault plane (Fig. 44).

When, as is quite generally the case, the actual plane of displacement
of a fault is not open to inspection, the movement may be proven by
the observation of abrupt, as contrasted with gradual, changes in the
strikes and dips of neighboring exposures (Fig. 45); or by noting
that some easily recognized formation has been sharply offset in its
outcrops (Fig. 46).


FIG. 45.—Map to show how a fault may be indicated in abrupt changes of
the strike and dip of neighboring exposures.]


FIG. 46.—A series of parallel faults indicated by successive offsets
in the course of an easily recognizable rock formation.]

There are in addition many indications rather than proofs of the
presence of faults, which must be taken account of in every general
study of the geology of a district. Thus the outcrops of all
neighboring formations may terminate abruptly upon a straight line
which intersects all alike. Deep-seated fissure springs may be aligned
in a striking manner, and so indicate the course of a prominent
fracture, though not necessarily of a fault. Much the same may be said
of the dikes of cooled magma which have been injected along preëxisting

=The base of the geological map.=—Modern topographic maps form an
important part of the library of the serious student of physiography;
they are the gazetteer of this branch of science. Every civilized
nation has to-day either completed a topographic atlas of its
territory, or it is vigorously prosecuting a survey to furnish maps
which represent the relief with some detail, and publishing the results
in the form of an atlas of quadrangles. Thus a relief map will erelong
be obtainable of any part of the civilized world, and may be purchased
in separate sections. Nowhere is this work being taken up with greater
vigor than in the United States, where a vast domain representing
every type of topographic peculiarity is being attacked from many
centers. Here and elsewhere the relief of the land is being expressed
by so-called contours or lines of equal altitude upon the earth’s
surface. It is as though a series of horizontal planes, separated by
uniform intervals of 20 or 40 or 100 feet, had been made to intersect
the surface, and the intersection curves, after consecutive numeration,
had been dropped into a single plane for printing.

Where the slopes are steep, the contour lines in the topographic map
will appear crowded together and so produce a deep shade upon the map;
whereas with relatively flat surfaces white patches will stand out
prominently upon the map. More and more the topographic map is coming
into use, and for the student of nature in particular it is important
to acquire facility in interpreting the relief from the topographic
map. To further this end, a special model has been devised, and its use
is described in appendix C. Usually before any satisfactory geological
map can be prepared, a contoured topographic map of the district to be
studied must be available.

=The field map and the areal geological map.=—As the atlas of
topographic maps is the physiographic gazetteer, so geological maps
together constitute the reference dictionary of descriptive geology.
Not only are topographic maps of many districts now generally
available, but more and more it has become the policy of governments to
supply geological maps in the same quadrangle form which is the unit
of the topographic map. The geological map is, however, a complex of
so many conventional symbols, that without some practical experience
in the actual preparation of one, it is exceedingly difficult for
the student to comprehend its significance. A modern geological map
is usually a rectangular sheet printed in color, upon which are many
irregular areas of individual hue joined to each other like the parts
of a child’s picture puzzle.

The colored areas upon the geological map are each supposed to indicate
where a certain rock type or formation lies immediately below the
surface, and this distribution represents the best judgment of the
geologist who, after a study of the district, has prepared the map.
Unfortunately the conventions in use are such that his observation and
his theory have been hopelessly intermingled in the finished product.
Armed with the geological map, the student who visits the district
finds spread out before him, it may be, a landscape of hill and valley,
of green forest and brown farming land, which is as different as may be
from the colored puzzle which he holds in his hand. Hidden under the
farm vegetation or masked by the woods are scattered outcroppings of
rock which have been the basis of the geologist’s judgment in preparing
the map. Experience shows that in order to bridge the wide gap between
the geology in the landscape and the patches of color upon the map
something more than mere examination of the colored sheet is necessary.
We shall therefore describe, with the aid of laboratory models, the
various stages necessary to the preparation of a geological map, and
every student should be advised to follow this by practical study of
some small area where rocks are found in outcrop.

Though the published _areal geological map_ represents both fact and
theory, the map maker retains an unpublished _field map_ or map of
observations, upon which the final map has been based. This field map
shows the location of each outcrop that has been studied, with a record
of the kind of rock and of such observations as strike, dip, and pitch.
Our task will therefore be to prepare: (1) a field map; (2) an areal
geological map; and (3) some typical geological sections.

=Laboratory models for the study of geological maps.=—In order to
represent in the laboratory the disposition of rock outcrops in the
field, special laboratory tables are prepared with removable covers
and with fixed tops, which are divided into squares numbered like the
township sections of the national domain (Fig. 47). To represent the
rock outcrops, blocks are prepared which may be fixed in any desired
position by fitting a pin into a small augur hole bored through the
table. The outcrop blocks for the sedimentary rock types are so
constructed as to show the strike and dip of the beds. (See Appendix D.)

[Illustration: FIG. 47.—Field map prepared from a laboratory table.]

=The method of preparing the map.=—To prepare the map, use is made of
a geological compass with clinometer attachment, a protractor, and a
map base divided into sections like the top of the table, and on the
scale of one inch to the foot. Each exposure represented upon the table
is “visited” and then located upon the base map in its proper position
and attitude. The result is the field map (Fig. 47), which thus
represents the facts only, unless there have been uncertainties in the
correlation of exposures or in determining the position of the bedding

[Illustration: FIG. 48.—Areal geological map constructed from the
field map of Fig. 47, with two selected geological sections.]

To prepare the areal geological map from the field map, it is first
necessary to fix the _boundaries_ which separate formations at the
surface; and now perhaps for the first time it is realized how large
an element of uncertainty may enter if the exposures were widely
separated. It is clear that no two persons will draw these lines in
the same positions throughout, though certain portions of them—where
the facts are more nearly adequate—may correspond. In Fig. 48 is
represented the areal geological map constructed from the field map,
with the doubtful area at one side left blank.

Some conclusions from this map may now be profitably considered. The
complexly folded sandstone formation at the left of the map appears
as the oldest member represented, since its area has been cut through
by the intrusive granite which does not intrude other formations,
and is unconformably overlaid by the limestone and its basal layer
of conglomerate. The limestone in turn is unconformably overlaid by
the merely tilted sandstone beds at the right of the map. These three
sedimentary formations clearly represent decreasing amounts of close
folding, from which it is clear that each earlier formation has passed
through an episode not shared by that of next younger age. Of the
other intrusive rocks, the dike of porphyry is younger than all the
other formations, with the possible exception of the upper sandstone.
Offsetting of the formations has disclosed the course of a fault, and
from its relations to the dikes we may learn that of these the porphyry
is younger and the basalt older than the date of the faulting.

The dashed lines upon the map (_AB_ and _CD_) have been selected as
appropriate lines along which to construct geological sections (Fig.
48, below map), and from these sections the _exposed_ thicknesses of
the different formations may be calculated. In one instance only,
that of the conglomerate, can we be sure that this exposed thickness
measures the entire formation.

=Fold _versus_ fault topography.=—The more resistant or “stronger”
rock beds, as regards attacks of the atmosphere, in the course of time
come to stand in relief, separated by depressions which overlie the
“weaker” formations. Simple open folds which are not plunging exercise
an influence upon topography by producing generally long and straight
ridges. More complex flexures, since they generally plunge, make
themselves apparent by features which in the map are represented by
curves. Fracture structures, and especially block displacements, are
differentiated from these curving features by the dominance of straight
or nearly rectilinear lines upon the map. The effect of erosion is to
reduce the asperity of features and to mold them with flowing curves.
The fracture structures are for this reason much more likely to be
overlooked, and if they are not to elude the observer, they must be
sought out with care. Fold and fracture structures may both be revealed
upon the same map.


 Joint systems:—

 JOHN PHILLIPS. Observations made in the Neighborhood of Ferrybridge in
 the Years 1826-1828, Phil. Mag., 2d ser., vol. 4, 1828, pp. 401-409;
 Illustrations of the geology of Yorkshire, Pt. II, The Limestone
 District. London, 1836, pp. 90-98.

 SAMUEL HAUGHTON. On the Physical Structure of the Old Red Sandstone of
 the County of Waterford, considered with reference to cleavage, joint
 surfaces, and faults, Trans. Roy. Soc. London, vol. 148, 1858, pp.

 W. C. BRÖGGER. Spaltenverwerfungen in der Gegend Langesund-Skien, Nyt
 Magazin for Naturvidernskaberne, vol. 28, 1884, pp. 253-419.

 WM. H. HOBBS. The Newark System of the Pomperaug Valley, Connecticut,
 21st Ann. Rept. U. S. Geol. Surv., Pt. III, 1901, pp. 85-143.

Geological map:—

 WM. H. HOBBS. The Interpretation of Geological Maps, School Science
 and Mathematics, vol. 9, 1909, pp. 644-653.



=Nature of earthquake shocks.=—Man’s belief in the stability of
Mother Earth—the _terra firma_—is so inbred in his nature that even
a light shock of earthquake brings a rude awakening. The terror which
it inspires is no doubt largely to be explained by this disillusionment
from the most fundamental of his beliefs. Were he better advised, the
long periods of quiet which separate earthquakes, and not the lighter
shocks which follow all grander disturbances, would occasion him

[Illustration: FIG. 49.—View of a portion of the ruins of Messina
after the earthquake of December 28, 1908.]

Earthquakes are the sensible manifestations of changes in level or of
lateral adjustments of portions of the continents, and the seismic
disturbances upon the sea—seaquakes and seismic sea waves—relate to
similar changes upon the floor of the ocean.

During the grander or catastrophic earthquakes, the changes are
indeed terrifying, and have usually been accompanied by losses to
life and property, which are only to be compared with those of great
conflagrations or of inundations on thickly populated plains. The
conflagration has all too frequently been an aftermath of the great
historic earthquakes. The earthquake of December 28, 1908, in southern
Italy, destroyed almost the entire population of a great city, and left
of its massive buildings only a confused heap of rubble (Fig. 49). Two
years later a heavy earthquake resulted in great damage to cities in
Costa Rica (Fig. 50), while two years earlier our own country was first
really awakened to the danger in which it stands from these convulsive
earth throes; though, as we shall see, these dangers can be largely met
through proper methods of construction.


FIG. 50.—Ruins of the Carnegie Palace of Peace at Cartago, Costa Rica,
destroyed when almost completed by the great earthquake of May 4, 1910
(after a photograph by Rear-Admiral Singer, U.S.N.).]

Earthquakes are usually preceded for a brief instant by subterranean
rumblings whose intensity appears to bear no relation to the shocks
which follow. The ground then rocks in wavelike motions, which, if of
large amplitude, may induce nausea, prevent animals from keeping upon
their feet, and wreck all structures not specially adapted to withstand
them. Heavy bodies are sometimes thrown up from the ground (Fig. 51),
and at other times similar heavy masses are, apparently because of
their inertia, more deeply imbedded in the earth. Thus gravestones and
heavy stone posts are often sunk more deeply in the ground and are
surrounded by a hollow and perhaps by small open cracks in the surface
(Fig. 52). When bodies are thrown upward, it would imply that a quick
upward movement of the ground had been suddenly arrested, while the
burial of heavy bodies in the earth is probably due to a movement which
begins suddenly and is less abruptly terminated.

[Illustration: FIG. 51.—Bowlders thrown into the air and overturned
during the Assam earthquake of 1897 (after R. D. Oldham).]


FIG. 52.—Heavy post sunk deeper into the ground during the Charleston
earthquake of August 31, 1886 (after Dutton).]

=Seaquakes and seismic sea waves.=—Upon the ocean the quakes which
emanate from the sea floor are felt on shipboard as sudden joltings
which produce the impression that the ship has struck upon a shoal,
though in most instances there is no visible commotion in the
water. The distribution of these shocks, as indicated either by the
experiences of neighboring ships at the time of a particular shock, or
by the records of vessels which at different times have sailed over an
area of frequent seismic disturbance, appears to be limited to narrow
zones or lines (Fig. 53). The same tendency of under-sea disturbances
to be localized upon definite straight lines has been often illustrated
by the behavior of deep-sea cables which are laid in proximity to one
another and which have been known to part simultaneously at points
ranged upon a straight line.


FIG. 53.—Map showing the localities at which shocks have been reported
at sea off Cape Mendocino, California.]

Far grander disturbances upon the floor of the ocean have been revealed
by the great sea waves—the so-called “tidal waves”, properly referred
to as _tsunamis_—which recur in those sea districts which adjoin the
special earthquake zones upon the continents (p. 86). The forerunner of
such a sea wave approaching the shore is usually a sudden withdrawal
of the water so as to lay bare a portion of the bottom, but this is
well-recognized to be the premonition of a gigantic oncoming wave
which sweeps all before it and is only halted when it has rolled over
all the low-lying country and encountered a mountain wall. Such
seismic waves have been especially common upon the Pacific shore of
South America and upon the Japanese littoral (Fig. 54). These waves
proceed from above the great deeps upon the ocean bottom, and clearly
result from the grander earth movements to which these depressions owe
their exceptional depth. The withdrawal of the water from neighboring
shores may be presumed to be connected with a descent of the floor
of the depression and the consequent drawing-in of the ocean surface
above. The later high wave would thus represent the dispersion of the
mountain of water which is raised by the meeting of the waters from the
different sides of the depression.

[Illustration: FIG. 54.—Effect of a seismic water wave at Kamaishi,
Japan, in 1896 (after E. R. Scidmore).]

[Illustration: FIG. 55.—A fault of vertical displacement.]

=The grander and the lesser earth movements.=—Upon the land the
grander and so-called catastrophic earthquakes are usually the
accompaniment of important changes in the surface of the ground that
will be discussed in later sections. Those shocks which do little
damage to structures produce no visible changes in the earth’s surface,
except, it may be, to shake down some water-soaked masses of earth upon
the steeper slopes. Still other movements, and these too slight to be
felt even in the night when the animal world is at rest, may yet be
distinguished by their sounds, the unmistakable rumblings which are
characteristic alike of the heaviest and the lightest of earthquake


FIG. 56.—Escarpment produced by an earthquake fault of vertical
displacement which cut across the Chedrang River and thus produced a
waterfall, Assam earthquake of 1897 (after R. D. Oldham).]

=Changes in the earth’s surface during earthquakes—faults and
fissures.=—Each of the grander among historic earthquakes has been
accompanied by noteworthy changes in the configuration of the earth’s
surface within the district where the shocks were most intense. A
section of the ground is usually found to have moved with reference
to another upon the other side of a vertical plane which is usually
to be seen; we have here to do with the actual making of a fault
or displacement such as we find the fossil examples of within the
rocks. The displacement, or throw, upon the fault plane may be either
upward or downward or laterally in one direction or the other, or
these movements may be combined. A movement of adjacent sections of
the ground upward or downward with reference to each other (Fig. 55)
has been often observed, notably at Midori after the great Japanese
earthquake of 1891, and in the Chedrang valley of Assam after the
earthquake of 1897 (Fig. 56).

[Illustration: FIG. 57.—A fault of lateral displacement.]


FIG. 58.—Fence parted and displaced fifteen feet by a transverse fault
formed during the California earthquake of 1906 (after W. B. Scott).]


FIG. 59.—Fault with vertical and lateral displacements combined.]

A lateral throw, unaccompanied by appreciable vertical displacement
(Fig. 57), is especially well illustrated by the fault in California
which was formed during the earthquake of 1906 (Fig. 58). A combination
of the two types of displacement in one (Fig. 59) is exemplified by the
Baishiko fault of Formosa at the place shown in plate 3 A.

=The measure of displacement.=—To afford some measure of the
displacements which have been observed upon earthquake faults, it may
be stated that the maximum vertical throw measured upon the fault in
the Neo valley of Japan (1891) was 18 feet, in the Chedrang valley of
Assam (1897) 35 feet, and of the Alaskan coast (1899) 47 feet. Large
sections of land were bodily uplifted in these cases within the space
of a few seconds, or at most a few minutes, by the amounts given. The
largest recorded lateral displacement measured upon an earthquake fault
is about 21 feet upon the California rift after the earthquake of 1906;
though an amount only slightly less than this is indicated in the
shifting of roads and arroyas dating from the earthquake of 1872 in the
Owens valley, California. Fault lines once established are planes of
special weakness and become later the seat of repeated movements of the
same kind.

│  PLATE 3.                                                            │
│                                                                      │
│ [Illustration: _A._ An earthquake fault opened in Formosa in 1906,   │
│ with vertical and lateral displacements combined (after Omori).]     │
│                                                                      │
│ [Illustration: _B._ Earthquake faults opened in Alaska in 1889, on   │
│ which vertical slices of the earth’s shell have undergone individual │
│ adjustments (after Tarr and Martin).]                                │


FIG. 60.—Diagram to show how small faults in the rock basement may be
masked at the surface through adjustments within the loose rock mantle.]

The greater number of earthquake faults are found in the loose rock
cover which so generally mantles the firmer rock basement, and it is
almost certain that the throws within the solid rock are considerably
larger than those which are here measured at the surface, owing to
the adjustments which so readily take place in the looser materials.
Those lighter shocks of earthquake which are accompanied by no visible
displacements at the surface do, however, in some instances affect
in a measure the flow of water upon the surface, and thus indicate
that small changes of surface level have occurred without breaks
sufficiently sharp to be perceived (Fig. 60). Intermediate between the
steep escarpment and the masked displacement just described is the
so-called “mole-hill” effect,—a rounded and variously cracked slope or
ridge above the position of a buried fault (Fig. 61).


FIG. 61.—Diagram to show the appearance of a “mole hill” above a
buried earthquake fault (after Kotô).]

The escarpments due to earthquake faults in loose materials at the
earth’s surface can obviously retain their steepness for a few years
or decades at the most; for because of their verticality they must
gradually disappear in rounded slopes under the action of the elements.
Smaller displacements within a rock which rapidly disintegrates under
the action of frost and sun will likewise before long be effaced. In
those exceptional instances where a resistant rock type has had all
altered upper layers planed away until a fresh and hard surface is
exposed, and has further been protected from the frost and sun beneath
a thin layer of soil, its original surface may be retained unaltered
for many centuries. Upon such a surface the lightest of sensible
shocks, or even the smaller earth movements which are not perceived
at the time, may leave an almost indelible record. Such records
particularly show that the movements which they register occur upon the
planes of jointing within the rock, and that these ready formed cracks
have probably been the seats of repeated and cumulative adjustments
(Fig. 62).


FIG. 62.—Post-glacial earthquake faults of small but cumulative
displacement, eastern New York (after Woodworth).]

[Illustration: FIG. 63.—Earthquake cracks in Colorado desert (after a
photograph by Sauerven).]

=Contraction of the earth’s surface during earthquakes.=—The wide
variations in the amount of the lateral displacement upon earthquake
faults, like those opened in California in 1906, show that at the
time of a heavy earthquake there must be large local changes in the
density of the surface materials. Literally, thousands of fissures may
appear in the lowlands, many of them no doubt a secondary effect of
the shaking, but others, like the _quebradas_ of the southern Andes or
the “earthquake cracks” in the Colorado desert (Fig. 63), may have a
deeper-seated origin. Many facts go to show, however, that though local
expansion does occur in some localities, a surface contraction is a
far more general consequence of earth movement. In civilized countries
of high industrial development, where lines of metal of one kind or
another run for long distances beneath or upon the surface of the
ground, such general contraction of the surface may be easily proven.
Comparatively seldom are lines of metal pulled apart in such a way as
to show an expansion of the surface; whereas bucklings and kinkings of
the lines appear in many places to prove that the area within which
they are found has, as a whole, been reduced.

[Illustration: FIG. 64.—Diagrams to show how railway tracks are either
broken or buckled locally within the district visited by an earthquake.]

[Illustration: FIG. 65.—The Biwajima railroad bridge in Japan after
the earthquake of 1891 (after Milne and Burton).]


FIG. 66.—Diagrams to show how the compression of a district and its
consequent contraction during an earthquake may close up the joint
spaces within the rock basement and concentrate the contraction of the
overlying mantle where this is partially cut through and so weakened in
the valley sections.]

Water pipes laid in the ground at a depth of some feet may be bowed up
into an arch which appears above the surface; lines of curbing are
raised into broken arches, and the tracks of railways are thrown into
local loops and kinks which imply a very considerable local contraction
of the surface (Fig. 64). With unvarying regularity railway or other
bridges which cross rivers or ravines, if the structures are seriously
damaged, indicate that the river banks have drawn nearer together at
the time of the disturbance. In such cases, whenever the bridge girder
has remained in place upon its abutments, these have either been broken
or back-tilted as a whole in such a manner as to indicate an approach
of the foundations which was prevented at the top by the stiffness of
the girder (Fig. 65).


FIG. 67.—Map of the Chedrang fault which made its appearance during
the Assam earthquake of 1897. The figures give the amounts of the local
vertical displacement measured in feet (after R. D. Oldham).]

The simplest explanation of such an approach of the banks at the
sides of the valleys cut in loose surface material is to be found in
a general closing up of the joint spaces within the underlying rock,
and an adjustment of the mantle upon the floor mainly in the valley
sections (Fig. 66).


FIG. 68.—Map giving the displacements in feet measured along an
earthquake fault formed in Alaska in 1899 (after Tarr and Martin).]

=The plan of an earthquake fault.=—In our consideration of earthquake
faults we have thus far given our attention to the displacement as
viewed at a single locality only. Such displacements are, however,
continued for many miles, and sometimes for hundreds of miles; and when
now we examine a map or plan of such a line of faulting, new facts of
large significance make their appearance. This may be well illustrated
by a study of the plan of the Chedrang fault which appeared at the
time of the Assam earthquake of 1897 (Fig. 67). From this map it
will be noticed that the upward or downward displacement upon the
perpendicular plane of the fault is not uniform, but is subject to
large and _sudden_ changes. Thus in order the measurements in feet are
32, 0, 18, 35, 0, 8, 25, 12, 8, 2, 0. The fault formed in 1899 upon
the shores of Russell Fjord in Alaska (Fig. 68) reveals similar sudden
changes of throw, only that here the direction of the movement is often
reversed; or, otherwise expressed, the upthrow is suddenly transferred
from one side of the fault to the other. Such abrupt changes in the
direction of the displacement have been observed upon many earthquake
faults, and a particularly striking one is represented in Fig. 69.


FIG. 69.—Abrupt change in the direction of throw upon an earthquake
fault which was formed in the Owens valley, California, in 1872. The
observer looks directly along the course of the fault from the left
foreground to the cliff beyond and to the left of the impounded water
(after a photograph by W. D. Johnson).]

=The block movements of the disturbed district.=—The displacements
upon earthquake faults are thus seen to be subdivided into sections,
each of which differs from its neighbors upon either side and is
sharply separated from them, at least in many instances. These points
of abrupt change of displacement are, in many cases at least, the
intersection points with transverse faults (Fig. 69). Such points of
abrupt change in the degree or in the direction of the displacement may
be, when looked at from above, abrupt turning points in the direction
of extension of the fault, whose course upon the map appears as a
zigzag line made up of straight sections connected by sharp elbows
(Fig. 70).


FIG. 70.—Map of the faults within an area of the Owens valley,
California, formed in part during the earthquake of 1872, and in part
due to early disturbances, In the western portions the displacements
cut across firm rock and alluvial deposits alike without deviation of
direction (after a map by W. D. Johnson).]

Such a grouping of surface faults as are represented upon the map is
evidence that the area of the earth’s shell, which is included, has at
the time of the earthquake been subject to adjustments as a series of
separate units or blocks, certain of the boundaries of which are the
fault lines represented. The changes in displacement measured upon the
larger faults make it clear that the observed faults can represent but
a fraction of the total number of lines of displacement, the others
being masked by variations in the compactness of the loose mantling
deposits. Could we but have this mantle removed, we should doubtless
find a rock floor separated into parts like an ancient Pompeiian
pavement, the individual blocks in which have been thrown, some upward
and some downward, by varying amounts. Less than a hundred miles away
to the eastward from the Owens Valley, a portion of this pavement has
been uncovered in the extensive operations of the Tonapah Mining
District, so that there we may study in all its detail the elaborate
pattern of earth marquetry (Fig. 71) which for the floor of the Owens
valley is as yet denied us.


FIG. 71.—Marquetry of the rock floor of the Tonapah Mining District,
Nevada (after Spurr).]


FIG. 72.—Map of a portion of the Alaskan coast to show the adjustments
in level during the earthquake of 1899 (after Tarr and Martin).]

=The earth blocks adjusted during the Alaskan earthquake of 1899.=—For
a study of the adjustments which take place between neighboring earth
blocks during a great earthquake, the recent Alaskan disturbance
has offered the advantage that the most affected district was upon
the seacoast, where changes of level could be referred to the datum
of the sea’s surface. Here a great island and large sections of the
neighboring shore underwent movements both as a whole in large blocks
and in adjustments of their subordinate parts among themselves (Fig.
72). Some sections of the coast were here elevated by as much as 47
feet, while neighboring sections were uplifted by smaller amounts (Fig.
73), and certain smaller sections were even dropped below the level
of the sea.


FIG. 73.—View on Haencke Island, Disenchantment Bay, Alaska, revealing
the shore that rose seventeen feet above the sea during the earthquake
of 1899, and was found with barnacles still clinging to the rock (after
Tarr and Martin).]

The amount of such subsidence is, however, difficult
to ascertain, for the reason that the former shore features are now
covered with water and thus removed from observation. In favorable
localities the minimum amount of submergence may sometimes be measured
upon forest trees which are now flooded with sea water. In Fig. 74
a portion of the coast is represented where the beach sand is now
extended back into the spruce forest, a distance of a hundred feet or
more, and where sedgy beach grass is growing among trees whose roots
are now laved in salt water. At the front of this forest the great
storm waves overturn the trees and pile the wreckage in front of those
that still remain standing.


FIG. 74.—Partially submerged forest upon the shore of Knight Island,
Alaska, due to the sinking of a section of the coast during the
earthquake of 1899 (after Tarr and Martin).]


FIG. 75.—Settlement of a section of the shore at Port Royal, Jamaica,
during the earthquake of January 14, 1907, adjacent to a similar but
larger settlement of the near shore during the earthquake of 1692
(after a photograph by Brown).]

Upon the glaciated rock surfaces of the Alaskan coast, exceptionally
favorable opportunities are found for study of the intricate pattern
of the earth mosaic which is under adjustment at the time of an
earthquake. Upon Gannett Nunatak the surface was found divided by
parallel faults into distinct slices which individually underwent small
changes of level (plate 3 B).



=Experimental demonstration of earth movements.=—The study of the
Alaskan earthquake of 1899 showed that during this adjustment within
the earth’s shell some of the local blocks moved upward and by larger
amounts than their neighbors, and that still others were actually
depressed so that the sea flowed over them. It must be evident that
such differential vertical movements of neighboring blocks at the
earth’s surface can only take place if lateral transfers of material
are made beneath it. From under those strips of coast land which were
depressed, material must have been moved so as to fill the void which
would otherwise have formed beneath the sections that were uplifted.
If we take into consideration much larger fractions upon the surface
of our planet, we are taught by the great seaquakes which are now
registered upon earthquake instruments at distant stations that large
_downward_ movements are to-day in progress beneath the sea much more
than sufficient to compensate all extensions of the earth’s surface
within those districts where the land is rising in mountains. From
under the offshore deeps of the ocean to beneath the growing mountains
upon the shore, a transfer of earth material must be assumed to take
place when disturbances are registered.

Within the time interval that separates the sudden adjustments of
the surface which are manifested in earthquakes, the condition of
strain which brings them about is steadily accumulating, due, as we
generally assume, to earth contraction through loss of its heat. It
seems probable that the resistance to an immediate adjustment is found
in the rigidity of the shell because of the compression to which it is
subjected. To illustrate: a row of blocks well fitted to each other
may be held firmly as a bridge between the jaws of a vice, because so
soon as each block starts to fall a large resistance from friction upon
its surface is called into existence, a force which increases with the
degree of compression.

It is thus possible upon this assumption crudely to demonstrate the
adjustment of earth blocks by the simple device represented in plate
4 A. The construction of this experimental tank is so simple that
little explanation is necessary. Wooden blocks of different heights
are supported in water within a tank having a glass front, and are
kept in a strained condition at other than their natural positions of
flotation by the compression of a simple vice at the top. Held firmly
in this position, they may thus represent the neighboring blocks within
the earth’s outer shell which are supported upon relatively yielding
materials beneath, and prevented from at once adjusting themselves
to their natural positions through the compression to which they are
subjected. Held as they now are, the water near the ends of the tank
is forced up beneath the blocks to higher than its natural level, and
thus tends to flow from both ends toward the center. Such a movement
would permit the end blocks to drop and force the middle ones to rise.
The end blocks are, let us say, the sections of Alaskan coast line
which sunk during the earthquake, as the center blocks are the sections
which rose the full measure of 47 feet. Upon a larger scale the end
blocks may equally well be considered as the floor of the great deeps
off the Alaskan coast, whose sinking at the time of the earthquake was
the cause of the great sea wave. Upon this assumption the center blocks
would represent the Alaskan coast regarded as a whole, which underwent
a general uplift.

Though we may not, in our experiment, vary the tendency to adjustment
by any contractional changes in either the water or the blocks, we may
reduce the compression of the vice, which leads to the same general
result. As the compression of the vice is slowly relaxed, a point is
at last reached at which friction upon the block surfaces is no longer
sufficient to prevent an adjustment taking place, and this now suddenly
occurs with the result shown in plate 4 B. In the case of the earth
blocks, this sudden adjustment is accompanied by mass movements of the
ground separated by faults, and these movements produce successional
vibrations that are particularly large near the block margins, and
other frictional vibrations of such small measure as to be generally
appreciated by sounds only. The jolt of the adjustments has thrown some
blocks beyond their natural position of rest, and these sink and rise
subsequently in order to readjust themselves with lighter vibrations,
which may be repeated and continued for some time. In the case of the
earth these later adjustments are the so-called _aftershocks_ which
usually continue throughout a considerable period following every great
earthquake. Gradually they fall off in intensity and frequency until
they can no longer be felt, and are thereafter continued for a time as
rumblings only.

│  PLATE 4.                                                             │
│                                                                       │
│ [Illustration:                                                        │
│                                                                       │
│ _A._ Experimental tank to illustrate the earth movements which are    │
│ manifested in earthquakes. The sections of the earth’s shell are here │
│ represented before adjustment has taken place.]                       │
│                                                                       │
│ [Illustration: _B._ The same apparatus after a sudden adjustment.]    │
│                                                                       │
│ [Illustration: _C._ Model to illustrate a block displacement in rocks │
│ which are intersected by master joints.]                              │

=Derangement of water flow by earth movement.=—The water which
supported the blocks in our experiment has represented the more mobile
portion of the earth’s substance beneath its outer zone of fracture.
The surface water layers in the tank may, however, be considered
in a different way, since their behavior is remarkably like that
of the water within and upon the earth’s surface during an earth
adjustment. At the instant when adjustment takes place in the tank,
water frequently spurts upward from the cracks between the sinking end
blocks; and if in place of one of the higher center blocks we insert
one whose top is below the level of the water in the tank, a “lake”
will be formed above it. When the adjustment occurs, this lake is
immediately drained by outflow of the water at its bottom along one of
the cracks between the blocks (Fig. 76).


FIG. 76.—Diagrams to illustrate the draining of lakes during

Such derangements of water flow as have been illustrated by the
experiment are among the commonest of the phenomena which accompany
earthquakes. Lakes and swamp lands have during earthquakes been
suddenly drained, fountains of water have been seen to shoot up
from the surface and have played for some minutes or hours before
their sudden disappearance in a sucking down of the water with later
readjustment. During the great earthquake of the lower Mississippi
valley in 1811, known as the New Madrid earthquake, the earlier Lake
Eulalie was completely drained, and upon the now exposed bed there
appeared parallel fissures on which were ranged funnel-like openings
down which the water had been sucked. In other sections of the affected
region the water shot up in sheets along fissures to the tops of high
trees. Areas where such spurting up of the water has been observed have
in most cases been shown to correspond to areas of depression, and such
areas have sometimes been left flooded with water. During the Indian
earthquake of 1819 an area of some 200 square miles suddenly sank and
was transformed into a lake.


FIG. 77.—Diagram to illustrate-the derangements of flow of water at
the time of an earthquake; water issuing at the surface over downthrown
rocks, and being sucked down in upthrown blocks.]


FIG. 78.—Mud cones aligned upon a fissure opened at Moraza, Servia,
during the earthquake of April 4, 1904 (after Michailovitch).]

=Sand or mud cones and craterlets.=—From a very moderate depth below
the surface to that of several miles, all pore spaces and all larger
openings within the rock are completely filled with water, the “trunk
lines” of whose circulation is by way of the joints or along the
bedding planes of the rocks. The principal reservoirs, so to speak, of
this water inclosed within the rock are the porous sand formations.
When, now, during an earthquake a block of the earth’s shell is
suddenly sunk and as suddenly arrested in its downward movement, the
effect is to compress the porous layers and so force the contained
water upward along the joints to the surface, carrying with it large
quantities of the sand (Fig. 77).


FIG. 79.—One of the many craterlets formed near Charleston, South
Carolina, during the earthquake of August 31, 1886. The opening is
twenty feet across, and the leaves about it are encased in sand as were
those upon the branches of the overhanging trees to a height of some
twenty feet (after Dutton).]


FIG. 80.—Cross section of a craterlet to show the trumpet-like form of
the sand column.]

Ejected at the surface this water appears in fountains usually arranged
in line over joints, or even in continuous sheets, and the sand
collecting about the jets builds up lines of _sand_ or _mud cones_
sometimes described as “mud volcanoes” (Fig. 78). The amount of sand
thus poured out is sometimes so great that blankets of quicksand are
spread over large sections of the country. Most frequently, however,
the sand is not built above the general level of the surface, but forms
a series of _craterlets_ which are largely shaped as the water is
sucked down at the time of the readjustment with which the play of such
earthquake fountains is terminated (Fig. 79). Subsequent excavations
made about such craterlets have shown them to have the form of a
trumpet, and that in the sand which so largely fills them there are
generally found scales of mica and such light bodies as would be picked
out from the heterogeneous materials of the sand layers and carried
upward in the rush of water to the surface (Fig. 80).

=The earth’s zones of heavy earthquake.=—Since earthquakes give notice
of a change of level of the ground, the special danger zones from this
source are the growing mountain systems which are usually found near
the borders of the sea. Such lines of mountains are to-day rising where
for long periods in the past were the basins of deposition of former
seas. They thus represent the zones upon the earth’s surface which
are the most unstable—which in the recent period have undergone the
greatest changes of level.


FIG. 81.—Map of the island of Ischia to show how the shocks of recent
earthquakes have been concentrated at the crossing point of two
fractures (after Mercalli and Johnston-Lavis).]

By far the most unstable belt upon the earth’s surface is the rim
surrounding the Pacific Ocean, within which margin it has been
estimated that about 54 per cent of the recorded shocks of earthquake
have occurred. Next in importance for seismic instability is the
zone which borders both the Mediterranean Sea and the Caribbean—the
American Mediterranean—and is extended across central Asia through
the Himalayas into Malaysia. Both zones approximate to great circles
upon the earth’s surface and intersect each other at an angle of about
67°. It has been estimated that about 95 per cent of the recorded
continental earthquakes have emanated from these belts.


FIG. 82.—A line of earth fracture indicated in the plan of the relief,
which may at any time become the seat of movement and resultant shock.]

=The special lines of heavy shock.=—Within any earthquake district
the shocks are not felt with equal severity at all places, but there
are, on the contrary, definite lines which the disturbance seems to
search out for special damage. From their relations to the relief of
the land these lines would appear to be lines of fracture upon the
boundaries of those sections of the crust that play individual rôles in
the block adjustment which takes place. More or less masked as these
lines are beneath the rounded curves of the landscape, they are given
an altogether unenviable prominence with each succeeding earthquake. At
such times we may think of the earth’s surface as specially sensitized
for laying bare its hidden structure, as is the sensitized plate under
the magical influence of the X rays.

When, at the time of an earthquake, blocks are adjusted with reference
to their neighbors, the movements of oscillation are greatest in
those marginal portions of direct contact. Corners of blocks—the
intersecting points of the important faults—should for the same
reason be shaken with a double violence, and this assumption appears
to be confirmed by observation. Upon the island of Ischia, off the
Bay of Naples, the shocks from recent earthquakes have been strangely
concentrated near the town of Casamicciola, which was last destroyed
in 1883. This unfortunate city lies at the crossing point of important
fractures whose course upon the island is marked by numerous springs
and _suffioni_ (Fig. 81).

=Seismotectonic lines.=—The lines of important earth fractures, as
will be more clearly shown in the sequel (p. 227), are often indicated
with some clearness by straight lines in the plan of the surface
relief (Fig. 82). Lines of this nature are easily made out upon
the map of the West Indies, and if we represent upon it by circles
of different diameters the combined intensities of the recorded
earthquakes in the various cities, it appears that the heavily shaken
localities are ranged upon lines stamped out in the relief, with
the most severely damaged places at their intersections (Fig. 83).
These lines of exceptional instability are known as _seismotectonic
lines_—earthquake structure lines.

[Illustration: FIG. 83.—Seismotectonic lines of the West Indies.]

=The heavy shocks above loose foundations.=—It is characteristic of
faults that they soon bury themselves from sight under loose materials,
and are thus made difficult of inspection. The escarpment which is the
direct consequence of a vertical displacement upon a fault tends to
migrate from the place of its formation, rounding the surface as it
does so and burying the fault line beneath its deposits (Fig. 43, p.

This is not, however, the sole reason why loose foundations should be
places of special danger at the time of earth shocks, for the reason
that earthquake waves are sent out in all directions from the surfaces
of displacement through the medium of the underlying rock. These
waves travel within the firm rock for considerable distances with
only a gradual dissipation of their energy, but with their entry into
the loose surface deposits their energy is quickly used up in local
vibrations of large amplitude, and hence destructive to buildings.


FIG. 84.—Device to illustrate the different effects upon the
transmission and the character of shocks which are produced by firm
rock and by loose materials.]

The essential difference between firm rock and such loose materials as
are found upon a river bottom or in the “made land” about our cities
may be illustrated by the simple device which is represented in Fig.
84. Two similar metal pans are suspended from a firm support by bands
of steel and “elastic” braid of similar size and shape, and carry each
a small block of wood standing upon its end. Similar light blows are
now administered directly to the pans with the effect of upsetting that
block which is supported by the loose braid because of the large range
or amplitude of movement that is imparted to the pan. The “elastic”
braid, because of these large vibrations of which it is susceptible,
may represent the loose materials when an earthquake wave passes into
them. In the case of the steel support, the energy of the blow, instead
of being dissipated in local swingings of the pan, is to a large extent
transmitted through the elastic metal to materials beyond. The steel
thus resembles in its high elasticity the firmer rock basement, which
receives and transmits the earthquake shocks, but except when ruptured
in a fault is subject to vibrations of small amplitude only.

=Construction in earthquake regions.=—Wherever earthquakes have
been felt, they are certain to occur again; and wherever mountains
are growing or changes of level are in progress, there no record of
past earthquakes is required in order to forecast the future seismic
history. Although the future earthquakes may be predicted, the time of
their coming is, fortunately or unfortunately, still hidden from us. If
one’s lot is to be cast in an earthquake country, the only sane course
to pursue is to build with due regard to future contingencies.

The danger, from destructive fires may to-day be largely met by methods
of construction which levy an additional burden of cost. Though the
danger from seismic disturbances can hardly be met as fully as that
from fire, yet it is true that buildings may be so constructed as to
withstand all save those heaviest shocks in the immediate vicinity of
the lines of large displacement. Here, also, a considerable additional
expense is involved in the method of construction, in the case of
residences particularly.

From what has been said, it is obvious that much of the danger from
earthquakes can be met by a choice of site away from lines of important
fracture and from areas of relatively loose foundation. The choice of
building materials is next of importance. Those buildings which succumb
to earthquakes are in most cases racked or shaken apart, and thus they
become a prey to their own inherent properties of inertia. Each part of
a structure may be regarded as a weight which is balanced upon a stiff
rod and pivoted upon the ground. When shocks arrive, each part tends
to be thrown into vibration after the manner of an inverted pendulum.
In proportion, therefore, as the weights are large and rest upon long
supports, the danger of overthrow and of tearing apart is increased.
In general, structures are best constructed of light materials whose
weight is concentrated near the ground. Masonry structures, and
especially high ones, are, therefore, the least suited for resisting
earthquakes, of which the late complete destruction of the city of
Messina is a grewsome reminder. Despite repeated warnings in the past,
the buildings of that stricken city were generally constructed of heavy
rubble, which in addition had been poorly cemented (Fig. 49, p. 67).
Such structures are usually first ruptured at the edges and corners,
since here the vibrations which tend to tear the building asunder are
resisted by no supports and are reënforced from neighboring walls.


FIG. 85.—House wrecked in San Francisco earthquake of 1906 because the
floors and partitions were not securely fastened to the walls (after R.
L. Humphrey).]

An advantage of the first importance is evidently secured if the rods
of the pendulum, of which the building is conceived to be composed,
have sufficient elasticity to be considerably distorted without
rupture and to again recover their original position. This is the
supreme advantage of structural steel for all large buildings, which
is coupled, however, with the disadvantage that the riveted fastenings
are apt to be quickly sheered off under the vibrations. Large and high
buildings, when sufficiently elastic, have fortunately the property of
destroying the earth waves by interference before they have traveled
above the lower stories.

For large structures in which wood cannot be used, strongly reënforced
concrete is well adapted, for it has in general the same advantages
as steel with somewhat reduced elasticity, but with a more effective
binding together of the parts. This requirement of thorough bracing
and tying together of the several parts of a building causes it to
vibrate, not as many pendulums, but as one body. If met, it removes
largely the danger from racking strains, and for small structures
particularly it is the requirement which is most easily complied with.
For such buildings it is therefore necessary that the framework should
be built in a close network with every joint firmly braced and with
all parts securely tied together. Especial attention should be given
to the fastenings of floor and partition ends. The house shown in Fig.
85 could not have been subjected to heavy shocks, for though the walls
are thrown down, the floors and partitions have been left near their
original positions.


FIG. 86.—Building wrecked at San Mateo, California, during the late
earthquake. The heavy roof and upper floor, acting as a unit, have
battered down the upper walls (after J. C. Branner).]

This tendency of the walls, floors, partitions, and roof to act as
individual units in the vibration, is one that must be reckoned with
and be met by specially effective bracing and tying at the junctions.
Otherwise these larger parts of the structure may act like battering
rams to throw over the walls or portions of them (Fig. 86).


 General works:—

 JOHN MILNE. Seismology. London, 1898, pp. 320.

 C. E. DUTTON. Earthquakes in the Light of the New Seismology. Putnam,
 New York, 1904, pp. 314.

 A. SIEBERG. Handbuch der Erdbebenkunde. Braunschweig, 1904, pp. 362.

 COUNT F. DE MONTESSUS DE BALLORE. Les Tremblements de Terre,
 Géographie Séismologique. Paris, 1906, pp. 475; La Science
 Séismologique. Paris, 1907, pp. 579.

 WILLIAM HERBERT HOBBS. Earthquakes, an Introduction to Seismic
 Geology. Appleton, New York, 1907, pp. 336.

 C. G. KNOTT. The Physics of Earthquake Phenomena. Clarendon Press,
 Oxford, 1908, pp. 283.

 E. RUDOLPH. Ueber Submarine Erdbeben und Eruptionen, Beiträge zur
 Geophysik, vol. 1, 1887, pp. 133-365; vol. 2, 1895, pp. 537-666; vol.
 3, 1898, pp. 273-536.

Descriptive reports of some important earthquakes:—

 C. E. DUTTON. The Charleston Earthquake of August 31, 1886, 9th Ann.
 Rept. U. S. Geol. Surv., 1889, pp. 203-528.

 B. KOTÔ. On the Cause of the Great Earthquake in Central Japan, 1891,
 Jour. Coll. Sci. Imp. Univ., Tokyo, Japan, vol. 5, 1893, pp. 295-353,
 pls. 28-35.

 JOHN MILNE and W. K. BURTON. The Great Earthquake of Central Japan.
 1891, pp. 10, pls. 30.

 R. D. OLDHAM. Report on the Great Earthquake of 12th June, 1897, Mem.
 Geol. Surv. India. Vol. 29, 1899, pp. 379, pls. 42.

 A. C. LAWSON, and others. The California Earthquake of April 18, 1906,
 Report of the State Earthquake Investigation Commission, three quarto
 vols. (Carnegie Institution of Washington); many plates and figures.

 _Italian Photographic Society_, Messina and Reggio before and after
 the Earthquake of December 28, 1908 (an interesting collection of
 pictures). Florence, 1909.

 R. S. TARR and L. MARTIN. Recent Changes of Level in the Yakutat Bay
 Region, Alaska, Bull. Geol. Soc. Am., vol. 17, 1906, pp. 29-64, pls.

 WILLIAM HERBERT HOBBS. The Earthquake of 1872 in the Owens Valley,
 California, Beiträge zur Geophysik, vol. 10, 1910, pp. 352-385, pls,

Faults in connection with earthquakes:—

 WILLIAM H. HOBBS. On Some Principles of Seismic Geology, Beiträge zur
 Geophysik, vol. 8, 1907, Chapters iv-v.

Expansion or contraction of the earth’s surface during earthquakes:—

 WILLIAM H. HOBBS. A Study of the Damage to Bridges during Earthquakes,
 Jour. Geol., vol. 16, 1908, pp. 636-653; The Evolution and the Outlook
 of Seismic Geology, Proc. Am. Phil. Soc., vol. 48, 1909, pp. 27-29.

Earthquake construction:—

 JOHN MILNE. Construction in Earthquake Countries, Trans. Seis. Soc.,
 Japan, vol. 14, 1889-1890, pp. 1-246.

 F. DE MONTESSUS DE BALLORE. L’art de bâtir dans les pays à
 tremblements de terre (34th Congress of French Architects),
 L’Architecture, 193 Année, 1906, pp. 1-31.

 GILBERT, HUMPHREY, SEWELL, and SOULÉ. The San Francisco Earthquake and
 Fire of April 18, 1906, and their Effects on Structures and Structural
 Materials, Bull. 324, U. S. Geol. Surv., 1907, pp. 1-170, pls. 1-57.

 WILLIAM H. HOBBS. Construction in Earthquake Countries, The
 Engineering Magazine, vol. 37, 1909, pp. 1-19.

 LEWIS ALDEN ESTES. Earthquake-proof Construction, a discussion of the
 effects of earthquakes on building construction with special reference
 to structures of reënforced concrete, published by Trussed Concrete
 Steel Company. Detroit, 1911, pp. 46.




=Prevalent misconceptions about volcanoes.=—The more or less common
impression that a volcano is a “burning mountain” or a “smoking
mountain” has been much fostered by the school texts in physical
geography in use during an earlier period. The best introduction
to a discussion of volcanoes is, therefore, a disillusionment from
this notion. Far from being burning or smoking, there is normally
no combustion whatever in connection with a volcanic eruption. The
unsophisticated tourist who, looking out from Naples, sees the steam
cap which overhangs the Vesuvian crater tinged with brown, easily
receives the impression that the material of the cloud is smoke. Even
more at night, when a bright glow is reflected to his eye and soon
fades away, only to again glow brightly after a few moments have
passed, is it difficult to remove the impression that one is watching
an intermittent combustion within the crater. The cloud which floats
away from the crest of the mountain is in reality composed of steam
with which is admixed a larger or smaller proportion of fine rock
powder which gives to the cloud its brownish tone. The glow observed at
night is only a reflection from molten lava within the crater, and the
variation of its brightness is explained by the alternating rise and
fall of the lava surface by a process presently to be explained.

Not only is there no combustion in connection with volcanic eruptions,
but so far as the volcano is a mountain it is a product of its own
action. The grandest of volcanic eruptions have produced no mountains
whatever, but only vast plains or plateaus of consolidated molten rock,
and every volcanic mountain at some time in its history has risen out
of a relatively level surface.

When the traditional notions about volcanoes grew up, it was supposed
that the solid earth was merely a “crust” enveloping still molten
material. As has already been pointed out in an earlier chapter, this
view is no longer tenable, for we now know that the condition of matter
within the earth’s interior, while perhaps not directly comparable to
any that is known, yet has properties most resembling known matter in
a solid state; it is much more rigid than the best tool steel. While
there must be reservoirs of molten rock beneath active volcanoes, it
is none the less clear that they are small, local, and temporary.
This is shown by the comparative study of volcanic outlets within any
circumscribed district.

It is perhaps not easy to frame a definition of a volcano, but its
essential part, instead of being a mountain, is rather a vent or
channel which opens up connection between a subsurface reservoir of
molten rock and the surface of the earth. An eruption occurs whenever
there is a rise of this material, together with more or less steam and
admixed gases, to the surface. Such molten rock arriving at the surface
is designated _lava_. The changes in pressure upon this material during
its elevation induce secondary phenomena as the surface is approached,
and these manifestations are often most awe inspiring. While often
locally destructive, the geological importance of such phenomena is by
reason of their terrifying aspect likely to be greatly exaggerated.

=Early views concerning volcanic mountains.=—As already pointed out,
a volcano at its birth is not a mountain at all, but only, so to
speak, a shaft or channel of communication between the surface and a
subterranean reservoir of molten rock. By bringing this melted rock to
the surface there is built up a local elevation which may be designated
a mountain, except where the volume of the material is so large and is
spread to such distances as to produce a plain (see fissure eruptions

In the early history of geology it was the view of the great German
geologist von Buch and his friend and colleague von Humboldt, that a
volcanic mountain was produced in much the same manner as is a blister
upon the body. The fluids which push up the cuticle in the blister were
here replaced by fluid rock which elevated the sedimentary rock layers
at the surface into a dome or mound which was open at the top—the
so-called _crater_. This “elevation-crater” theory of volcanoes long
held the stage in geological science, although it ignored the very
patent fact that the layers on the flanks of volcanic cones are not
of sedimentary rock at all, but, on the contrary, of the volcanic
materials which are brought up to the surface during the eruption.
The observational phase of science was, however, dawning, and the
English geologists Scrope and Lyell were able to show by study of
volcanic mountains that the mound about the volcanic vent was due to
the accumulation of once molten rock which had been either exuded or
ejected. Making use of data derived from New Zealand, Scrope showed
that, instead of being elevated during the formation of a volcanic
mountain, the sedimentary strata of the vicinity may be depressed near
the volcanic vent (Fig. 87).


FIG. 87.—Breached volcanic cone near Auckland, New Zealand, showing
the bending down of the sedimentary strata in the neighborhood of the
vent (after Heaphy and Scrope).]

=The birth of volcanoes.=—To confirm the impression that the formation
of the volcanic mountain is in reality a secondary phenomenon
connected with eruptions, we may cite the observed birth of a number
of volcanoes. On the 20th of September, 1538, a new volcano, since
known as Monte Nuovo (new mountain), rose on the border of the ancient
Lake Lucrinus to the westward of Naples. This small mountain attained
a height of 440 feet, and is still to be seen on the shore of the bay
of Naples. From Mexico have been recorded the births of several new
volcanoes: Jorullo in 1759, Pochutla in 1870, and in 1881 a new volcano
in the Ajusco Mountains about midway between the Gulf of Mexico and the
Pacific Ocean. The latest of new volcanoes is that raised in Japan on
November 9, 1910, in connection with the eruption of Usu-san. This “New
Mountain” reached an elevation of 690 feet.


FIG. 88.—View of the new Camiguin volcano from the sea. It was formed
in 1871 over a nearly level plain. The town of Catarman appears at the
right near the shore (after an unpublished photograph by Professor Dean
C. Worcester).]

As described by von Humboldt, Jorullo rose in the night of the 28th
of September, 1759, from a fissure which opened in a broad plain at
a point 35 miles distant from any then existing volcano. The most
remarkable of new volcanoes rose in 1871 on the island of Camiguin
northward from Mindanao in the Philippine archipelago. This mountain
was visited by the _Challenger_ expedition in 1875, and was first
ascended and studied thirty years later by a party under the leadership
of Professor Dean C. Worcester, the Secretary of the Interior of the
Philippine Islands, to whom the writer is indebted for this description
and the accompanying illustration of this largest and most interesting
of new-born volcanoes. As in the case of Jorullo, the eruption began
with the formation of a fissure in a level plain, some 400 yards
distant from the town of Catarman (Fig. 88). The eruption continued
for four years, at the end of which time the height of the summit was
estimated by the _Challenger_ expedition to be 1900 feet. At the time
of the first ascent in 1905, the height was determined by aneroid as
1750 feet, with sharp rock pinnacles projecting some 50 or 75 feet

=Active and extinct volcanoes.=—The terms “active” and “extinct”
have come into more or less common use to describe respectively those
volcanoes which show signs of eruptive activity, and those which are
not at the time active. The term “dormant” is applied to volcanoes
recently active and supposed to be in a doubtfully extinct condition.
From a well-known volcano in the vicinity of Naples, volcanoes which no
longer erupt lava or cinder, but show gaseous emanations (_fumeroles_)
are said to be in the _solfatara_ condition, or to show _solfataric_

Experience shows that the term “extinct”, while useful, must always be
interpreted to mean apparently extinct. This may be illustrated by the
history of Mount Vesuvius, which before the Christian era was forested
in the crater and showed no signs of activity; and in fact it is
known that for several centuries no eruption of the volcano had taken
place. Following a premonitory earthquake felt in the year 63, the
mountain burst out in grand explosive eruption in 79 A.D. This eruption
profoundly altered the aspect of the mountain and buried the cities
of Pompeii, Stabeii, and Herculaneum from sight. Once more, this time
during the middle ages, for nearly five centuries (1139 to 1631) there
was complete inactivity, if we except a light ash eruption in the year
1500. During this period of rest the crater was again forested, but the
repose was suddenly terminated by one of the grandest eruptions in the
mountain’s history.

[Illustration: FIG. 89.—Map showing the location of the belts of
active volcanoes.]

=The earth’s volcano belts.=—The distribution of volcanoes is not
uniform, but, on the contrary, volcanic vents appear in definite zones
or belts, either upon the margins of the continents or included within
the oceanic areas (Fig. 89). The most important of these belts girdles
the Pacific Ocean, and is represented either by chains or by more
widely spaced volcanic mountains throughout the Cordilleran Mountain
system of South and Central America and Mexico, by the volcanoes of the
Coast and Cascade ranges of North America, the festooned volcanic chain
of the Aleutian Islands, and the similar island arcs off the eastern
coast of the Eurasian continent. The belt is further continued through
the islands of Malaysia to New Zealand, and on the Pacific’s southern
margin are found the volcanoes of Victoria Land, King Edward Land, and
West Antarctica.

[Illustration: FIG. 90.—A portion of the “fire girdle” of the Pacific,
showing the relation of the chains of volcanic mountains to the deeps
of the neighboring ocean floor.]

This volcano girdle is by no means a perfect one, for in addition
to the principal festoons of the western border there are many
secondary ones, and still other arcs are found well toward the center
of the oceanic area. Another broad belt of volcanoes borders the
Mediterranean Sea, and is extended westward into the Atlantic Ocean.
Narrower belts are found in both the northern and southern portions
of the Atlantic Ocean, on the margins of the Caribbean Sea, etc. The
fact of greatest significance in the distribution seems to be that
bands of active volcanoes are to be found wherever mountain ranges are
paralleled by deeps on the neighboring ocean floor (Fig. 90). As has
been already pointed out in the chapter upon earthquakes, it is just
such places as these which are the seat of earthquakes; these are zones
of the earth’s crust which are undergoing the most rapid changes of
level at the present time. Thus the rise of the land in mountains is
proceeding simultaneously with the sinking of the sea floor to form the
neighboring deeps.

[Illustration: FIG. 91.—Volcanic cones formed in 1783 above the
Skaptár fissure in Iceland (after Helland).]


FIG. 92.—Diagrams to illustrate the location of volcanic vents upon
fissure lines, _a_, openings caused by lateral movement of fissure
walls; _b_, openings formed at fissure intersections.]

=Arrangement of volcanic vents along fissures and especially at their
intersections.=—Within those districts in which volcanoes are widely
separated from their neighbors, the law of their arrangement is
difficult to decipher, but the view that volcanic vents are aligned
over fissures is now supported by so much evidence that illustrations
may be supplied from many regions. An exceptionally perfect line of
small cones is found along the Skaptár cleft in Iceland, upon which
stands the large volcano of Laki. This fissure reopened in 1783, and
great volumes of lava were exuded. Over the cleft there was left a long
line of volcanic cones (Fig. 91). There are in Iceland two dominating
series of parallel fissures of the same character which take their
directions respectively northeast-southwest and north-south. Many such
fissures are traceable at the surface as deep and nearly straight
clefts or _gjás_, usually a few yards in width, but extending for many
miles. The Eldgjá has a length of more than 18 English miles and a
depth varying from 400 to 600 feet. On some of these fissures no lava
has risen to the surface, whereas others have at numerous points exuded
molten rock. Sometimes one end only of a fissure, the more widely
gaping portion, has supplied the conduits for the molten lava. This
is well illustrated by the cratered monticules raised by the common
ant over the cracks which separate the blocks of cement sidewalk, the
hillocks being located where the most favorable channel was found for
the elevation of the materials.


FIG. 93.—Outline map of the eastern portion of the island of Java,
displaying the arrangement of volcanic vents in alignment upon fissures
with the larger mountains at fissure intersections (after Verbeek).]

Those places upon fissures which become lava conduits appear to be the
ones where the cleft gapes widest so as to furnish the widest channel.
Wherever a differential lateral movement of the walls has occurred,
openings will be found in the neighborhood of each minor variation from
a straight line (Fig. 92_a_). Wherever there are two or more series
of fissures, and this would appear to be the normal condition, places
favorable for lava conduits occur at fissure intersections. Within such
veritable volcano gardens as are to be found in Malaysia, the law of
volcano distribution became apparent so soon as accurate maps had been
prepared. Thus the outline map of a portion of the island of Java (Fig.
93) shows us that while the volcanoes of the island present at first
sight a more or less irregular band or zone, there are a number of
fissures intersecting in a network, and that the volcanoes are aligned
upon the fissures with the larger cones located at the intersections.
So also in Iceland, the great eruption of Askja in 1875 occurred at
the intersection of two lines of fissure.

Outside these closely packed volcanic regions, similar though less
marked networks are indicated; as, for example, in and near the Gulf of
Guinea. If now, instead of reducing the scale of our volcano maps, we
increase it, the same law of distribution is no less clearly brought
out. The monticules or small volcanic cones which form upon the flanks
of larger volcanic mountains are likewise built up over fissures which
on numerous occasions have been observed to open and the cones to form
upon them.


FIG. 94.—Map of the Puy Pariou in the Auvergne of central France. The
seat of eruption has migrated along the fissure upon which the earlier
cone had been built up (after Scrope).]

Still further reducing now the area of our studies and considering for
the moment the “frozen” surface of the boiling lava within the caldron
of Kilauea, this when observed at night reveals in great perfection
the sudden formation of fissures in the crust with the appearance
of miniature volcanoes rising successively at more or less regular
intervals along them.

It not infrequently happens that after a volcanic vent has become
established above some conduit in a fissure, the conduit migrates along
the fissure, thus establishing a new cone with more or less complete
destruction of the old one (Fig. 94).

=The so-called fissure eruptions.=—The grandest of all volcanic
eruptions have been those in which the entire length and breadth of
the fissures have been the passageway for the upwelling lava. Such
grander eruptions have been for the most part prehistoric, and in later
geologic history have occurred chiefly in India, in Abyssinia, in
northwestern Europe, and in the northwestern United States. In western
India the singularly horizontal plateaus of basaltic lava, the Dekkan
traps, cover some 200,000 square miles and are more than a mile in
depth. The underlying basement where it appears about the margins of
the basalt is in many places intersected by dikes or fissure fillings
of the same material. No cones or definite vents have been found.


FIG. 95.—Basaltic plateau of the northwestern United States due to
fissure eruptions of lava.]

The larger portion of the northwestern British Isles would appear to
have been at one time similarly blanketed by nearly horizontal beds of
basaltic lava, which beds extended northwestward across the sea through
the Orkney and Faroe islands to Iceland. Remnants of this vast plateau
are to-day found in all the island groups as well as in large areas of
northeastern Ireland, and fissure fillings of the same material occur
throughout large areas of the British Isles. In many cases these dikes
represent once molten rock which may never have communicated with the
surface at the time of the lava outpouring, yet they well illustrate
what we might expect to find if the basalt sheets of Iceland or Ireland
were to be removed.

The floods of basaltic lava which in the northwestern United States
have yielded the barren plateau of the Cascade Mountains (Fig. 95)
would appear to offer another example of fissure eruption, though cones
appear upon the surface and perhaps indicate the position of lava
outlets during the later phases of the eruptive period. The barrenness
and desolation of these lava plains is suggested by Fig. 96.

[Illustration: FIG. 96.—Lava plains about the Snake River in Idaho.]

Though the greater effusions of lava have occurred in prehistoric
times, and the manner of extrusion has necessarily been largely
inferred from the immense volume of the exuded materials and the
existence of basaltic dikes in neighboring regions, yet in Iceland
we are able to observe the connection between the dikes and the lava
outflows. Professor Thoroddsen has stated that in the great basaltic
plateau of Iceland, lava has welled out quietly from the whole length
of fissures and often on both sides without giving rise to the
formation of cones. At three wider portions of the great Eld cleft,
lava welled out quietly without the formation of cones, though here in
the southern prolongation of the fissure, where it was narrower, a row
of low slag cones appeared. Where the lava outwellings occurred, an
area of 270 square miles was flooded.

[Illustration: FIG. 97.—Characteristic profiles of lava volcanoes. 1,
basaltic lava mountain; 2, mountain of siliceous lava (after Judd).]

=The composition and the properties of lava.=—In our study of igneous
rocks (Chapter IV) it was learned that they are composed for the most
part of silicate minerals, and that in their chemical composition they
represent various proportions of silica, alumina, iron, magnesia,
lime, potash, and soda. The more abundant of these constituents is
silica, which varies from 35 to 70 per cent of the whole. Whenever the
content of silica is relatively low,—basic or basaltic lava,—the
cooled rock is dark in color and relatively heavy. It melts at a
relatively low temperature, and is in consequence relatively fluid
at the temperatures which lavas usually have on reaching the earth’s
surface. Furthermore, from being more fluid, the water which is nearly
always present in large quantity within the lava more readily makes its
escape upon reaching the surface. Eruptions of such lava are for this
reason without the violent aspects which belong to extrusions of more
siliceous (more “acidic”) lavas. For the same reason, also, basaltic
lava flows more freely and can spread much farther before it has
cooled sufficiently to consolidate. This is equivalent to saying that
its surface will assume a flatter angle of slope, which in the case
of basaltic lava seldom exceeds ten degrees and may be less than one
degree (Fig. 97).

[Illustration: FIG. 98.—A driblet cone (after J. D. Dana).]

Siliceous lavas, on the other hand, are, when consolidated,
relatively light both in color and weight and melt at relatively high
temperatures. They are, therefore, usually but partly fused and of a
viscous consistency when they arrive at the earth’s surface. Because
of this viscosity they offer much resistance to the liberation of the
contained water, which therefore is released only to the accompaniment
of more or less violent explosions. The lava is blown into the air
and usually falls as consolidated fragments of various degrees of


FIG. 99.—View of Leffingwell crater, a cinder cone in the Owens
valley, California (after an unpublished photograph by W. D. Johnson).]

It must not, however, be assumed that the temperature of lava is always
the same when it arrives at the surface, and hence it may happen that
a siliceous lava is exuded at so high a temperature that it behaves
like a normal basaltic lava. On the other hand, basaltic lavas may be
extruded at unusually low temperatures, in which case their behavior
may resemble that of the normal siliceous lavas. If, however, as is
generally the case, the energy of explosion of a basaltic lava is
relatively small, any ejected portions of the liquid lava travel to
a moderate height only in the air, so that on falling they are still
sufficiently pasty to adhere to rock surfaces and thus build up the
remarkably steep cones and spines known as “spatter cones” or “driblet
cones” (Fig. 98). When, on the other hand, the energy of explosion is
great, as is normally the case with siliceous lavas, the portions of
ejected lava have been fully consolidated before their fall to the
surface, so that they build up the same type of accumulation as would
sand falling in the same manner. The structures which they form are
known as tuff, cinder, or ash cones (Fig. 99).

Whenever the contained water passes off from siliceous lavas without
violent explosions, the lava may flow from the vent, but in contrast to
basaltic lavas it travels a short distance only before consolidating.
The resulting mountain is in consequence proportionately high and steep
(Fig. 97). Eruptions characterized by violent explosions accompanied by
a fall of cinder are described as _explosive_ eruptions. Those which
are relatively quiet, and in which the chief product is in the form of
streams of flowing lava, are spoken of as _convulsive_ eruptions.

=The three main types of volcanic mountain.=—If the eruptions at a
volcanic vent are exclusively of the explosive type, the material
of the mountain which results is throughout tuff or cinder, and the
volcano is described as a _cinder cone_. If, on the other hand, the
vent at every eruption exudes lava, a mountain of solid rock results
which is a _lava dome_. It is, however, the exception for a volcano
which has a long history to manifest but a single kind of eruption. At
one time exuding lava comparatively quietly, at another the violence
with which the steam is liberated yields only cinder, and the mountain
is a composite of the two materials and is known as a _composite
volcanic cone_.

=The lava dome.=—When successive lava flows come from a crater, the
structure which results has the form of a more or less perfect dome. If
the lava be of the basaltic or fluid type, the slopes are flat, seldom
making an angle of as much as ten degrees with the horizon and flatter
toward the summit (Fig. 101, p. 106). If of siliceous or viscous lava,
on the other hand, the slopes are correspondingly steep and in some
cases precipitous. To this latter class belong some of the _Kuppen_ of
Germany, the _puys_ of central France, and the _mamelons_ of the Island
of Bourbon.


FIG. 100.—Map of Hawaii and the lava volcanoes of Mokuaweoweo (Mauna
Loa) and Kilauea (after the government map by Alexander).]

=The basaltic lava domes of Hawaii.=—At the “crossroads of the
Pacific” rises a double line of lava volcanoes which reach from 20,000
to 30,000 feet above the floor of the ocean, some of them among
the grandest volcanic mountains that are known. More than half the
height and a much larger proportion of the bulk of the largest of
these are hidden beneath the ocean’s surface. The two great active
vents are Mokuaweoweo (on Mauna Loa) and Kilauea, distinct volcanoes
notwithstanding the fact that their lava extravasations have been
merged in a single mass. The rim of the crater of Mauna Loa is at an
elevation of 13,675 feet above the sea, whereas that of Kilauea is
less than 4000 feet and appears to rest upon the flank of the larger
mountain (Figs. 100 and 101). Although one crater is but 20 miles
distant from the other and nearly 10,000 feet lower, their eruptions
have apparently been unsympathetic. Nowhere have still active lava
mountains been subjected to such frequent observations extending
throughout a long period, and the dynamics of their eruptions are
fairly well understood. To put this before the reader, it will be
best to consider both mountains, for though they have much in common,
the observations from one are strangely complementary to those of the
other. The lower crater being easily accessible, Kilauea has been often
visited, and there exists a long series of more or less consecutive
observations upon it, which have been assembled and studied by Dana and
Hitchcock. The place of outflow of the Kilauea lavas has not generally
been visible, whereas Mokuaweoweo has slopes rising nearly 14,000 feet
above the sea and displays the records of outflow of many eruptions,
some of which were accompanied by the grandest of volcanic phenomena.

[Illustration: FIG. 101.—Section through Mauna Loa and Kilauea.]

=Lava movements within the caldron of Kilauea.=—The craters of these
mountains are the largest of active ones, each being in excess of
seven miles in circumference. In shape they are irregularly elliptical
and consist of a series of steps or terraces descending to a pit at the
bottom, in which are open lakes of boiling lava. Enough is known of the
history of Kilauea to state that the steep cliffs bounding the terraces
are fault walls produced by inbreak of a frozen lava surface. The cliff
below the so-called “black ledge” was produced by the falling in of
the frozen lava surface at the time of the outflow of 1840, the lava
issuing upon the eastern flank of the mountain and pouring into the sea
near Nanawale. Since that date the floor of the pit below the level of
this ledge has been essentially a movable platform of frozen lava of
unknown and doubtless variable thickness which has risen and descended
like the floor of an elevator car between its guiding ways (Fig. 102).
The floor has, however, never been complete, for one or more open lakes
are always to be seen, that of Halemaumau located near the southwestern
margin having been much the most persistent. Within the open lakes the
boiling lava is apparently white hot at the depth of but a few inches
below the surface, and in the overturnings of the mass these hotter
portions are brought to the surface and appear as white streaks marking
the redder surface portions. From time to time the surface freezes
over, then cracks open and erupt at favored points along the fissures,
sending up jets and fountains of lava, the material of which falls in
pasty fragments that build up driblet cones. Small fluid clots are
shot out, carrying a threadlike line of lava glass behind them, the
well-known “Pelé’s hair.” Sometimes the open lakes build up congealed
walls, rising above the general level of the pit, and from their rim
the lava spills over in cascades to spread out upon the frozen floor,
thus increasing its thickness from above (Fig. 103). At other times
a great dome of lava has been pushed up from the pit of Halemaumau
under a frozen shell, the molten lava shining red through cracks in
its surface and exuding so as to heal each widely opened fissure as it


FIG. 102.—Schematic diagram to illustrate the moving platform of
frozen lava which rises and falls in the crater of Kilauea.]

At intervals of from a few years to nine or ten years the crater has
been periodically drained, at which times the moving platform of
frozen lava has sunk more or less rapidly to levels far below the black
ledge and from 900 to 1700 feet below the crater rim. Following this
descent a slow progressive rise is inaugurated, which has sometimes
gone on at a rate of more than a hundred feet per year, though it is
usually much slower than this. When the platform has reached a height
varying from 700 to 350 feet below the crater rim, another sudden
settlement occurs which again carries the pit floor downward a distance
of from 300 to 700 feet.


FIG. 103.—View of the open lava lake of Halemaumau within the crater
of Kilauea, the molten lava shown cascading over the raised lava walls
on to the floor of the pit (after Pavlow).]

=The draining of the lava caldrons.=—The changes which go on within
the crater of Mokuaweoweo, though less studied than those of Kilauea,
appear to be in some respects different. Here every eruption seems
to be preceded by a more or less rapid influx of melted lava to the
pit of the crater, this phenomenon being observed from a distance as
a brilliant light above the crater—the reflection of the glow from
overhanging vapor clouds. The uprising of the lava has often been
accompanied by the formation of high lava fountains upon the surface,
and the molten lava sometimes appears in fissures near the crater rim
at levels well above the lava surface within the pit.

Although in many cases the lava which has thus flooded the crater has
suddenly drained away without again becoming visible, it is probable
that in such cases an outlet has been found to some submarine exit,
since under-ocean discharge effects have been observed in connection
with eruptions of each of the volcanoes.


FIG. 104.—Map showing the manner of outflow of lava from Kilauea
during the eruption of 1840. The outflowing lava made its appearance
successively at the points _A_, _B_, _C_, _m_, _n_, and finally at a
point below _n_, from whence it issued in volume and flowed down to the
sea at Nanawale (after J. D. Dana).]

Inasmuch as no earthquakes are felt in connection with such outflows
as have been described, it is probable that the hot lava fuses a
passageway for itself into some open channel underneath the flanks
of the mountain. Such a course is well illustrated by the outflow
of Kilauea in 1840, when, it will be remembered, occurred the great
down-plunge of the crater that yielded the pit below the black ledge.
At this time the lava first made its appearance upon the flanks of the
mountain at the bottom of a small pit or inbreak crater which opened
five miles southeast of the main crater of Kilauea (Fig. 104). Within
this new crater the lava rose, and small ejections soon followed from
fissures formed in its neighborhood. Some time after, the lava sank
in the first new crater, only to reappear successively at other small
openings (Fig. 104, _B_, _C_, _m_, _n_) and finally to issue in volume
at a point eleven miles from the shore and flow thereafter _upon
the surface_ of the mountain until it had reached the sea. Only the
slightest earth tremors were felt, and as no rumblings were heard, it
is evident that the lava fused its way along a buried channel largely
open at the time (see below, p. 112).

In a majority of the eruptions of Mokuaweoweo, when the outflowing
lavas have become visible, the molten rock has apparently fused its way
out to the surface of the mountain at points from 1000 to 3000 feet
below the bottom of the crater, and this discharge has corresponded
in time to the lowering of the lava surface within the crater. There
are, however, three instances upon record in which the lava issued
from definite rents which were formed upon the mountain flanks at
comparatively low levels. In contrast to the formation of fused
outlets, these ruptures of a portion of the mountain’s flank were
always accompanied by vigorous local earthquakes of short duration. In
one instance (the eruption of 1851) such a rent appeared under the same
conditions but at an elevation of 12,500 feet, or near the level of the
lava in the crater.

=The outflow of the lava floods.=—In order to properly comprehend
these and many otherwise puzzling phenomena connected with
volcanoes, it is necessary to keep ever in mind the quite remarkable
heat-insulating property of congealed lava. So soon as a thin crust
has formed upon the surface of molten rock, the heat of the underlying
fluid mass is given off with extreme slowness, so that lava streams no
longer connected with their internal lava reservoirs may remain molten
for decades.


FIG. 105.—Lava of Matavanu upon the Island of Savaii flowing down to
the sea during the eruption of 1906. The course may be followed by the
jets of steam escaping from the surface down to the great steam cloud
which rises where the fluid lava discharges into the sea (after H. I.

We have seen that for Mokuaweoweo and Kilauea, lava either quietly
melts its way to the surface at the time of outflow, or else produces a
rent for its egress to the accompaniment of vigorous local earthquakes.
In either case if the lava issues at a point far below the crater,
gigantic lava fountains arise at the point of outflow, the fluid rock
shooting up to heights which range from 250 to 600 or more feet above
the surface. A certain proportion of this fluid lava is sufficiently
cooled to consolidate while traveling in the air, and falling, it
builds up a cinder cone which is left as a location monument for
the place of discharge. From this outlet the molten lava begins its
journey down the slope of the mountain, and quickly freezes over to
produce a tunnel, beneath the roof of which the fluid lava flows with
comparatively slow further loss of heat. Save for occasional steam jets
issuing from its surface, it may give little indication of its presence
until it has reached the sea (Fig. 105).

[Illustration: FIG. 106.—Lava stream discharging into the sea from
beneath the frozen roof of a lava tunnel. Eruption of Matavanu on
Savaii in 1906 (after Sapper).]

If sufficient in volume and the shore be not too distant, the stream
of lava arrives at the sea, where, discharging from the mouth of its
tunnel, it throws up vast volumes of steam and induces ebullition of
the water over a wide area (Fig. 106). Professor Dana, who visited
Hawaii a few months only after the great outflow of 1840, states that
the lava, upon reaching the ocean, was shivered like melted glass and
thrown up in millions of particles which darkened the sky and fell like
hail over the surrounding country. The light was so bright that at a
distance of forty miles fine print could be read at midnight.


FIG. 107.—Diagrammatic representation of the structure of the flanks
of lava volcanoes as a result of the draining of frozen lava streams.]

Protected from any extensive consolidation by its congealed cover, the
lava within a stream may all drain away, leaving behind an empty lava
tunnel, which in the case of the Hawaiian volcanoes sometimes has its
roof hung with beautiful lava stalactites and its floor studded with
thin lava spines. Later lava outflows over the same or neighboring
courses bury such tunnels beneath others of similar nature, giving
to the mountain flanks an elongated cellular structure illustrated
schematically in Fig. 107. These buried channels may in the future be
again utilized for outflows similar in character to that of Kilauea in


FIG. 108.—Diagram to show the manner of formation of mesas or table
mountains by the outflow of lava in valleys and the subsequent more
rapid erosion of the intervening ridges. _R_, earlier river valley;
_R’R’_, later valleys.]

While the formation of lava stalactites of such perfection and beauty
is peculiar to the Hawaiian lava tunnels, the formation of the tunnel
in connection with lava outflow is the rule wherever a dissipation at
the end has permitted of drainage. A few hours only after the flow has
begun, the frozen surface has usually a thickness of a few inches, and
this cover may be walked over with the lava still molten below. At
first in part supported by the molten lava, the tunnel roof sometimes
caves in so soon as drainage has occurred.

[Illustration: FIG. 109.—Surface of lava of the Pahoehoe type.]

Wherever basaltic lava has spread out in valleys on the surface of
more easily eroded material, either cinder or sedimentary formations,
the softer intervening ridges are first carried away by the eroding
agencies, leaving the lava as cappings upon residual elevations.
Thus are derived a type of table mountain or _mesa_ of the sort well
illustrated upon the western slopes of the Sierra Nevadas in California
(Fig. 108).


FIG. 110.—Three successive views to illustrate the growth of the
Island of Savaii from the outflow of lava at Matavanu in the year 1906.
_a_, near the beginning of the outflow; _b_, some weeks later than _a_;
_c_, some weeks later than _b_ (after H. I. Jensen).]

The surface which flowing lava assumes, while subject to considerable
variation, may yet be classified into two rather distinct types. On
the one hand there is the billowy surface in which ellipsoidal or
kidney-shaped masses, each with dimensions of from one to several feet,
lie merged in one another, not unlike an irregular collection of sofa
pillows. This type of lava has become known as the _Pahoehoe_, from
the Hawaiian occurrence (Fig. 109). A variation from this type is the
“corded” or “ropy” lava, the surface of which much resembles rope as it
is coiled along the deck of a vessel, the coils being here the lines of
scum or scoriæ arranged in this manner by the currents at the surface
of the stream (Fig. 123, p. 124). A quite different type is the block
lava (_Aa_ type) which usually has a ragged scoriaceous surface and
consists of more or less separate fragments of cooled lava (Fig. 131,
p. 130).

Wherever lava flows into the sea in quantity, it extends the margin of
the shore, often by considerable areas. The outflow of Kilauea in 1840
extended the shore of Hawaii outward for the distance of a quarter of a
mile, and a more recent illustration of such extension of land masses
is furnished by Fig. 110.




=The mechanics of crater explosions.=—If we now turn from the lava
volcano to the active cinder cone, we encounter an entire change of
scene. In place of the quiet flow and convulsive movements of the
molten lava, we here meet with repeated explosions of greater or less
violence. If we are to profitably study the manner of the explosions,
considering the volcanic vent as a great experimental apparatus, it
would be well to select for our purpose a volcano which is in a not too
violent mood. The well-known cinder cone of Stromboli in the Eolian
group of islands north of Sicily has, with short and unimportant
interruptions, remained in a state of light explosive activity since
the beginning of the Christian era. Rising as it does some three
thousand feet directly out of the Mediterranean, and displaying by day
a white steam cap and an intermittent glow by night, its summit can be
seen for a distance of a hundred miles at sea and it has justly been
called the “Lighthouse of the Mediterranean.” The “flash” interval
of this beacon may vary from one to twenty minutes, and it may show,
furthermore, considerable variation of intensity.

For the reason that the crater of the mountain is located at one side
and at a considerable distance below the actual summit, the opportunity
here afforded of looking into the crater is most favorable whenever
the direction of the wind is such as to push aside the overhanging
steam cloud (Fig. 111). Long ago the Italian vulcanologist Spallanzani
undertook to make observations from above the crater, and many others
since his day have profited by his example.

Within the crater of the volcano there is seen a lava surface lightly
frozen over and traversed by many cracks from which vapor jets
are issuing. Here, as in the Kilauea crater, there are open pools
of boiling lava. From some of these, lava is seen welling out to
overflow the frozen surface; from others, steam is ejected in puffs
as though from the stack of a locomotive. Within others lava is seen
heaving up and down in violent ebullition, and at intervals a great
bubble of steam is ejected with explosive violence, carrying up with
it a considerable quantity of the still molten lava, together with
its scumlike surface, to fall outside the crater and rattle down
the mountain’s slope into the sea. Following this explosion the
lava surface in the pool is lowered and the agitation is renewed,
to culminate after the further lapse of a few minutes in a second
explosion of the same nature. The rise of the lava which precedes the
ejection appears at night as a brighter reflection or glow from the
overhanging steam cloud—the flash seen by the mariner from his vessel.


FIG. 111.—The volcano of Stromboli, showing the excentric position of
the crater (after a sketch by Judd).]

What is going on within the crater of Stromboli we may perhaps best
illustrate by the boiling of a stiff porridge over a hot fire. Any one
who has made corn mush over a hot camp fire is fully aware that in
proportion as the mush becomes thicker by the addition of the meal, it
is necessary to stir the mass with redoubled vigor if anything is to be
retained within the kettle. The thickening of the mush increases its
viscosity to such an extent that the steam which is generated within it
is unable to make its escape unless aided by openings continually made
for it by the stirring spoon. If the stirring motion be stopped for a
moment, the steam expands to form great bubbles which soon eject the
pasty mass from the kettle.

For the crater of Stromboli this process is illustrated by the series
of diagrams in Fig. 112. As the lava rises toward the surface,
presumably as a result of convectional currents within the chimney of
the volcano, the contained steam is relieved from pressure, so that
at some depth below the surface it begins to separate out in minute
vesicles or bubbles, which, expanding as they rise, acquire a rapidly
accelerating velocity. Soon they flow together with a quite sudden
increase of their expansive energy, and now shooting upward with
further accelerated velocity, a layer of liquid lava with its cover of
scum is raised on the surface of a gigantic bubble and thrown high into
the air. Cooled during their flight, the quickly congealed lava masses
become the tuff or volcanic ash which is the material of the cinder

[Illustration: FIG. 112.—Diagrams to illustrate the nature of
eruptions within the crater of Stromboli.]

=Grander volcanic eruptions of cinder cones.=—Most cinder and
composite cones, in the intervals between their grander eruptions, if
not entirely quiescent, lapse into a period, of light activity during
which their crater eruptions appear to be in all essential respects
like the habitual explosions within the Strombolian crater. This phase
of activity is, therefore, described as _Strombolian_. By contrast,
the occasional grander eruptions which have punctuated the history
of all larger volcanoes are described in the language of Mercalli as
_Vulcanian_ eruptions, from the best studied example.

Just what it is that at intervals brings on the grander Vulcanian
outburst within a volcano is not known with certainty; but it is
important to note that there is an approach to periodicity in the
grander eruptions. It is generally possible to distinguish eruptions of
at least two orders of intensity greater than the Strombolian phase;
a grander one, the examples of which may be separated by centuries,
and one or more orders of relatively moderate intensity which recur
at intervals perhaps of decades, their time intervals subdividing the
larger periods marked off by the eruptions of the first order.


FIG. 113.—Map of Volcano in the Eolian group of islands. The smaller
craters partially dissected by the waves belong to Vulcanello (after

=The eruption of Volcano in 1888.=—In the Eolian Islands to the north
of Sicily was located the mythical forge of Vulcan. From this locality
has come our word “volcano”, and both the island and the mountain bear
no other name to-day (Fig. 113). There is in the structure of the
island the record of a somewhat complex volcanic history, but the form
of the large central cinder cone was, according to Scrope, acquired
during the eruption of 1786, at which time the crater is reported to
have vomited ash for a period of fifteen days. Passing after this
eruption into the solfatara condition, with the exception of a light
eruption in 1873, the volcano remained quiet until 1886. So active
had been the fumeroles within the crater during the latter part of
this period that an extensive plant had been established there for
the collection especially of boracic acid. In 1886 occurred a slight
eruption, sufficient to clear out the bottom of the crater, though
not seriously to disturb the English planter whose vineyards and fig
orchards were in the valley or _atrio_ near the point _d_ upon the map
(Fig. 113), nearly a mile from the crater rim. On the 3d of August,
1888, came the opening discharge of an eruption, which, while not of
the first order of magnitude, was yet the greatest in more than a
century of the mountain’s history, and may serve us to illustrate the
Vulcanian phase of activity within a cinder cone. During the day, to
the accompaniment of explosions of considerable violence, projectiles
fell outside the crater rim and rolled down the steep slopes toward the
_atrio_. These explosions were repeated at intervals of from twenty to
thirty minutes, each beginning in a great upward rush of steam and
ash, accompanied by a low rumbling sound. During the following night
the eruptions increased in violence, and the anxious planter remained
on watch in his villa a mile from the crater. Falling asleep toward
morning, he was rudely awakened by a rain of projectiles falling upon
his roof. Hastily snatching up his two children he ran toward the door
just as a red hot projectile, some two feet in diameter, descended
through the roof, ceiling, and floor of the drawing room, setting fire
to the building. A second projectile similar to the first was smashed
into fragments at his feet as he was emerging from the house, burning
one of the children. Making his escape to Vulcanello at the extremity
of the island, the remainder of the night and the following day, until
rescue came from Lipari, were spent just beyond the range of the
falling masses.


FIG. 114.—“Bread-crust” lava projectile from the eruption of Volcano
in 1888 (after Mercalli).]

When the writer visited the island some months later, the eruption was
still so vigorous that the crater could not be reached. The ruined
villa, smashed and charred, stood with its walls half buried in ash and
lapilli, among which were partly smashed pumiceous lava projectiles.
The entire _atrio_ about the mountain lay buried in cinder to the depth
of several feet and was strewn with projectiles which varied in size
from a man’s fist to several feet in diameter (Fig. 114). The larger of
these exhibited the peculiar “bread-crust” surface and had generally
been smashed by the force of their fall after the manner of a pumpkin
which has been thrown hard against the ground. One of these projectiles
fully three feet in diameter was found at the distance of a mile and
a half from the crater. Though diminished considerably in intensity,
the rhythmic explosions within the crater still recurred at intervals
varying from four minutes to half an hour, and were accompanied
by a dull roar easily heard at Lipari on a neighboring island six
miles away. Simultaneously, a dark cloud of “smoke”, the peculiar
“cauliflower cloud” or _pino_ mounted for a couple of miles above the
crater (Fig. 115), and the rise was succeeded by a rain of small lava
fragments or _lapilli_ outside the crater rim.


FIG. 115.—Peculiar “cauliflower cloud” or _pino_ composed of steam and
ash, rising above the cinder cone of Volcano during the waning phases
of the explosive eruption of 1888 (after a photograph by B. Hobson).]

There seems to be no good reason to doubt that Vulcanian cinder
eruptions of this type differ chiefly in magnitude from the rhythmic
explosion within the crater of Stromboli, if we except the elevation of
a considerable quantity of accessory and older tuff which is derived
from the inner walls of the crater and carried upward into the air
together with the pasty cakes of fresh lava derived from the chimney.
It is this accessory material which gives to the _pino_ its dark or
even black appearance.


FIG. 116.—Double explosive eruption of Taal volcano on the morning of
January 30, 1911.]

=The eruption of Taal volcano on January 30, 1911.=—The recent
eruption of the cinder cone known as Taal volcano is of interest,
not only because so fresh in mind, but because two neighboring vents
erupted simultaneously with explosions of nearly equal violence (Fig.
116). This Philippine volcano lies near the center of a lake some
fifteen miles in diameter and about fifty miles south of the city of
Manila. After a period of rest extending over one hundred and fifty
years, the symptoms of the coming eruption developed rapidly, and on
the morning of January 30 grand explosions of steam and ash occurred
simultaneously in the neighboring craters, and the condensed moisture
brought down the ash in an avalanche of scalding mud which buried the
entire island. Almost the entire population of the island, numbering
several hundreds, was literally buried in the blistering mud (Fig.
117); and the gases from the explosions carried to the distant shores
of the lake added to this number many hundred victims.


FIG. 117.—The thick mud veneer upon the island of Taal (after a
photograph by Deniston).]

[Illustration: FIG. 118.—A pear-shaped lava projectile.]

The shocks which accompanied the explosions raised a great wave upon
the surface of the lake, which, advancing upon the shores, washed away
structures for a distance of nearly a half mile.

=The materials and the structure of cinder cones.=—Obviously the
materials which compose cinder cones are the cooled lava fragments of
various degrees of coarseness which have been ejected from the crater.
If larger than a finger joint, such fragments are referred to as
_volcanic projectiles_, or, incorrectly, as “volcanic bombs.” Of the
larger masses it is often true that the force of expulsion has not been
applied opposite the center of mass of the body. Thus it follows that
they undergo complex whirling motions during their flight, and being
still semiliquid, they develop curious pear-shaped or less regular
forms (Fig. 118). When crystals have already separated out in the lava
before its rise in the chimney of the volcano, the surrounding fluid
lava may be blown to finely divided volcanic dust which floats away
upon the wind, thus leaving the crystals intact to descend as a crystal
rain about the crater. Such a shower occurred in connection with the
eruption of Etna in 1669, and the black augite crystals may to-day be
gathered by the handful from the slopes of the Monti Rossi (Fig. 125,
p. 125).


FIG. 119.—Artificial production of the structure of a cinder cone with
use of colored sands carried up in alternation by a current of air
(after G. Linck).]

The term _lapilli_, or sometimes _rapilli_, is applied to the ejected
lava fragments when of the average size of a finger joint. This is the
material which still partially covers the unexhumed portions of the
city of Pompeii. Volcanic _sand_, _ash_, and _dust_ are terms applied
in order to increasingly fine particles of the ejected lava. The
finest material, the volcanic dust, is often carried for hundreds and
sometimes even for thousands of miles from the crater in the high-level
currents of the atmosphere. Inasmuch as this material is deposited far
from the crater and in layers more or less horizontal, such material
plays a small rôle in the formation of the cinder cone. The coarser
sands and ash, on the other hand, are the materials from which the
cinder cone is largely constructed.

The manner of formation and the structure of cinder cones may be
illustrated by use of a simple laboratory apparatus (Fig. 119). Through
an opening in a board, first white and then colored sand is sent up in
a light current of air or gas supplied from suitable apparatus. The
alternating layers of the sand form in the attitudes shown; that is
to say, dipping inward or toward the chimney of the volcano at all
points within the crater rim, and outward or away from it at all points
outside (Fig. 119). If the experiment is carried so far that at its
termination sand slides down the crater walls into the chimney below,
the inward dipping layers will be truncated, or even removed entirely,
as shown in Fig. 119 _b_.


FIG. 120.—Diagram to show the contrast between a lava dome and a
cinder cone. _AAA_, cinder cone; _BabC_, lava dome; _DE_, line of low
cinder cones above a fissure (after Thoroddsen).]

=The profile lines of cinder cones.=—The shapes of cinder cones are
notably different from those of lava mountains. While the latter are
domes, the mountains constructed of cinder are conical and have curves
of profile that are concave upward instead of convex (Fig. 120). In
the earlier stages of its growth the cinder cone has a crater which in
proportion to the height of the mountain is relatively broad (Fig. 99,
p. 104).


FIG. 121.—Mayon volcano on the island of Luzon, P.I. A remarkably
perfect high cinder cone.]

Speaking broadly, the diameter of the crater is a measure of the
violence of the explosions within the chimney. A single series of short
and violent explosive eruptions builds a low and broad cinder cone.
A long-continued succession of moderately violent explosions, on the
other hand, builds a high cone with crater diameter small if compared
with the mountain’s altitude, and the profile afforded is a remarkably
beautiful sweeping curve (Fig. 121). Toward the summit of such a cone
the loose materials of which it is composed are at as steep an angle as
they can lie, the so-called angle of repose of the material; whereas
lower down the flatter slopes have been determined by the distribution
of the cinder during its fall from the air. When one makes the ascent
of such a mountain, he encounters continually steeper grades, with the
most difficult slope just below the crest.


FIG. 122.—A series of breached cinder cones where the place of
eruption has migrated along the underlying fissure. The Puys Noir,
Solas, and La Vache in the Mont Dore Province of central France (after

=The composite cone.=—The life histories of volcanoes are generally
so varied that lava domes and the pure types of cinder cones are less
common than volcanoes in which paroxysmal eruptions have alternated
with explosions, and where, therefore, the structure of the mountain
represents a composite of lava and cinder. Such composite cones possess
a skeleton of solid rock upon which have been built up alternate
sloping layers of cinder and lava. In most respects such cones stand in
an intermediate position between lava domes and cinder cones.


FIG. 123.—The _bocca_ or mouth upon the inner cone of Mount Vesuvius
from which flowed the lava stream of 1872. This lava stream appears in
the foreground with its characteristic “ropy” surface.]

Regarded as a retaining wall for the lava which mounts in the chimney,
the cinder cone is obviously the weakest of all. Should lava rise in a
cinder cone without an explosion occurring, the cone is at once broken
through upon one side by the outwelling of the lava near the base. Thus
arises the characteristic _breached_ cone of horseshoe form (Fig. 122).


FIG. 124.—A row of parasitic cones raised above a fissure which was
opened upon the flanks of Mount Etna during the eruption of 1892 (after
De Lorenzo).]

Quite in contrast with the weak cinder cone is the lava dome with
its rock walls and relatively flat slopes. Considered as a retaining
wall for lava it is much the strongest type of volcanic mountain, and
it is likely that the hydrostatic pressure of the lava within the
crater would seldom suffice to rupture the walls, were it not that
the molten rock first fuses its way into old stream tunnels buried
under the mountain slopes (see _ante_, p. 112). Composite cones have
a strength as retaining walls for lava which is intermediate between
that of the other types. Their Vulcanian eruptions of the convulsive
type are initiated by the formation of a rent or fissure upon the
mountain flanks at elevations well above the base, the opening of the
fissure being generally accompanied by a local earthquake of greater or
less violence.

From one or more such fissures the lava issues usually with sufficient
violence at the place of outflow to build up over it either an enlarged
type of driblet cone, referred to as a “mouth”, or _bocca_[1] (Fig.
123), or one or more cinder cones which from their position upon the
flanks of the larger volcano are referred to as _parasitic cones_ (Fig.
124). The lava of Vesuvius more frequently yields _bocchi_ at the place
of outflow, whereas the flanks of Etna are pimpled with great numbers
of parasitic cinder cones, each the monument to some earlier eruption
(Fig. 125).


FIG. 125.—View looking toward the summit of Etna from a position upon
the southern flank near the village of Nicolosi. The two breached
parasitic cones seen behind this village are the Monti Rossi which were
thrown up in 1669 and from which flowed the lava which overran Catania
(after a photograph by Sommer).]

It is generally the case that a single eruption makes but a relatively
small contribution to the bulk of the mountain. From each new cone or
_bocca_ there proceeds a stream of lava spread in a relatively narrow
stream extending down the slopes (Fig. 126).


FIG. 126.—Sketch map of Etna, showing the individual surface lava
streams (in black) and the tuff covered surface (stippled).]

=The caldera of composite cones.=—Because of the varied episodes
in the history of composite cones, they lack the regular lines
characteristic of the two simpler types. The larger number of the more
important composite cones have been built up within an outer crater of
relatively large diameter, the Somma cone or _caldera_, which surrounds
them like a gigantic ruff or collar. This caldera is clearly in most
cases at least the relic of an earlier explosive crater, after which
successive eruptions of lesser violence have built a more sharply
conical structure. This can only be interpreted to mean that most
larger and long-active volcanoes have been born in the grandest throes
of their life history, and that a larger or smaller lateral migration
of the vent has been responsible for the partial destruction of the
explosion crater. Upon Vesuvius we find the crescent-like rim of Monte
Somma; on Etna it is the Val del Bove, etc. It is this caldera of
composite cones which gave rise to the theory of the “elevation crater”
of von Buch (see _ante_, p. 95, and Fig. 127).


FIG. 127.—Panum crater, showing the caldera and the later interior
cones (after Russell).]

=The eruption of Vesuvius in 1906.=—The volcano Vesuvius rises on the
shores of the beautiful bay of Naples only about ten miles distant
from the city of Naples. The mountain consists of the remnant of an
earlier broad-mouthed explosion crater, the Monte Somma, and an inner,
more conical elevation, the Monte Vesuvio. Before the eruption of 1906
this central cone was sharply conical and rose to a height of about
4300 feet above the surface of the bay, or above the highest point of
the ancient caldera. The base of this inner cone is at an elevation of
something less than half that of the entire mass, and is separated from
the encircling ring wall of the old crater by the _atrio_, to which
corresponds in height a perceptible shelf or _piano_ upon the slope
toward the bay of Naples (Fig. 128).


FIG. 128.—View of Mount Vesuvius as it appeared from the Bay of Naples
shortly before the eruption of 1906. The horn to the left is Monte

An active composite cone like that of Vesuvius is for the greater part
of the time in the Strombolian condition; that is to say, light crater
explosions continue with varying intensity and interval, except when
the mountain has been excited to the periodic Vulcanian outbreaks with
which its history has been punctuated. The Strombolian explosions
have sufficient violence to eject small fragments of hot lava, which,
falling about the crater, slowly built up a rather sharp cone. The
period of Strombolian activity has, therefore, been called the
_cone-producing period_. Just before each new outbreak of the Vulcanian
type, the altitude of the mountain has, therefore, reached a maximum,
and since the larger explosive eruptions remove portions of this cone
at the same time that they increase the dimensions of the crater,
the Vulcanian stage in contrast to the other has been called the
_crater-producing period_. In this period, then, the material ejected
during the explosions does not consist solely of fresh lava cakes,
but in part of the older débris derived from the crater walls, whence
it is avalanched upon the chimney after each larger explosion. The
overhanging cloud, which during the Strombolian period has consisted
largely of steam and is noticeably white, now assumes a darker tone,
the “smoke” which characterizes the Vulcanian eruption.


FIG. 129.—A series of consecutive sketches of the summit of the
Vesuvian cone, showing the modifications in its outline (after Sir
William Hamilton).]

On several historical occasions the cone of Vesuvius has been lowered
by several hundred feet, the greatest of relatively recent truncations
having occurred in 1822 and in 1906. Between Vulcanian eruptions the
Strombolian activity is by no means uniform, and so the upward growth
of the cone is subject to lesser interruptions and truncations (Fig.

The Vesuvian eruption of 1906 has been selected as a type of the
larger Vulcanian eruption of composite cones, because it combined the
explosive and paroxysmal elements, and because it has been observed and
studied with greater thoroughness than any other. The latest previous
eruption of the Vulcanian order had occurred in 1872. Some two years
later the period of active cone building began and proceeded with
such rapidity that by 1880 the new cone began to appear above the rim
of the crater of 1872. From this time on occasional light eruptions
interrupted the upbuilding process, and as the repairs were not in
all cases completed before a new interruption, a nest of cones, each
smaller than the last, arose in series like the outdrawn sections of
an old-time spyglass. At one time no less than five concentric craters
were to be seen.

For a brief period in the fall of 1904 Vesuvius had been in almost
absolute repose, but soon thereafter the Strombolian crater explosions
were resumed. On May 25, 1905, a small stream of lava began to issue
from a fissure high up upon the central cone, and from this time on
the lava continued to flow down to the valley or _atrio_, separating
the inner cone from the caldera remnant of Monte Somma. Seen in the
night, this stream of lava appeared from Naples like a red hot wire
laid against the mountain’s side (Fig. 130). With gradual augmentation
of Strombolian explosions and increase in volume of the flowing lava
stream, the same condition continued until the first days of April in
1906. The flowing lava had then overrun the tracks of the mountain
railway and accumulated in considerable quantity within the _atrio_
(Fig. 131).


FIG. 130.—Night view of Vesuvius from Naples before the outbreak of
1906. A small lava stream is seen descending from a high point upon the
central cone (after Mercalli).]

On the morning of April 4, a preliminary stage of the eruption was
inaugurated by the opening of a new radial fissure about 500 feet
below the summit of the cone (Fig. 132 _a_), and by early afternoon
the cone-destroying stage began with the rise of a dark “cauliflower
cloud” or _pino_ to replace the lighter colored steam cloud. The
cone was beginning to fall into the crater, and old lava débris was
mingled in the ejections with the lava clots blown from the still fluid
material within the chimney. From now on short and snappy lightning
flashes played about the black cloud, giving out a sharp staccato
“tack-a-tack.” The volume and density of the cloud and the intensity
of the crater explosions continued to increase until the culmination
on April 7. On April 5 at midnight a new lava mouth appeared upon
the same fissure which had opened near the summit, but now some 300
feet lower (Fig. 132 _b_). The lava now welled out in larger volume
corresponding to its greater head, and the stream which for ten months
had been flowing from the highest outlet upon the cone now ceased to
flow. The next morning, April 6, at about 8 o’clock, lava broke out at
several points some distance east of the opening _b_, and evidently
upon another fissure transverse to the first (Fig. 132 _c_). The lava
surface within the chimney must still have remained near its old
level,—effective draining had not yet begun,—since early upon the
following morning a small outflow began nearly at the top of the cone
upon the opposite side and at least a thousand feet higher.

[Illustration: FIG. 131.—Scoriaceous lava encroaching upon the tracks
of the Vesuvian railway (after a photograph by Sommer).]


FIG. 132.—Map of Vesuvius, showing the position and order of formation
of the lava mouths upon its flanks during the eruption of 1906 (after

The culmination of the eruption came in the evening of April 7, when,
to the accompaniment of light earthquakes felt as far as Naples, lava
issued for the first time in great volume from a mouth more than
halfway down the mountain side (Fig. 132 _f_), and thus began the
drainage of the chimney. At about the same time with loud detonations
a huge black cloud rose above the crater in connection with heavy
explosions, and a rain of cinder was general in the region about the
mountain but especially within the northeast quadrant. Those who were
so fortunate as to be in Pompeii had a clear view of the mountain’s
summit where red hot masses of lava were thrown far into the air. The
direction of these projections was reported to have been not directly
upward, but inclined toward the northeast quadrant of the mountain; but
since with a northeast surface wind the heaviest deposit of ash and
dust should have been upon the southwestern quadrant of the mountain,
it is evident that the material was carried upward until it reached the
contrary upper currents of the atmosphere, to be by them distributed.

[Illustration: FIG. 133.—The ash curtain which had overhung Vesuvius
lifting and disclosing the outlines of the mountain on April 10, 1911
(after De Lorenzo).]


FIG. 134.—The central cone of Vesuvius as it appeared after the
eruption of 1906, but with the earlier profile indicated. The
truncation represents a lowering of the summit by some five hundred
feet, with corresponding increase in the diameter of the crater (after

When the heavy curtain of ash, which now for a number of succeeding
days overhung all the circum-Vesuvian country, began to lift (Fig.
133), it was seen that the summit of the cone had been truncated
an average of some 500 feet (Fig. 134). All the slopes and much of
the surrounding country had the aspect of being buried beneath a
cocoa-colored snow of a depth to the northeastward of several feet,
where it had drifted into all the hollow ways so as almost to efface
them (Fig. 135). More than thrice as heavy as water, the weak roof
timbers of the houses at the base of the mountain gave way beneath the
added load upon them, thus making many victims. Inasmuch, however, as
the ash-fall partakes of the same general characters as in eruptions
from cinder cones, we may here give our attention especially to the
streams of lava which issued upon the opposite flank of the mountain
(Fig. 136).


FIG. 135.—A sunken road filled with indrifted cocoa-colored ash from
the Vesuvian eruption of 1906.]


FIG. 136.—View of Vesuvius taken from the southwest during the waning
stages of the eruption of 1906. In the middle distance may be discerned
the several lava mouths aligned upon a fissure, and the courses of the
streams which descend from them. In the foreground is the main lava
stream with scoriaceous surface (after W. Prinz).]

The main lava stream descended the first steep slopes with the velocity
of a mile in twenty-five minutes, about the strolling speed of a
pedestrian, but this rate was gradually reduced as the stream advanced
farther from the mouth. Taking advantage of each depression of the
surface, the black stream advanced slowly but relentlessly toward
the cities at the southwest base of the mountain. With a motion not
unlike that of a heap of coal falling over itself down a slope, the
block lava advances without burning the objects in its path (Fig.


FIG. 137.—The main lava stream of 1906 advancing upon the village of


FIG. 138.—An Italian pine snapped off by the lava and carried forward
upon its surface as a passenger (after Haug).]

The beautiful pines are merely charred where snapped off and are
carried forward upon the surface of the stream (Fig. 138). When a real
obstruction, such as a bridge or a villa, is encountered, the stream is
at first halted, but the rear crowding upon the van, unless a passage
is found at the side, the lava front rises higher and higher until by
its weight the obstruction is forced to give way (Figs. 139 and 140).


FIG. 139.—Lava front both pushing over and running around a wall which
lies athwart its course (after Johnston-Lavis).]


FIG. 140.—One of the villas in Boscotrecase which was ruined by the
Vesuvian lava flow of 1906. The fragments of masonry from the ruined
walls traveled upon the lava current, where they sometimes became
incased in lava.]

=The sequence of events within the chimney.=—The thorough study of
this Vesuvian eruption has placed us in a position to infer with
some confidence in our conclusions the sequence of events within the
chimney and crater of the volcano, both before and during the eruption.
Anticipating some conclusions derived from the observed dissection of
volcanoes, which will be discussed below, it may be stated that what
might be termed the core of the composite cone—the chimney—is a more
or less cylindrical plug of cooled lava which during the active period
of the vent has an interior bore of probably variable caliber. This
plug in its lower section appears in solid black in all the diagrams
of Fig. 141. During the cone-building period (Fig. 141 _a_ and _b_)
the plug is obviously built upward along with the cone, for lava often
flows out at a level a few hundred feet only below the crater rim. By
what process this chimney building goes on is not well understood,
though some light is thrown upon it by the post-eruption stage of Mont
Pelé in 1902-1903 (see below).


FIG. 141.—Three diagrams to illustrate the sequence of events
within the crater of a composite cone during the cone-building and
crater-producing periods. _a_ and _b_, two successive stages of the
cone building or Strombolian period; _c_, enlargement of the crater,
truncation of the cone, and destruction of the upper chimney during the
relatively brief crater-producing or Vulcanian period.]

Both the older and newer sections of this plug or chimney are furnished
some support against the outward pressure of the contained lava by
the surrounding wall of tuff; and they are, therefore, in a condition
not unlike that of the inner barrel of a great gun over which sleeves
of metal have been shrunk so as to give support against bursting
pressures. On the other hand, when not sustaining the hydrostatic
pressure of the liquid lava within, the chimney would tend to be
crushed in by the pressure of the surrounding tuff. Its strength to
withstand bursting pressures is dependent not alone upon the thickness
of its rock walls, but also upon its internal diameter or caliber. A
steam cylinder of given thickness of wall, as is well known, can resist
bursting pressures in proportion as its internal diameter is small. So
in the volcanic chimney, any tendency to remelt from within the chimney
walls must weaken them in a twofold ratio.

We are yet without accurate temperature observations upon the lava in
volcanic chimneys, but it seems almost certain that these temperatures
rise as the Vulcanian stage is approaching, and such elevation of
temperature must be followed by a greater or less re-fusion of the
chimney walls. The sequence of events during the late Vesuvian eruption
is, then, naturally explained by progressive re-fusion and consequent
weakening of the chimney walls, thus permitting a radial fissure to
open near the top and gradually extend downwards. Thus at first small
and high outlets were opened insufficient to drain the chimney, but
later, on April 7, after this fissure had been much extended and a new
and larger one had opened at a lower level, the draining began and the
surface of lava commenced rapidly to sink.

[Illustration: FIG. 142.—The spine of Pelé rising above the chimney of
the volcano after the eruption of 1902 (after Hovey).]

When the rapid sinking of the lava surface occurred, the lower lava
layers were almost immediately relieved of pressure, thus causing a
sudden expansion of the contained steam and resulting in grand crater
explosions. The partially refused and fissured upper chimney, now
unable to withstand the inward pressure of the surrounding tuff walls,
since outward pressures no longer existed, crushed in and contributed
its materials and those of the surrounding tuff to the fragments of
fresh lava rising in volume in the grand explosions (Fig. 141 _c_). In
outline, then, these seem to be the conditions which are indicated by
the sequence of observed events in connection with the late Vesuvian


FIG. 143.—Outlines of the Pelé spine upon successive dates. The full
line represents its outline on December 26, 1902; the dotted-dashed
line is a profile of January 3, 1902; while the dotted line is that of
January 9, 1903. The dark line is a fissure (after E. O. Hovey).]

=The spine of Pelé.=—The disastrous eruption of Mont Pelé upon
Martinique in the year 1902 is of importance in connection with the
interesting problem of the upward growth of volcanic chimneys during
the cone-building period of a volcano. After the conclusion of this
great Vulcanian eruption, a spine of lava grew upward from the
chimney of the main crater until it had reached an elevation of more
then a thousand feet above its base, a figure of the same order of
magnitude as the probable height of the upper section of the Vesuvian
chimney previous to the eruption of 1906. The Pelé spine (Fig. 142)
did not grow at a uniform rate, but was subject to smaller or larger
truncations, but for a period of 18 days the upward growth was at the
rate of about 41 feet per day. Later, the mass split upon a vertical
plane revealing a concave inner surface, and was somewhat rapidly
reduced in altitude to 600 feet (Fig. 143), only to rise again to its
full height of about 1000 feet some three months later.

While apparently unique as an observed phenomenon, and not free from
uncertainty as to its interpretation, the growth of this obelisk has
at least shown us that a mass of rock can push its way up above the
chimney of an active volcano even when there are no walls of tuff about
it to sustain its outward pressures.


FIG. 144.—Corrugated surface of the Vesuvian cone after the mud flows
which followed the eruption in 1906 (after Johnston-Lavis).]

=The aftermath of mud flows.=—When the late Vulcanian explosions of
Vesuvius had come to an end, all slopes of the mountain, but especially
the higher ones, were buried in thick deposits of the cocoa-colored
ash, included in which were larger and smaller projectiles. As this
material is extremely porous, it greedily sucks up the water which
falls during the first succeeding rains. When nearly saturated, it
begins to descend the slopes of the mountain and soon develops a
velocity quite in contrast with that of the slow-moving lava. The upper
slopes are thus denuded, while the fields and even the houses about the
base are invaded by these torrents of mud (_lava d’acqua_). Inasmuch as
these mud flows are the inevitable aftermath of all grander explosive
eruptions, the Italian government has of late spent large sums of
money in the construction of dikes intended to arrest their progress
in the future. It was streams of this sort that buried the city of
Herculaneum after the explosive eruption of 79 A.D.

After the mud flows have occurred, the Vesuvian cone, like all
similar volcanic cones under the same conditions, is found with deep
radial corrugations (Fig. 144), such as were long ago described as
“barrancoes” and supposed to support the “elevation crater” theory of
volcano formation.

=The dissection of volcanoes.=—To the uninitiated it might appear a
hopeless undertaking to attempt to learn by observation the internal
structure of a volcano, and especially of a complex volcano of the
composite type. The earliest successful attempt appears to have been
made by Count Caspar von Sternberg in order to prove the correctness
of the theory of his friend, the poet Goethe. Goethe had claimed that
a little hill in the vicinity of Eger, on the borders of Bohemia, was
an extinct volcano, though the foremost geologist of the time the
famous Werner, had promulgated the doctrine that this hill, in common
with others of similar aspect, originated in the combustion of a bed
of coal. The elevation in question, which is known as the Kammerbühl,
consists mainly of cinder, and Goethe had maintained that if a tunnel
were to be driven horizontally into the mountain from one of its
slopes, a core or plug of lava would be encountered beneath the summit.
The excavations, which were completed in 1837, fully verified the
poet’s view, for a lava plug was found to occupy the center of the mass
and to connect with a small lava stream upon the side of the hill (Fig.


FIG. 145.—The Kammerbühl near Eger, showing the tunnel completed in
1837 which proved the volcanic nature of the mountain (after Judd).]

It is not, however, to such expensive projects that reference is here
made, but rather to processes which are continually going on in nature,
and on a far grander scale. The most important dissecting agent for our
purpose is running water, which is continually paring down the earth’s
surface and disclosing its buried structures. How much more convincing
than any results of artificial excavation, as evidence of the internal
structure of a volcano, is the monument represented in Fig. 146,
since here the lava plug stands in relief like a gigantic thumb still
surrounded by a remnant of cinder deposits. Such exposed chimneys of
former volcanoes are found in many regions, and have become known as
volcanic _necks_, _pipes_, or _plugs_.

[Illustration: FIG. 146.—Volcanic plug exposed by natural dissection
of a volcanic cone in Colorado (U. S. G. S.).]


FIG. 147.—A dike cutting beds of tuff in a partly dissected volcano of
southwestern Colorado (after Howe, U. S. G. S.).]

Not infrequently the beds of tuff composing the flanks of the volcano,
upon dissection by the same process, bring to light walls of cooled
lava standing in relief (Fig. 147)—the filling of the fissure which
gave outlet to the flanks of the mountain at the time of the eruption.
Study of exposed dikes formed in connection with recent eruptions of
Vesuvius has shown that in many instances they are still hollow, the
lava having drained from them before complete consolidation.

Another agent which is effective in uncovering the buried structures
of volcanoes is the action of waves on shores. Always a relatively
vigorous erosive agency, the softer structures of volcanic cones are
removed with especial facility by this agent. On the shores of the
island of Volcano, the little cone of Vulcanello has been nearly half
carried away by the waves, so as to reveal with especial perfection the
structure of the cinder beds as well as the internal rock skeleton of
the mass. Here the characteristic dips of lava streams, intercalated as
they now are between tuff deposits and the lava which consolidated in
fissures, are both revealed.


FIG. 148.—Map and general view of St. Paul’s Rocks, a volcanic cone
dissected by waves.]

In mid-Atlantic a quite perfect crater, the St. Paul’s Rocks, has been
cut nearly in half so as to produce a natural harbor (Fig. 148).

In still other instances we may thank the volcano itself for opening
up the interior of the mountain for our inspection. The eruption in
1888 of the Japanese volcano of Bandai-san, by removing a considerable
part of the ancient cone, has afforded us a section completely through
the mountain. The summit and one side of the small Bandai was carried
completely away, and there was substituted a yawning crater eccentric
to the former mountain and having its highest wall no less than 1500
feet in height (Fig. 149). In two hours from the first warning of the
explosion the catastrophe was complete and the eruption over.


FIG. 149.—Dissection by explosion of Little Bandai-san in 1888 (after

The eruption of Krakatoa in 1883, probably the grandest observed
volcanic explosion in historic times, left a volcanic cone divided
almost in half and open to inspection (Fig. 150). Rakata, Danan,
and Perbuatan had before constituted a line of cones built up round
individual craters subsequent to the partial destruction of an
earlier caldera, portions of which were still existent in the islands
Verlaten and Lang. By the eruption of 1883 all the exposed parts and
considerable submerged portions of the two smaller cones were entirely
destroyed, and the larger one, known as Rakata, was divided just
outside the plug so as to leave a precipitous wall rising directly from
the sea and showing lava streams in alternation with somewhat thicker
tuff layers, the whole knit together by numerous lava dikes.


FIG. 150.—The half-submerged volcano of Krakatoa in the Sunda Straits
before and after the eruption of 1883 (after Verbeek).]

In order to carry our dissecting process down to levels below the
base of the volcanic mountain, it is usually necessary to inspect the
results of erosion by running water. Here the plug or chimney, instead
of being surrounded by tuff, is inclosed by the country rock of the
region, which is commonly a sedimentary formation. Such exposed lower
sections of volcanic chimneys are numerous along the northwestern
shores of the British Isles. Where aligned upon a dislocation or
noteworthy fissure in the rocks, the group of plugs has been referred
to as a scar or _cicatrice_ (Fig. 151). Associated with the plugs of
the cicatrice are not infrequently dikes, or, it may be, sheets of lava
extended between layers of sediment and known as _sills_.

[Illustration: FIG. 151.—The cicatrice of the Banat (after Suess).]

If we are able to continue the dissection process to still greater
depths, we encounter at last igneous rock having a texture known as
granitic and indicating that the process of consolidation was not
only exceedingly slow but also uninterrupted. This rock is found in
masses of larger dimensions, and though generally of more or less
irregular form, no one dimension is of a different order of magnitude
from the others. Such masses are commonly described as _bosses_, or,
if especially large, as _batholites_ (Fig. 152). Wherever the rock
beds appear as though they had been forced up by the upward pressure
of the igneous mass, the latter takes the form of a mushroom and has
been described as a _laccolite_ (Figs. 479-481, pp. 441-442). Evidence
seems, however, to accumulate that in the greater number of cases the
molten rock has fused its way upward, in part assimilating and in part
inclosing the rock which it encountered. This process of upward fusion
has been likened to the progress of a red hot iron burning its way
through a board.

=The formation of lava reservoirs.=—The discarding of the earlier
notion that the earth has a liquid interior makes it proper in
discussing the subject of volcanoes to at least touch upon the origin
of the molten rock material. As already pointed out, such reservoirs as
exist must be local and temporary, or it would be difficult to see how
the existing condition of earth rigidity could be maintained. From the
rate at which rock temperatures rise, at increasing depths below the
surface, it is clear that all rocks would be melted at very moderate
depths only, if they were not kept in a solid state by the prodigious
loads which they sustain. Any relief from this load should at once
result in fusion of the rock.


FIG. 152.—Diagram to illustrate a probable cause of formation of lava
reservoirs, and to show the connection between such reservoirs and the
volcanoes at the surface.]

Now the restriction of active volcanoes to those zones of the earth’s
surface within which mountains are rising, and where in consequence
earthquakes are felt, has furnished us at least a clew to the origin
of the lava. Regarded as a structure capable of sustaining a load,
the competency of an arch is something quite remarkable, so that the
arching up of strong rock formations into anticlines within the upper
layers of the zone of flow, or of combined fracture and flow, would
be sufficient to remove the load from relatively weak underlying beds,
which in consequence would be fused and form local reservoirs of lava
(Figs. 152 and 153).

It has been further quite generally observed that lines of volcanoes,
in so far as they betray any relation in position to neighboring
mountain ranges, tend to appear upon the rear or flatter limb of
unsymmetrical arches, or where local tension would favor the opening of
channels toward the surface. Moreover, wherever recent block movements
of surface portions of the earth’s shell have been disclosed in the
neighborhood of volcanoes, the latter appear to be connected with
downthrown blocks, as though the lava had, so to speak, been squeezed
out from beneath the depressed block or blocks.


FIG. 153.—Result of experiment with layers of composition to
illustrate the effect of relief of load upon rocks by arching of
competent formation (after Willis).]

We must not, however, forget that the igneous rocks are greatly
restricted in the range of their chemical composition. No igneous
rock type is known which could be formed by the fusion of any of
the carbonate rocks such as limestone or dolomite, or of the more
siliceous rocks, such as sandstone or quartzite. There remains only
the argillaceous class of sediments, the shales and slates, and so
soon as we examine the composition of these rocks we are struck by
the remarkable resemblance to that of the class of igneous rocks. For
purposes of comparison there is given below the composite or average
constitution of igneous rocks in parallel column, with the average
attained by combining the analyses of 56 slates and shales, the latter
recalculated with water excluded:

                │     AVERAGE IGNEOUS ROCK         │
                ├——————————————————┬———————————————┤  AVERAGE SHALE
                │    (Clark)       │  (Washington) │
                │                  │               │
  SiO_2         │    61.25         │   61.69       │   63.34
  Al_{2}O_3     │    15.81         │   15.94       │   16.56
  Fe_{2}O_3     │     2.70} 6.31   │    1.88} 4.53 │    4.41} 7.89
  FeO           │     3.61}        │    2.65}      │    3.48}
  MgO           │     4.47         │    4.90       │    3.54
  CaO           │     5.03         │    5.02       │    3.33
  Na_{2}O       │     3.64         │    4.09       │    1.29
  K_{2}O        │     2.87         │    3.35       │    3.52
  TiO_2         │      .62         │     .48       │     .53
                │   ——————         │  ——————       │  ——————
                │   100.00         │  100.00       │  100.00

This close resemblance is probably of deep significance, for the reason
that shales and slates are structurally the weakest of all rocks and
for the further reason that they rather generally directly underlie
the carbonate rocks, which are by contrast the strongest (see _ante_,
p. 37). For these reasons shales and slates are the only rocks which
are likely to be fused by relief from load through the formation of
anticlinal arches within the earth’s zone of flow. If this view is
well founded, lavas and other igneous rocks are in large part fused
argillaceous sediments formed in connection with the process of
folding, or are refused rocks of igneous origin and similar composition.

=Character profiles.=—The character profiles of features connected in
their origin with volcanoes are particularly easy to recognize, and in
a few cases in which they might be confused with others of a different
origin, an examination of the materials of the features should lead to
a definitive judgment.

The lava plains which result from massive outflows of basalt might
perhaps strictly be regarded as lack of feature, so great may be their
continuous extent. Wherever definite vents exist, a broad flat dome is
the usual result of the extravasation of a basaltic lava. The puys of
France and many of the Kuppen of Germany, being formed from less fluid
lava, have afforded profiles with relatively small radius of curvature.

In its youthful stage, the cinder cone usually presents a broad summit
sag and relatively short side slopes, whereas the cone of later stages
is apt to present long sweeping and upwardly concave curves with both
the gradient and the radius of curvature increasing rapidly toward
the summit. In contrast, too, with the earlier stage, the crest is
relatively small. A marked reduction in the high symmetry of such
profiles is noted wherever a breaching by lava outflow has occurred
(Fig. 154).

With the composite cone, complexity and corresponding lack of symmetry
is introduced, especially in the partially ruined caldera, and by
the more or less accidental distribution of parasitic cones, as well
as by migrations of the central cone. Peculiarly similar acuminated
profiles result from spatter-cone formation, from the formation of
a superchimney spine, and by the uncovering of the chimney through
denudational processes—the volcanic neck.

[Illustration: FIG. 154.—Character profiles connected with volcanoes.]

Another important feature resulting from denudation is the Mesa or
table mountain with its protecting basalt cap above softer rocks. Its
profile most resembles that of table mountains due to differential
erosion of alternately strong and weak horizontally bedded rocks, such
as compose the upper portion of the section in the Grand Cañon of the
Colorado. Here, however, in place of a single unusually strong top
layer there are found several strong layers in alternation with weaker
ones so as to produce additional steps in the profile.


 General works:—

 PAULETT SCROPE. The Geology of the Extinct Volcanoes of Central
 France. John Murray, London, 1858, pp. 258. (An epoch-making work of
 early date which, like the following reference, may be studied to
 advantage to-day.)

 SIR CHARLES LYELL. Principles of Geology, vol. 1, Chapters xxiii-xxv.

 MELCHIOR NEUMAYR. Erdgeschichte, vol. 1, Allgemeine Geologie, revised
 edition by v. Uhlig, 1897, pp. 133-277 (a storehouse of valuable
 information clearly presented).

 J. D. DANA. Characteristics of Volcanoes, with Contributions of Facts
 and Principles from the Hawaiian Islands. Dodd, Mead, and Company, New
 York, 1890, pp. 397.

 TEMPEST ANDERSON. Volcanic Studies in Many Lands, being reproductions
 of photographs by the author with explanatory notes. John Murray,
 London, 1903, pp. 200, pls. 105.

 T. G. BONNEY. Volcanoes, their Structure and Significance. John
 Murray, London, 1899, pp. 331.

 I. C. RUSSELL. Volcanoes of North America. Macmillan, New York, 1897,
 pp. 346.

 ELISÉE RÉCLUS. Les volcans de la terre, Belgian Society of Astronomy,
 Meteorology, and Physics of the Globe, 1906-1910 (a valuable
 descriptive geographical and bibliographical work of reference).

 G. MERCALLI. I vulcani attivi della terre. Hoepli, Milan, 1907, pp.
 421. (A most valuable work, beautifully illustrated, but in the
 Italian language.)

Arrangement of volcanic vents:—

 TH. THORODDSEN. Die Bruchlinien und ihre Beziehungen zu den Vulkanen,
 Pet. Mitt., vol. 51, 1905, pp. 1-5, pl. 5.

 R. D. M. VERBEEK. Various volumes and atlases of maps covering the
 Dutch East Indies and fully cited in the following reference (p. 21).

 WILLIAM H. HOBBS. The Evolution and the Outlook of Seismic Geology,
 Proc. Am. Phil. Soc., vol. 48, 1909, pp. 17-27.

Birth of volcanoes:—

 F. OMORI. The Usu-san Eruption and Earthquake and Elevation Phenomena,
 Bull. Earthq. Inv. Com., Japan, vol. 5, No. 1, 1911, pp. 1-37, pls.

Fissure eruptions:—

 TH. THORODDSEN. Island, IV, Vulkane, Pet. Mitt., Ergänzungsh. 153,
 1906, pp. 108-111.

 A. GEIKIE. Text-book of Geology, 4th ed., pp. 342-346.

Lava domes of Hawaii:—

 J. D. DANA. Characteristics of Volcanoes (as above).

 C. H. HITCHCOCK. Hawaii and Its Volcanoes. Honolulu, 1909, pp. 314.

Eruption of Matavanu volcano in 1906:—

 KARL SAPPER. Der Matavanu-Ausbruch auf Savaii, 1905-1906, Zeit. d.
 Gesell. f. Erdk. z. Berlin, vol. 19, 1906, pp. 686-709, 4 pls.

 H. J. JENSEN. The Geology of Samoa, and the Eruptions in Savaii, Proc.
 Linn. Soc., New South Wales, vol. 31, 1906, pp. 641-672, pls. 54-64.

 TEMPEST ANDERSON. The Volcano of Matavanu in Savaii, Quart. Jour.
 Geol. Soc., London, vol. 66, 1910, pp. 621-639, pls. 45-52.

Eruption of Volcano in 1888:—

 H. J. JOHNSTON-LAVIS. The South Italian Volcanoes. Naples, 1891, pp.
 342, pls. 16.

Eruption of Taal volcano in 1911:—

 W. E. PRATT. The Eruption of Taal Volcano, January 30, 1911, Phil.
 Jour. Sci., vol. 6, No. 2, Sec. A, 1911, pp. 63-86, pls. 1-14.

 F. H. NOBLE. Taal Volcano, album of views of 1911 eruption, Manila,
 1911, pp. 1-48.

The volcano of Etna:—

 G. VOM RATH. Der Aetna. Bonn, 1872, pp. 1-33. (A beautiful piece of
 descriptive writing from both the geological and scenic standpoints.)

 SARTORIUS VON WALTERSHAUSEN. Der Aetna. Leipzig, 1880, 2 quarto vols.,
 pp. 371 and 548.

The eruption of Vesuvius in 1906:—

 H. J. JOHNSTON-LAVIS. Geological Map of Monte Somma and Vesuvius, with
 a short and concise account, etc. Geo. Philip & Son, London, 1891.

 H. J. JOHNSTON-LAVIS. The Eruption of Vesuvius in April, 1906, Trans.
 Roy. Dublin Soc., vol. 9, 1909, Pt. VIII, pp. 139-200, pls. 3-23 (the
 most authoritative work upon the subject).

 T. A. JAGGAR, JR. The Volcano Vesuvius in 1906, Tech. Quart., vol. 19,
 1906, pp. 105-115.

 W. PRINZ. L’éruption du Vesuv d’avril, 1906, Ciel et Terre, 27e Année,
 1906, pp. 1-49.

 FRANK A. PERRET. Notes on the Electrical Phenomena of the Vesuvian
 Eruption, April, 1906, Sci. Bull., Brooklyn Inst. Arts and Sci., vol.
 1, No. 11, pp. 307-312; Vesuvius, Characteristics and Phenomena of the
 Present Repose Period, Am. Jour. Sci., vol. 28, 1909, pp. 413-430.

 WILLIAM H. HOBBS. The Grand Eruption of Vesuvius in 1906, Jour. Geol.,
 vol. 14, 1906, pp. 636-655.

The spine of Pelée:—

 E. O. HOVEY. The New Cone of Mont Pelée and the Gorge of the Rivière
 Blanche, Martinique, Am. Jour. Sci., vol. 16, 1903, pp. 269-281, pls.

 A. HEILPRIN. The Tower of Pelée. Philadelphia, 1904, pp. 62, pls. 22.

 A. LACROIX. La montagne Pelée et ses éruptions, Acad. des Sciences,
 Paris, 1904, Chapter iii.

 KARL SAPPER. In den Vulkangebieten Mittelamerikas und Westindiens,
 Stuttgart, 1905, pp. 172-178.

 A. C. LANE. Absorbed Gases of Vulcanism, Science, N.S., vol. 18, 1903,
 p. 760.

 G. K. GILBERT. The Mechanism of the Mont Pelée Spine, _ibid._, vol.
 19, 1904, pp. 927-928.

 I. C. RUSSELL. Pelée Obelisk once More, _ibid._, vol. 21, 1905, pp.

The dissection of volcanoes:—

 J. W. JUDD. Volcanoes, Chapter v.

 S. SEKYA and Y. KIKUCHI. The Eruption of Bandai-San, Trans. Seis.
 Soc., Japan, vol. 13, Pt. 2, 1890, pp. 140-222, pls. 1-9.

 R. D. M. VERBEEK. Krakatau. Batavia, 1885, pp. 557, pls. 25.

 ROYAL SOCIETY. The Eruption of Krakatoa and Subsequent Phenomena.
 London, 1888, pp. 494.

 G. K. GILBERT. Report on the Geology of the Henry Mountains, U.S.
 Geogr. and Geol. Surv., Rocky Mt. Region, Washington, 1877, pp. 22-60.

 SIR A. GEIKIE. Ancient Volcanoes of Great Britain, vol. 2 especially.

 D. W. JOHNSON. Volcanic Necks of the Mount Taylor Region, New Mexico,
 Bull. Geol. Soc. Am., vol. 18, 1907, pp. 303-324, pls. 25-30.



=The two contrasted processes of weathering.=—It has already been
pointed out that change and not stability is the order of nature.
Within the earth’s outer shell and upon it rock alteration goes on
continually, and from some portions of its surface the changed material
is as constantly migrating to neighboring or even far distant regions.
Before such transportation can begin the hard rock must first be broken
down and reduced to fragments which the transporting agencies are
competent to move.

To accomplish this breaking down, or _degeneration_, of the rock
masses, either a wide range in temperature or chemical reaction is
essential. In the atmosphere are found such active chemical agents as
oxygen and carbon dioxide, the so-called carbonic acid gas; and these
agents in the presence of water react chemically with the minerals of
the rocks and form other minerals such as the hydrates and carbonates,
which are lighter in weight and more soluble. This _chemical_ attack
upon the outer shell of the lithosphere is described as _decomposition_.

On the other hand the rock may succumb to changes which are purely
mechanical and are due either to the stresses set up by differences
between surface and interior temperatures, or to the prying action
of the frost in the crevices. Such purely mechanical degeneration
of the rocks is in contrast with decomposition and is described as
_disintegration_. The two processes of decomposition and disintegration
may, however, go on together; and the changes of volume that are caused
by decomposition may result directly in considerable disintegration, as
we are to see.

=The rôle of the percolating water.=—In order to effect chemical
change or reaction, it is essential that the substances which are
to react must be brought into such intimate contact with each other
as it is seldom possible to attain except by solution. The chemical
reactions which go on between the gaseous atmosphere and the solid
lithosphere are accomplished through solution of the gases in water.
This water, derived from rain or snow, percolates into the ground or
descends along the crevices in the rocks, carrying with it a certain
measure of dissolved air. This air differs from that of the surrounding
atmospheric envelope by containing relatively large amounts of oxygen
and of the other active element carbon dioxide. It follows from the
important rôle thus performed by the percolating water that the process
of decomposition will be relatively important in humid regions where
the atmospheric precipitation is sufficient for the purpose.


FIG. 155.—Successive diagrams to show the effect of decomposition and
resulting disintegration upon joint blocks so as to produce spheroidal
bowlders by weathering.]

Within hot and dry regions there is a larger measure of rock
disintegration, and distinct chemical changes unlike those of humid
regions take place in the higher temperatures and with the more
concentrated saline solutions. The discussion of such changes will be
deferred until desert conditions are treated in another chapter.

=Mechanical results of decomposition—spheroidal weathering.=—From
an earlier chapter it has been learned that the rocks of the earth’s
outermost shell are generally intersected by a system of vertical
fissures which at each locality tend to divide the rock into parallel
and upright rectangular prisms. It is these joints which offer
relatively easy paths for the descent of the water into the rocks.
In rocks of sedimentary origin there are found, in addition to the
vertical joints, planes of bedding originally horizontal, and in the
intrusive and volcanic rocks a somewhat similar parting, likewise
parallel to the surface of the ground. The combined effect of the
joints and the additional parting planes is thus to separate the rock
mass into more or less perfect squared blocks (Fig. 155, upper figure)
which stand in vertical columns.

The water which percolates downward upon the joints, finds its way
laterally along the parting planes, and so subjects the entire surface
of each block to simultaneous attack by its reagents. Though all
parts of the surface of each block are alike subject to attack, it
is the angles and the edges which are most vigorously acted upon. In
the narrow crevices the solutions move but sluggishly, and as they
are soon impoverished of their reagents in the attack upon the rock,
fresh solution can reach the middle of the faces from relatively few
directions. The edges are at the same time being reached from many more
directions, and the corners from a still larger number.

The minerals newly formed by these chemical processes of hydration
and carbonization are notably lighter, and hence more bulky than the
minerals from whose constituents they have been largely formed. Strains
are thus set up which tend to separate the bulkier new material from
the core of unaltered rock below. As the process continues, distinct
channels for the moving waters are developed favorable to action at
the edges and corners of the blocks. Eventually, the squared block is
by this process transformed into a spheroidal core of still unaltered
rock wrapped in layers of decomposed material, like the outer wrappings
of an onion. These in turn are usually imbedded in more thoroughly
disintegrated material from which the shell structure has disappeared
(Fig. 156).


FIG. 156.—Spheroidal weathering of an igneous rock.]

=Exfoliation or scaling.=—A fact of much importance to geologists, but
one far too often overlooked, is that rocks are but poor conductors
for heat. It results from this that in the bright sun of a summer’s
day a thin skin, as it were, upon the rock surface may be heated to
a relatively high temperature, although the layer immediately below
it is practically unaffected. The consequent expansion of the surface
layer causes stresses that tend to scale it off from the layer below,
which, uncovered in its turn, develops new strains of the same sort.
This process of exfoliation acquires exceptional importance in desert
regions where the rock surfaces are daily elevated to excessively high
temperatures (see Chapter XV).


FIG. 157.—Dome structure in granite mass, Yosemite valley, California
(after a photograph by Sinclair).]

=Dome structure in granite masses.=—In large granite masses, such
as are to be found in the ranges of the Sierra Nevada of California,
a peculiar dome structure is sometimes found developed upon a large
scale, and has had an important influence upon the breaking down of the
rock and upon the shaping of the mountain (Fig. 157). Such a structure,
made up as it is of prodigious layers, can have little in common with
the veneers of weathered minerals which are the result of exfoliation,
and it is quite likely that the dome structure is in some way connected
with the relief of these massive rocks from their load—the rock which
once rested upon them, but has been carried away by erosion since the
uplift of the range.

=The prying work of frost.=—In all countries where winter temperatures
range below the freezing point of water, a most potent agent of rock
disintegration is the frost which pries at every crevice and cranny
of the surface rock. Important in the temperate zones, in the polar
regions it becomes almost the sole effective agent of rock weathering.
There, as elsewhere, its efficiency as a disintegrating agent is
directly dependent upon the nature of the crevices within the rock, so
that the omnipresent joints are able to exercise a degree of control
over the sculpturing of the surface features which is hardly to be
looked for elsewhere (see plate 10 A).

[Illustration: FIG. 158.—Talus slope beneath a cliff.]

=Talus.=—Wherever the earth’s surface rises in steep cliffs, the
rock fragments derived from frost action, or by other processes of
disintegration, as they become detached either fall or slide rapidly
downward until arrested upon a flatter slope. Upon the earlier
accumulations of this kind, the later ones are deposited, until their
surface slopes up to the cliff face as steeply as the material will
lie—the angle of repose. Such débris accumulations at the base of a
cliff (Fig. 158) are known as _talus_, and the slope is described as a
talus slope, or in Scotland as a “scree.”


FIG. 159.—Striped ground from soil flow of chipped rock fragments upon
a slope, Snow Hill Island, West Antarctica (after Otto Nordenskiöld).]

=Soil flow in the continued presence of thaw water.=—So soon as the
rocks are broken down by the weathering processes, they are easily
moved, usually to lower levels. In part this transportation may be
accomplished by gravity slowly acting upon the disintegrated rock and
causing it to creep down the slope. Yet even in such cases water is
usually present in quantity sufficient to fill the spaces between the
grains, and so act as a lubricant to facilitate the migration.

Upon a large scale rocks which were either originally incoherent or
have been made so by weathering, after they have become saturated with
water, may start into sudden motion as great landslides or avalanches,
which in the space of a few moments materially change the face of the
country, and by burying the bottom lands leave disaster and misery in
their wake.


FIG. 160.—Pavement of horizontal surface due to soil flow, Spitzbergen
(after Otto Nordenskiöld).]

Within the subpolar regions, where a large part of the surface is for
much of the year covered with snow, the underlying rocks are for long
periods saturated with thaw water, and in alternation are repeatedly
frozen and thawed. Essentially similar conditions are met with in the
high, snow-capped mountains of temperate or torrid regions. For the
subpolar regions particularly it is now generally recognized that
somewhat special processes of soil flow, described under the name
_solifluction_, are characteristic. The exact nature of these processes
is as yet imperfectly understood, but there can be little doubt
concerning the large rôle which they have played in the transportation
of surface materials. Such soil flow is clearly manifested under
different aspects, and it is likely that by this comprehensive term
distinct processes have been brought together.


FIG. 161.—Tree roots entering fissured rock and prying its sections

Possibly the most striking aspect of the soil flow in subpolar regions
is furnished by the remarkable “stone rivers” and “rock glaciers”;
though the more generally characteristic are peculiar stripings or
other markings which appear upon the surface of the ground and thus
betray the movements of the underlying materials. Upon slopes it is not
uncommon for the surface to be composed of angular rock fragments riven
by the frost and crossed by broad parallel furrows as though a gigantic
plow had gone over it (Fig. 159). The direction of the furrows is
always up and down the slope, and the striping is marked in proportion
as the slope is steep. Where the bottom is reached, the furrows are
replaced by a sort of mosaic pavement of hexagonal repeating figures,
each of which may be an area of the surface six feet or more across
(Fig. 160, and Fig. 390, p. 368). The depressions which separate the
“blocks” of the pavement are often filled with clay, while the inclosed
surfaces are made up of coarsely chipped stone.

=The splitting wedges of roots and trees.=—In the mechanical breakdown
of the rocks within humid regions a not unimportant part is sometimes
taken by the trees, which insinuate the tenuous extremities of their
rootlets into the smallest cracks and by continued growth slowly wedge
even the firmer rocks apart (Fig. 161). In a similar manner the small
tree trunk growing within a crevice of the rock may in time split its
parts asunder (Fig. 162).


FIG. 162.—A large glacial bowlder split by a growing tree near East
Lansing, Michigan (after a photograph by Bertha Thompson).]

=The rock mantle and its shield in the mat of vegetation.=—Through
the action of weathering, the rocks, as we have seen, lose their
integrity within a surface layer, which, though it may be as much as a
hundred feet or more in thickness, must still be accounted a mere film
above the underlying bed rock. The mechanical agents of the breakdown
operate only within a few feet of the surface, and the agents of rock
decomposition, derived as they are from the atmosphere, become inert
before they have descended to any considerable depth. The surface layer
of incoherent rock is usually referred to as the _rock mantle_ (Fig.
163). Where the rock mantle is relatively deep, as it is in the states
south of the Ohio in the eastern United States, there is found, deep
below the outer layer of soil, a partially decomposed and disintegrated
rock, of which the unaltered minerals lie unchanged in position but
separated by the new minerals which have resulted from the breakdown
of their more susceptible associates. While thus in a certain sense
possessing the original structure, this altered material is essentially
incoherent and easily succumbs to attack by the pick and spade, so that
it is only at considerably greater depths that the unaltered rock is


FIG. 163.—Rock mantle consisting of broken rock, above which is
soil and a vegetable mat. Coast of California (after a photograph by

Because of the tendency of mantle rock to creep down upon slopes it is
generally found thicker upon the crests and at the bases of hills and
thinnest upon their slopes (Fig. 164).

In the transformation of the upper portion of the mantle rock into
soil, additional chemical processes to those of weathering are carried
through by the agency of earthworms, bacteria, and other organisms, and
by the action of humus and other acids derived from the decomposition
of vegetation. The bacteria particularly play a part in the formation
of carbonates, as they do also in changing the nitrogen of the air
into nitrates which become available as plant food. Within the humid
tropical regions ants and other insects enter as a large factor in rock
decomposition, as they do also in producing not unimportant surface


FIG. 164.—Diagram to show the varying thickness of mantle rock
upon the different portions of a hill surface (after Chamberlin and

How important is the cover of vegetation in retaining the rock
mantle and the upper soil layer in their respective positions, as
required for agricultural purposes, may be best illustrated by the
disastrous consequences of allowing it to be destroyed. Wherever, by
the destruction of forests, by the excessive grazing of animals, or
by other causes, the mat of turf has been destroyed, the surface is
opened in gullies by the first hard rain, and the fertile layer of soil
is carried from the slopes and distributed with the coarser mantle
upon the bottom lands. Thus the face of the country is completely
transformed from fertile hills into the most desolate of deserts where
no spear of grass is to be seen and no animal food to be obtained
(plate 5 A). The soil once washed away is not again renewed, for the
continuation of the gullying process now effectively prevents its

│  PLATE 5.                                                      │
│                                                                │
│ [Illustration: _A._ Once wooded region in China now reduced to │
│ desert                                                         │
│ through deforestation (after Willis).]                         │
│                                                                │
│ [Illustration: _B._ “Bad Lands” in the Colorado Desert (after  │
│ Mendenhall).]                                                  │


 Decomposition and disintegration:—

 GEORGE P. MERRILL. The Principles of Rock Weathering, Jour. Geol.,
 vol. 4, 1896, pp. 704-724, 850-871. Rocks, Rock Weathering, and Soils.
 Macmillan, New York, 1897, Pt. iii, pp. 172-411.

 ALEXIS A. JULIEN. On the Geological Action of the Humus Acids, Proc.
 Am. Assoc. Adv. Sci., vol. 28, 1879, pp. 311-410.

Corrosion of rocks:—

 C. W. HAYES. Solution of Silica under Atmospheric Conditions, Bull.
 Geol. Soc. Am., vol. 8, 1897, pp. 213-220, pls. 17-19.

 M. L. FULLER. Etching of Quartz in the Interior of Conglomerates,
 Jour. Geol., vol. 10, 1902, pp. 815-821.

 C. H. SMYTH, JR. Replacement of Quartz by Pyrites and Corrosion of
 Quartz Pebbles, Am. Jour. Sci. (4), vol. 19, 1905, pp. 282-285.

Dome structure of granite masses:—

 G. K. GILBERT. Domes and Dome Structure of the High Sierra, Bull.
 Geol. Soc. Am., vol. 15, 1904, pp. 29-36, pls. 1-4.

 RALPH ARNOLD. Dome Structure in Conglomerate, _ibid._, vol. 18, 1907,
 pp. 615-616.

Soil flow:—

 J. GUNNAR ANDERSSON. Solifluction, a Component of Subaërial
 Denudation, Jour. Geol., vol. 14, 1906, pp. 91-112.

 OTTO NORDENSKIÖLD. Die Polarwelt und ihre Nachbarländer, Leipzig,
 1909, pp. 60-65.

 ERNEST HOWE. Landslides in the San Juan Mountains, Colorado, etc.,
 Prof. Pap., 67 U. S. Geol. Surv., 1909, pp. 1-58, pls. 1-20.

 G. E. MITCHELL. Landslides and Rock Avalanches, Nat. Geogr. Mag., vol.
 21, 1910, pp. 277-287.

 WILLIAM H. HOBBS. Soil Stripes in Cold Humid Regions and a Kindred
 Phenomenon, 12th Rept. Mich. Acad. Sci., 1910, pp. 51-53, pls. 1-2.

Relation of deforestation to erosion:—

 N. S. SHALER. Origin and Nature of Soils, 12th Ann. Rept. U. S. Geol.
 Surv., 1891, Pt. 1, pp. 268-287.

 W. J. MCGEE. The Lafayette Formation, _ibid._, pp. 430-448.

 F. H. KING. Soils. Macmillan, New York, 1908, pp. 50-54.

 BAILEY WILLIS. Water Circulation and Its Control, Rept. Nat. Conserv.
 Com., 1909, vol. 2, pp. 687-710.

 W. J. MCGEE. Soil erosion, Bull. 71, U. S. Bureau of Soils, 1911, pp.
 60, pls. 33.



=The intricate pattern of river etchings.=—The attack of the weather
upon the solid lithosphere destroys the integrity of its surface
layer, and through reducing it to rock débris makes it the natural
prey of any agent competent to carry it along the surface. We have
seen how, for short distances, gravity unaided may pile up the débris
in accumulations of talus, and how, when assisted by thaw water which
has soaked into the material, it may accomplish a slow migration by a
peculiar type of soil flow. Yet far more potent transporting agencies
are at work, and of these the one of first importance is running
water. Only in the hearts of great deserts or in the equally remote
white deserts of the polar regions is the sound of its murmurings
never heard. Every other part of the earth’s surface has at some time
its running water coursing in valleys which it has itself etched into
the surface. It is this etching out of the continents in an intricate
pattern of anastomosing valleys which constitutes the chief difference
between the land surface and the relatively even floor of the oceans.

=The motive power of rivers.=—Every river is born in throes of Mother
Earth by which the land is uplifted and left at a higher level than
it was before. It is the difference of elevation thus brought about
between separated portions of the land areas that makes it possible for
the water which falls upon the higher portions to descend by gravity to
the lower. This natural “head” due to differences of elevation is the
motive power of the local streams, and for each increase in elevation
there is an immediate response in renewed vigor of the streams. The
elevated area off which the rivers flow is here termed an upland.

The velocity of a stream will be dependent not only upon the difference
in altitude between its source and its mouth, but upon the distance
which separates them, since this will determine the grade. The level
of the mouth being the lowest which the stream can reach is termed
the _base level_, and the current is fixed by the slope or declivity.
The capacity to lift and transport rock débris is augmented at a quite
surprising rate with every increase in current velocity, the law being
that the weight of the heaviest transportable fragment varies with the
sixth power of the velocity of the current. Thus if one stream flows
twice as rapidly as another, it can transport fragments which are
sixty-four times as heavy.

=Old land and new land.=—The uplifts of the continents may proceed
without changes in the position of the shore lines, in which case
areas, already carved by streams but no longer actively modified by
them, are worked upon by tools freshly sharpened and driven by greater
power. The land thus subjected to active stream cutting is described
as old land, and has already had engraved upon it the characteristic
pattern of river etchings, albeit the design has been in part effaced.

If, upon the other hand, the shore line migrates seaward with the
uplift, a portion of the relatively even sea floor, or new land, is
elevated and laid under the action of the running water. As we are to
see, stream cutting is to some extent modified when a river pattern is
inherited from the uplift. The uplift, whether of old land only or of
both old land and new land, marks the starting point of a new river
history, usually described as an _erosion cycle_.

[Illustration: FIG. 165.—Two successive forms of gullies from the
earliest stage of a river’s life (after Salisbury and Atwood).]

=The earlier aspects of rivers.=—Though geologists have sometimes
regarded the uplift of the continents as a sort of upwarping in a
continuous curved surface, the discussions of river histories and
the pictorial illustrations of them have alike clearly assumed that
the uplift has been essentially in blocks and that the elevated area
meets the lower lying country or the sea in a more or less definite
escarpment. The first rivers to develop after the uplift may be
described as gullies shaped by the sudden downrush of storm waters and
spaced more or less regularly along the margin of the escarpment (Fig.
165). These gullies are relatively short, straight, and steep; they
have precipitous walls and few, if any, tributaries.

[Illustration: FIG. 166.—Partially dissected upland (after Salisbury
and Atwood).]

With time the gully heads advance into the upland as they take on
tributaries; and so at length they in part invest it and dissect it
into numerous irregularly bounded and flat-topped tables which are
separated by cañons (Fig. 166). At the same time the grade of the
channel is becoming flatter, and its precipitous walls are being
replaced by curving slopes, as will be more fully described in the
sequel. It is because of this progressive reduction of grades with
increasing age that the early stages of a river’s life are much the
most turbulent of its history. The water then rushes down the steep
grades in rapids, and is often at times opened out in some basin to
form a lake where differences of uplift have been characteristic of
neighboring sections. For several reasons such basins in the course of
a stream are relatively short lived (Chapter XXX), and they disappear
with the earlier stages of the river history.

=The meshes of the river network.=—From the continued throwing out
of new tributaries by the streams, the meshes in the river network
draw more closely together as the stages of its history advance. The
closeness of texture which is at last developed upon the upland is in
part determined by the quantity of rainfall, so that in New Jersey with
heavy annual precipitation the meshes in the network are much smaller
than they are, for example, upon the semiarid or arid plains of the
western United States. Its design will, however, in either case more
or less clearly express the plan of rock architecture which is hidden
beneath the surface (Chapter XVII).

=The upper and lower reaches of a river contrasted.=—From the fact
that the river progressively invades new portions of the upland and
lays the acquired sections under more and more thorough investment, it
has near its headwaters for a long time a frontier district which may
be regarded as youthful even though the sections near its mouth have
reached a somewhat advanced stage. The newly acquired sections of river
valley may thus possess the steep grade and precipitous walls which are
characteristic of early gullies and cañons and are in contrast with the
more rounded and flat-bottomed sections below. Lateral streams, from
the fact that they are newer than the main or trunk stream to which
they are tributary, likewise descend upon somewhat steeper grades (Fig.

[Illustration: FIG. 167.—Characteristic longitudinal sections of the
upper portion of a river valley and its tributaries (after scaled
sections by Nussbaum).]

=The balance between degradation and aggradation.=—We have seen
that the power to transport rock fragments is augmented at a most
surprising rate with every increase in the current velocity. While the
lighter particles of rock may be carried as high up as the surface of
the water, the heavier ones are moved forward upon the bottom with
a combined rolling and hopping motion aided by local eddies. Those
particles which come in contact with the bottom or sides of the
channel abrade its surface so as ever to deepen and widen the valley.
This cutting accomplished by partially suspended débris in rapidly
moving currents of water is known as _corrasion_ and the stream is said
to be _incising_ its valley.

As the current is checked upon the lower and flatter grades, some of
its load of sediment, and especially the coarser portion, will be
deposited and so partially fill in the channel. A nice balance is thus
established between _degradation_ and the contrasted process known
as _aggradation_. The older the river valley the flatter become the
grades at any section of its course, and thus the point which separates
the lower zone of aggradation from the upper one of degradation moves
steadily upstream with the lapse of time.

=The accordance of tributary valleys.=—It is a consequence of the
great sensitiveness of stream corrasion to current velocity that
no side stream may enter the trunk valley at a level above that
of the main stream—the tributary streams enter the trunk stream
_accordantly_. Each has carved its own valley, and any abrupt increase
in gradient of the side streams near where they enter the main stream
would have increased the local corrasion at an accelerated rate and so
have cut down the channel to the level of the trunk stream.

=The grading of the flood plain.=—All rivers are subject to seasonal
variations in the volume of their waters. Where there are wet and
dry seasons these differences are greatest, and for a large part of
the year the valleys in such regions may be empty of water, and are
in fact often utilized for thoroughfares. In the temperate climates
of middle latitudes rivers are generally flooded in the spring when
the winter snows are melted, though they may dwindle to comparatively
small streams during the late summer. In the upper reaches of the
river the current velocities are such that the usual river channel may
carry all the water of flood time; but lower down and in the zone of
aggradation, where the current has been checked, the level of the water
rises in flood above the banks of its usual channel and spreads over
the surrounding lowlands. As a deposit of sediment is spread upon the
surface, the succession of the annual deposits from this source raises
the general level as a broad floor described as the _flood plain_ of
the river.

=The cycles of stream meanders.=—The annual flooding with water
and simultaneous deposition of silt is not, however, the only
grading process which is in operation upon the flood plain. It is
characteristic of swift currents that their course is maintained
in relatively straight lines because of the inertia of the rapidly
moving water. In proportion as their currents become sluggish, rivers
are turned aside by the smallest of obstructions; and once diverted
from their straight course, a law of nature becomes operative which
increases the curvature of the stream at an accelerated rate up to
a critical point, when by a change, sudden and catastrophic, a new
and direct course is taken, to be in its turn carried through a
similar cycle of changes. This so-called _meandering_ of a stream is
accompanied by a transfer of sediment from one bend or meander of the
river to those below and from one bank to the other. Inasmuch as the
later meanders cross the earlier ones and in time occupy all portions
of the plain to the same average extent, a process of rough grading is
accomplished to which the annual overflow deposit is supplementary.


FIG. 168.—Map and sections of a stream meander. The course of the main
current is indicated by the dashed line.]

The course of the current in consecutive meanders and the cross
sections of the channel which result directly from the meandering
process will be made clear from examination of Fig. 168. So soon
as diverted from its direct course, the current, by its inertia of
motion, is thrown against the outer or convex side so as to scour or
corrade that bank. Upon the concave or inner side of the curve there
is in consequence an area of slack water, and here the silt scoured
from higher meanders is deposited. The scouring of the current upon
the outer bank and the filling upon the inner thus gives to the cross
section of the stream a generally unsymmetrical character (Fig. 168
_ab_). Between meanders near the point of inflection of the curve, and
there only, the current is centered in the middle of the channel and
the cross section is symmetrical (Fig. 168 _cd_).

[Illustration: FIG. 169.—Tree in part undermined upon the outer bank
of a meander.]

The scour upon the convex side of a meander causes the river to swing
ever farther in that direction, and through invasion of the silted
flood plain to migrate across it. Trees which lie in its path are
undermined and fall outward in the stream with tops directed with the
current (Fig. 169). Whenever the flood plain is forested, the fallen
trees may be so numerous as to lie in ranks along the shore, and at the
time of the next flood they are carried downstream to jam in narrow
places along the channel and give the erroneous impression that the
flood has itself uprooted a section of forest (see p. 418).


FIG. 170.—Diagrams to show the successive positions of stream meanders
and the relatively stationary point near the sharpest curvature.]

=The cut-off of the meander.=—As the meander swings toward its
extreme position it becomes more and more closely looped. Adjacent
loops thus approach nearer and nearer to each other, but in the
successive positions a nearly stationary point is established near
where the river makes its sharpest turn (Fig. 170, _G_, and Fig. 454,
p. 417). At length the neck of land which separates meanders is so
narrow that in the next freshet a temporary jamming of logs within the
channel may direct the waters across the neck, and once started in
the new direction a channel is scoured out in the soft silt. Thus by
a breaking through of the bank of the stream, a so-called “crevasse”,
the river suddenly straightens its course, though up to this time it
has steadily become more and more sharply serpentine. After the cut-off
has occurred, the old channel may for a time continue to be used by the
stream in common with the new one, but the advantage in velocity of
current being with the cut-off, the old channel contains slacker water
and so begins to fill with silt both at the beginning and the end of
the loop. Eventually closed up at both ends, this loop or “ox-bow”
is entirely separated from the new channel, and once abandoned of the
stream is transformed into an ox-bow lake (Fig. 171 and p. 415).

[Illustration: FIG. 171.—An ox-bow lake in the flood plain of a river.]

=Meander scars.=—Swinging as it occasionally does in its meanderings
quite across the flood plain and against the bank of the earlier
degrading river in this section, the meander at times scours the high
bank which bounds the flood plain, and undermining it in the same
manner, it excavates a recess of amphitheatral form which is known as
a meander _scar_ (Fig. 172). At length the entire bank is scarred in
this manner so as to present to the stream a series of concave scallops
separated by sharp intermediate salients of cuspate form.


FIG. 172.—Schematic representation of a series of river terraces. _a_,
_b_, _c_, _e_, successive terraces in order of age. _d_, _d_, _d_, _d_,
terrace slopes formed of meander scars.]

=River terraces.=—Whenever the river’s history is interrupted by a
small uplift, or the base level is for any reason lowered, the stream
at once begins to sink its channel into the flood plain. Once more
flowing upon a low grade, it again meanders, and so produces new walls
at a lower level, but formed, like the first, of intersecting meander
scars. Thus there is produced a new flood plain with cliff and terrace
above, which is known as a _river terrace_. A succession of uplifts or
of depressions of the base level yields terraces in series, as they
appear schematically represented in Fig. 172. Such terraces are to be
found well developed upon most of our larger rivers to the northward of
the Ohio and Missouri. The highest terrace is obviously the remnant of
the earliest flood plain, as the lowest represents the latest.

=The delta of the river.=—As it approaches its mouth the river moves
more and more sluggishly over the flat grades, and swings in broader
meanders as it flows. Yet it still carries a quantity of silt which is
only laid down after its current has been stopped on meeting the body
of standing water into which it discharges. If this be the ocean, the
salinity of the sea water greatly aids in a quick precipitation of the
finest material. This clarifying effect upon the water of the dissolved
salt may be strikingly illustrated by taking two similar jars, the
one filled with fresh and the other with salt water, and stirring the
same quantity of fine clay into each. The clay in the salt water is
deposited and the water cleared long before the murkiness of the other
has disappeared.

By the laying down of the residue of its burden of sediment where it
meets the sea, the river builds up vast plains of silt and clay which
are known as deltas and which often form large local extensions of the
continents into the sea. Whereas in its upper reaches the river with
its tributary streams appears in the plan like a tree and its branches,
in the delta region the stream, by dividing into diverging channels
called distributaries (Fig. 458, p. 420), completes the resemblance to
the tree by adding the roots. From the divergence of the distributaries
upon the delta plain the Greek capital letter Δ is suggested and has
supplied the name for these deposits. Of great fertility, the delta
plains of rivers have become the densely populated regions of the
globe, among which it is necessary to mention only the delta of the
Nile in Egypt, those of the Ganges and Brahmaputra in India, and those
of the Hoang and Yangtse rivers in China.

=The levee.=—When the snows thaw upon the mountains at the headwaters
of large rivers, freshets result and the delta regions are flooded. At
such times heavily charged with sediment, a thin deposit of fertile
soil is left upon the surface of the delta plain, and in Egypt
particularly this is depended upon for the annual enrichment of the
cultivated fields. Though at this time the waters spread broadly over
the plain, the current still continues to flow largely within the
normal channel, so that the slack water upon either side becomes the
locus for the main deposit of the sediment. There is thus built up on
either side of the channel a ridge of silt which is known as a _levee_,
and this bank is steadily increased in height from year to year (Fig.

To prevent the danger of floods upon the inhabited plains, artificial
levees are usually raised upon the natural ones, and in a country like
Holland, such levees (dikes) involve a large expenditure of money and
no small degree of engineering skill and experience to construct. So
important to the life of the nation is the proper management of its
dikes, that in the past history of China each weak administration has
been marked by the development of graft in this important department
and by floods which have destroyed the lives of hundreds of thousands
of people.


FIG. 173.—“Bird-foot” delta of the Mississippi River.]

Wherever there has been a markedly rapid sinking upon a delta region,
and depressions are common in delta territory, no doubt as a result of
the loading down of the crust, the river may present the paradoxical
condition of flowing at a higher level than the surrounding country.
Between the levees of neighboring distributaries there are peculiar
saucer-shaped depressions of the country which easily become filled
with water. At the extremity of the delta the levee may be the only
land which shows above the ocean surface, and so present the peculiar
“bird-foot” outline which is characteristic of the extremity of the
Mississippi delta, though other processes than the mere sinking of the
deposits may contribute to this result (Fig. 173).

=The sections of delta deposits.=—If now we leave the plan of the
delta to consider the section of its deposits, we find them so
characteristic as to be easily recognized. Considered broadly, the
delta advances seaward after the manner of a railroad embankment
which is being carried across a lake. Though the greater portion of
the deposit is unloaded upon a steep slope at the front, a smaller
amount of material is dropped along the way, and a layer of extremely
fine material settles in advance as the water clears of its finely
suspended particles (Fig. 174). Simultaneous deposits within a delta
thus comprise a nearly horizontal layer of coarser materials, the
so-called top-set bed; the bulk of the deposit in a forward sloping
layer, the so-called fore-set bed; and a thin film of clay which is
extended far in advance, the bottom-set bed (Fig. 174, 2). If at any
point a vertical section is made through the deposits, beds deposited
in different periods are encountered; the oldest at the bottom in a
horizontal position, the next younger above them and with forward
dip, and the youngest and coarsest upon the top in nearly horizontal
position (Fig. 174, 3).


FIG. 174.—Diagrams to show the nature of delta deposits as exhibited
in section.]

It has been estimated that the surface of the United States is now
being pared down by erosion at the average rate of an inch in 760
years. The derived material is being deposited in the flood plain
and delta regions of its principal rivers. Some 513 million tons of
suspended matter is in the United States carried to tidewater each
year, and about half as much more goes out to sea as dissolved matter.
If this material were removed from the Panama Canal cutting, an 85-foot
sea-level canal would be excavated in about 73 days. The Mississippi
River alone carries annually to the sea 340 million tons of suspended
matter, or two thirds of the entire amount removed from the area of the
United States as a whole. It is thus little wonder that great deltas
have extended their boundaries so rapidly and that the crust is so
generally sinking beneath the load.



=The newly incised upland and its sharp salients.=—The successive
stages of incising, sculpturing, and finally of reducing an uplifted
land area, are each of them possessed of distinctive characters
which are all to be read either from the map or in the lines of the
landscape. Upon the newly uplifted plain the incising by the young
rivers is to be found chiefly in the neighborhood of the margins. In
this stage the valleys are described as V-shaped cañons, for the valley
wall meets the upland surface in sharp salients (plate 12 A), and the
lines of the landscape are throughout made up from straight elements.
Though the landscapes of this stage present the grandest scenery that
is known and may be cut out in massive proportions, often with rushing
river or placid lake to enhance the effect of crag and gorge, they lack
the softness and grace of outline which belong only to the maturer
erosion stages. The grand cañon of the Colorado presents the features
characteristic of this stage in the grandest and most sublime of all
examples, and the castled Rhine is a gorge of rugged beauty, carved out
from the newly elevated plateau of western Prussia, through which the
water swirls in eddying rapids (Fig. 175).


FIG. 175.—Gorge of the River Rhine near St. Goars, incised within an
uplifted plain which forms the hill tops.]

=The stage of adolescence.=—As the upland becomes more largely
invaded as a consequence of the headward advance of the cañons and
their sending out of tributary side cañons, the sharp angles in which
the cañon walls intersect the plain become gradually replaced by
well-rounded shoulders. Thus the lines in the landscape of this stage
are a combination of the straight line with a simple curve convex
toward the sky (Fig. 176). In this stage large sections of the original
plateau remain, though cut into small areas by the extensions of the
tributary valleys.


FIG. 176.—V-shaped valley with well-rounded shoulders characteristic of
the stage of adolescence. Allegheny plateau in West Virginia.]

=The maturely dissected upland.=—Continued ramifications by the
rivers eventually divide the entire upland area into separated parts,
and the rounding of the shoulders of valleys proceeds simultaneously
until of the original upland no easily recognizable compartments
are to be found. Where before were flat hilltops are now ridges or
watersheds, the well-known _divides_. The upland is now said to be
completely dissected or to have arrived at _maturity_. The streams are
still vigorous, for they make the full descent from the upland level
to base level, and yet a critical turning point of their history has
been reached, and from now on they are to show a steady falling off in
efficiency as sculpturing agents.


FIG. 177.—View of a maturely dissected upland from one of its
hilltops, Klamath Mountains, California (after a photograph by

Viewed from one of the hilltops, the landscape of this stage bears a
marked resemblance to a sea in which the numberless divides are the
crests of billows, and these, as distance reduces their importance in
the landscape, fade away into the even line of the horizon (Fig. 177).

=The Hogarthian line of beauty.=—Since the youthful stage of the
upland, when the lines of its landscape were straight, its character
rugged, and its rivers wild and turbulent, there has been effected
a complete transformation. The only straight line to be seen is the
distant horizon, for the landscape is now molded in softened outlines,
among which there is a repeated recurrence of the line of beauty made
famous by Hogarth in his “Analysis of Beauty.” As well known to all art
students, this is a sinuous line of reversed or double curvature—a
curve which passes insensibly at a point of inflection from convex
to concave (Fig. 178). The curve of beauty is now found in every
section of the hills, and it imparts to the landscape a gracefulness
and a measure of restfulness as well, which are not to be found in the
landscapes of earlier stages in the erosion cycle. In the bottoms of
the valleys also the initial windings of the rivers within their narrow
flood plains add silver beauty lines which stand out prominently from
the more somber background of the hills.

[Illustration: FIG. 178.—Hogarth’s line of beauty.]

Considered from the commercial viewpoint, the mature upland is one
of the least adaptable as a habitation for highly civilized man.
Direct lines of communication run up hill and down dale in monotonous
alternation, and almost the only way of carrying a railroad through the
region, without an expenditure for trestles which would be prohibitive,
is to follow the tortuous crest of a main divide or the equally winding
bed of one of the larger valleys.


FIG. 179.—View of the old land of New England, with Mount Monadnock
rising in the distance.]

=The final product of river sculpture—the peneplain.=—When maturity
has been reached in the history of a river, its energies are devoted
to a paring down of the valley slopes and crests so as to reduce the
general level. From this time on hill summits no longer fall into a
common level—that of the original upland—for some mount notably
higher than others, and with increasing age such differences become
accentuated. There is now also a larger aggradation of the valleys to
form the level floors of flood plains, out of which at length the now
slight elevations rise upon such gentle slopes that the process of
land sculpture approaches its end. Gradually the vigor of the stream
has faded away, and can now only be renewed through a fresh uplift of
the land, or, what would amount to the same thing, a depression of
the base level. Upland and river have reached old age together, and
the approximation to a new plain but little elevated above base level
is so marked that the name _peneplain_ is applied to it. Scattered
elevations, which because of some favoring circumstance rise to
greater heights above the general level of the peneplain, are known as
_monadnocks_ after the type example of Mount Monadnock in New Hampshire
(Fig. 179).

[Illustration: FIG. 180.—Comparison of the cross sections of river
valleys for the different stages of the erosion cycle.]

=The river cross sections of successive stages.=—To the successive
stages of a river’s life it has been common to carry over the names
from the well-marked periods of a human life. If neglecting for the
moment the general aspect of the upland, we fix our attention upon the
characteristic cross sections of the river valley, we find that here
also there are clearly marked characters to distinguish each stage
of the river’s life (Fig. 180). In infancy the steep, narrow, and
sharp-angled cañon is a characteristic; with youth the wider V-form
has already developed; in adolescence the angles of the cañon are
transformed into well-rounded shoulders, and the valley broadens so
as in the lower reaches to lay down a flood plain; in maturity the
divides and the double curves of the line of beauty appear; while in
the decline of old age the valleys are extremely broad and flat and are
floored by an extended flood plain.

=The entrenchment of meanders with renewed uplift.=—Upon the reduced
grades which are characteristic of the declining stage of a river’s
life, the current has little power to modify the surface configuration.
On the old land of this stage a renewed uplift starts the streams again
into action. This infusion of driving power into moving water, regarded
as a machine capable of accomplishing certain work, is like winding up
a clock that has run down. Once more the streams acquire a velocity
sufficient to enable them to cut their valleys into the land surface,
and so a new erosional cycle may be inaugurated upon the old land
surface—the peneplain. After such an uplift has been accomplished
and the rivers have sunk their early valleys within the new upland, we
may look out from this now elevated surface and the eye take in but a
single horizontal line, since we view the plain along its edge.

[Illustration: FIG. 181.—The Beavertail Bend of the Yakima Cañon in
central Washington (after George Otis Smith).]

By the uplift the meanders of the earlier rivers may become entrenched
in the new upland, the wide lobes of the individual meanders being now
separated by mountains where before had been plains of silt only. The
New River of the Cumberland plateau and the Yakima River of central
Washington (Fig. 181) furnish excellent American examples of intrenched
meanders, as the Moselle River does in Europe. Upon the course of the
latter river near the town of Zell a tunnel of the railroad a quarter
of a mile in length pierces a mountain in the neck of a meander lobe
in which the river itself travels a distance of more than six miles in
order to make the same advance. The Kaiser Wilhelm tunnel in the same
district penetrates a larger mountain included in a double meander of
the river. Although intrenched, river meanders are still competent to
scour and so undermine the outer bank, and with favoring conditions
they may by this process erode extended “bottoms” out of the plateau.
(See Lockport quadrangle, U. S. G. S.)

=The valley of the rejuvenated river.=—Whenever a new uplift occurs
before an erosional cycle has been completed, the rivers become
intrenched, not in a peneplain, but in the bottoms of broad valleys.
The sweeping curves which characterize mature landscapes may thus be
brought into striking contrast with the straight lines of youthful
cañons which with V-sections descend from their lowest levels (Fig.
182). The full cross section of such a valley shows a central V whose
sharp shoulders are extended outward and upward in the softened curves
of later erosion stages.

[Illustration: FIG. 182.—A rejuvenated river valley (after a
photograph by Fairbanks).]

=The arrest of stream erosion by the more resistant rocks.=—The
capacity of a river to erode and carry away the rock material that
lies along its course is dependent not only upon the velocity of the
current, but also upon the hardness, the firmness of texture, and the
solubility of the material. Particularly in arid and semiarid regions,
where no mantle of vegetation is at hand to mask the surfaces of the
firmer rock masses, differences of this kind are stamped deeply upon
the landscape. The rock terraces in the Grand Cañon of the Colorado
together represent the stronger rock formations of the region, while
sloping talus accumulations bury the weaker beds from sight.

[Illustration: FIG. 183.—Plan of a river narrows.]

Each area of harder rock which rises athwart the course of a stream
causes a temporary arrest in the process of valley erosion and is
responsible for a noteworthy local contraction of the river valley.
The valley is carved less widely as well as less deeply, and since a
river can never corrade below its base, a “temporary base level” is
for a time established above the area of harder rock. Owing to the
contraction of the valley under these conditions, the locality is
described as a river _narrows_ (Fig. 183). The narrows upon the Hudson
River occur in the Highlands where the river leaves a broad expanse
occupied by softer sediments to traverse an island-like area of hard
crystalline rocks. Within the narrows of a river the steep walls,
characteristic of youth and the turbulent current as well, are often
retained long after other portions of the river have acquired the more
restful lines of river maturity. The picturesque crag and the generally
rugged character of river narrows render them points of special
interest upon every navigable river.


FIG. 184.—Successive diagrams to illustrate repeated river piracy and
the development of “trellis drainage”, (after Russell).]

=The capture of one river’s territory by another.=—The effect of a
hard layer of rock interposed in the course of a stream is thus always
to delay the advance of the erosional process at all levels above the
obstruction. When a stream in incising its valley degrades its channel
through a veneer of softer rocks into harder materials below, it is
technically described as having _discovered_ the harder layer. Where
several neighboring streams flow by similar routes to their common base
level, those which discover a harder rock will advance their headwaters
less rapidly into the upland and so will be at a disadvantage in
extending their drainage territory. A stream which is not thus hindered
will in the course of time rob the others of a portion of their
territory, for it is able to erode its lower reaches nearer to base
level and thus acquire for its upper reaches, where erosion is chiefly
accomplished, an advantage in declivity. The divide which separates its
headwaters from those of its less favored neighbor will in consequence
migrate steadily into the neighbor’s territory. The divide is thus a
sort of boundary wall separating the drainage basins of neighboring
streams, and any migration must extend the territory of the one at
the expense of the other. As more and more territory is brought under
the dominion of the more favored stream, there will come a time when
the divide in its migration will arrive at the channel of the stream
that is being robbed, and so by a sudden act of annexation draw off
all the upper waters into its own basin. By this _capture_ the stream
whose territory has been invaded is said to have been _beheaded_.
By this act of _piracy_ the stronger stream now develops exceptional
activity because of the local steep grades near the point of capture,
and with this newly acquired cutting power the invader is competent to
advance still further and enter the territory of the stream that lies
next beyond. The type of drainage network which results from repeated
captures of this kind is known as “trellis drainage” (Fig. 184), a type
well illustrated by the rivers of the southern Appalachians.

In general it may be said that, other conditions being the same, of
two neighboring streams which have a common base level, that one which
takes the longest route will lose territory to the other, since it must
have the flatter average slope. Stream capture may thus come about
without the discovery of hard rock layers which are more unfavorable to
one stream than another.


FIG. 185.—Sketch maps to show the earlier and the present drainage
condition about the Blue Ridge near Harper’s Ferry.]

=Water and wind gaps.=—In the Allegheny plateau rivers cross, the
range of harder rocks in deep mountain narrows which upon the horizon
appear as gateways through the barrier of the mountain wall. Such
gateways are sometimes referred to as “water gaps”, of which the
Delaware Water Gap is perhaps the best known example, though the
Potomac crosses the Blue Ridge at the historic Harper’s Ferry through
a similar portal. The valley of the tributary Shenandoah has been the
scene of an interesting episode in the struggle of rival streams which
is typical of others in the same upland region. The records which may
be made out from the landscapes show clearly that in an earlier but
recent period, when the general surface stood at a higher level which
has been called the Kittatinny Plain, the younger Potomac of that time
and a younger but larger ancestor of Beaverdam Creek each crossed
the Blue Ridge of the time through similar water gaps (Fig. 185, map,
and Fig. 186). The Potomac of that time was, however, the more deeply
intrenched, and possessing an advantage in slope it was able to advance
the divide at the head of its tributary, the Shenandoah, into the
territory of Beaverdam Creek. Thus the beheading of the Beaverdam by
the Shenandoah was accomplished (Fig. 185, second map) and its upper
waters annexed to the Potomac system. With the subsequent lowering of
the general level of the country which yielded the present Shenandoah
Plain, the former water gap of Beaverdam Creek was abandoned of its
stream at a high level in the range. Known as Snickers Gap, it may
serve as a type of the “wind gaps” of similar origin which are not
altogether uncommon in the Appalachian Mountain system (Fig. 186).

[Illustration: FIG. 186.—Section to illustrate the history of Snickers

=Character profiles.=—For humid regions the landscapes possess
characters which, speaking broadly, depend upon the stage of the
erosion cycle. For the earliest stages the straight line enters as
almost the only element in the design; as the cycle advances to
adolescence the rounded forms begin to replace the angles of the
immature stages, and with full maturity the lines of beauty alone
are characteristic. As this critical stage is passed irregularity of
feature and ever more flattened curves are found to correspond to the
decline of the river’s vital energies. There are thus marks of senility
in the work of rivers (Fig. 187).

[Illustration: FIG. 187.—Character profiles of landscapes shaped by
stream erosion in humid climates.]



 SIR JOHN PLAYFAIR. Illustrations of the Huttonian Theory of the Earth.
 Edinburgh, 1802, pp. 350-371.

 J. W. POWELL. Exploration of the Colorado River of the West and its
 Tributaries. Washington, 1875, pp. 149-214.

 G. K. GILBERT. Report on the Geology of the Henry Mountains.
 Washington, 1877, pp. 99-150. (A classic upon the work of rivers.)

 C. E. DUTTON. Tertiary History of the Grand Cañon District (with
 atlas), Mon. 2, U. S. Geol. Surv., 1882, pp. 264.

 W. M. DAVIS. The Rivers and Valleys of Pennsylvania, Nat. Geogr. Mag.
 vol. 1, 1889, pp. 203-219; The Triassic Formation of Connecticut, 18th
 Ann. Rept. U. S. Geol. Surv., Pt. ii, 1898, pp. 144-153.

 SIR A. GEIKIE. The Scenery of Scotland. London, 1901, pp. 1-12.

 I. C. RUSSELL. Rivers of North America. Putnam. New York, 1898, pp.

 M. R. CAMPBELL. Drainage Modifications and their Interpretation, Jour.
 Geol., vol. 4, 1896, pp. 567-581, 657-678.

 HENRY GANNETT. Physiographic Types, U. S. Geol. Surv., Topographic
 Atlas, Folios 1-2, 1896, 1900.

 W. M. DAVIS. The Geographical Cycle, Geogr. Jour., vol. 14, 1899, pp.

The flood plain:—

 HENRY GANNETT. The Flood of April, 1897, in the Lower Mississippi,
 Scot. Geogr. Mag., vol. 13, 1897, pp. 419-421.

 W. M. DAVIS. The Development of River Meanders, Geol. Mag., Decade iv,
 vol. 10, 1903, pp. 145-148.

 W. S. TOWER. The Development of Cut-off Meanders, Bull. Am. Geogr.
 Soc., vol. 36, 1904, pp. 589-599.

River terraces:—

 W. M. DAVIS. The Terraces of the Westfield River, Massachusetts, Am.
 Jour. Sci., vol. 14, 1902, pp. 77-94, pl. 4; River Terraces in New
 England, Bull. Mus. Comp. Zoöl., vol. 38, 1902, pp. 281-346.

River deltas:—

 G. K. GILBERT. The Topographic Features of Lake Shores, 5th Ann.
 Rept. U. S. Geol. Surv., 1885, pp. 104-108; Lake Bonneville, Mon. I,
 U. S. Geol. Surv., 1890, pp. 153-167.

 Charts of Mississippi River Commission.

 G. R. CREDNER. Die Deltas, ihre Morphologie, geographische Verbreitung
 und Entstehungsbedingungen, Pet. Mitt. Ergh. 56, 1878, pp. 1-74, pls.

The peneplain:—

 W. M. DAVIS. Plains of Marine and Subaërial Denudation, Bull. Geol.
 Soc. Am., vol. 7, 1896, pp. 377-398; The Peneplain, Am. Geol., vol.
 23, 1899, pp. 207-239.

Intrenchment of meanders:—

 W. M. DAVIS. The Seine, the Meuse, and the Moselle, Nat. Geogr. Mag.,
 vol. 7, 1896, pp. 189-202.

Stream capture:—

 N. H. DARTON. Examples of Stream Robbing in the Catskill Mountains,
 Bull. Geol. Soc. Am., vol. 7, 1896, pp. 505-507, pl. 23.

 COLLIER COBB. A Recapture from a River Pirate, Science, vol. 22, 1893,
 p. 195.

 WILLIAM H. HOBBS. The Still Rivers of Western Connecticut, Bull. Geol.
 Soc. Am., vol. 13, 1902, pp. 17-22, pl. 1.

 ISAIAH BOWMAN. A Typical Case of Stream Capture in Michigan, Jour.
 Geol., vol. 12, 1904, pp. 326-334.



=The descent within the unsaturated zone.=—Of the moisture
precipitated from the atmosphere, that portion which neither evaporates
into the air nor runs off upon the surface, sinks into the ground and
is described as the _ground water_. Here it descends by gravity through
the pores and open spaces, and at a quite moderate depth arrives at a
zone which is completely saturated with water. The depth of the upper
surface of this saturated zone varies with the humidity of the climate,
with the altitude of the earth’s surface, and with many other similarly
varying factors. Within humid regions its depth may vary from a few
feet to a few hundred feet, while in desert areas the surface may lie
as low as a thousand feet or more.

The surface of the zone of the lithosphere that is saturated with water
is called the _water table_, and though less accentuated it conforms in
general to the relief of the country (Fig. 188). Its depth at any point
is found from the levels of all perennial streams and from the levels
at which water stands in wells.

[Illustration: FIG. 188.—Diagram to show the seasonal range in the
position of the water table and the cause of intermittent streams.]

During the season of small precipitation the water table is lowered,
and if at such times it falls below the bed of a valley, the surface
stream within the valley dries up, to be revived when, after heavier
precipitation, the water table has in turn been raised. Such streams
are said to be _intermittent_, and are especially characteristic of
semiarid regions (Fig. 188).

Wherever in descending from the surface an impervious layer, such as
clay, is encountered, the further downward progress of the water is
arrested. Now conducted in a lateral direction it issues at the surface
as a spring at the line of emergence of the upper surface of the
impervious layer (Fig. 189).

[Illustration: FIG. 189.—Diagram to show how an impervious layer
conducts the descending water in a lateral direction to issue in
surface springs.]

=The trunk channels of descending water.=—While within the
unconsolidated rock materials near the surface of the earth, it is
clear that water can circulate in proportion as the materials are
porous and so relatively pervious. As the pore spaces become minute and
capillary, the difficulty of permeation through the materials becomes
very great. Thus in the noncoherent rocks it is the coarse gravel and
the layers of sand which serve as the underground channels, while the
fine clays have the effect of an impervious wall upon the circulating
waters. In coarse sand as much as a third of the volume of the material
is pore space for the absorption and transmission of water. Even under
these favorable conditions the movement of the water is exceedingly
slow and usually less than a fifth of a mile a year.


FIG. 190.—Sketch map of the Oucane de Chabrières near Chorges in the
High Alps, to illustrate the corrosion of limestone along two series of
vertical joints (after Martel).]

Within the hard rocks it is the sandstones which have the largest pore
spaces, but in nearly all consolidated rocks there are additional
spaces along certain of the bedding planes, the joint openings (Fig.
190), and the crushed zones of displacement, so that these parting
planes become the trunk channels, so to speak, of the circulating
water. It is along such crevices that in the course of time the
mineral matter carried in solution by the water is deposited to produce
the ore veins and the associated crystallized minerals.

=The caverns of limestones.=—Where limestone formations have a nearly
flat upper surface, a large part of the surface water enters the rock
by way of the joint spaces, which it soon widens by solution into
broad crevices with well-rounded shoulders. At joint intersections
solution of the limestone is so favored that the water may here descend
in a sort of vertical shaft until it meets a bedding plane extending
laterally and offering more favorable conditions for corrosion. Its
journey now begins in a lateral direction, and solution of the rock
continuing, a tunnel may be etched out and extended until another
joint is encountered which is favorable to its further descent into
the formation. By this process on alternating shafts and galleries
the water descends to near the surface of the water table by a series
of steps, and is eventually discharged into the river system of the
district (Fig. 191). Within the larger caverns the water at the lowest
level usually flows as a subterranean river to emerge later into the
light from beneath a rock arch.


FIG. 191.—Diagram to show the relation of caverns in limestone to
the river system of the district and to the “swallow holes” upon the

From the plan of a system of connecting caverns it may often be
observed that the galleries of the several levels are alike directed
along two rectangular directions which indicate the master joint
directions within the limestone formation. This is especially clear
from the map of the galleries in the explored portions of the Mammoth
Cave (Fig. 192).

=Swallow holes and limestone sinks.=—Above the caverns of limestone
formations there are selected points where the water has descended
in the largest volume, and here funnel-shaped depressions have
been dissolved out from the surface of the rock. In different
districts such depressions have become known as “sinks”, “swallow
holes”, _entonnoirs_, and _Orgeln_. Wherever the depressions have
a characteristic circular outline, there can be little doubt that
they are the product of solution by the descending water, and have
relatively small connections only with the subterranean caverns. They
have thus naturally collected upon their bottoms the insoluble clay
which was contained in the impure limestone as well as a certain amount
of slope wash from the surface. Inasmuch as the clays are impervious
to water, the bottoms of these swallow holes are better supplied with
moisture than the surrounding rock surfaces, and by nourishing a more
vigorous plant growth are strongly impressed upon the landscape (Fig.

[Illustration: FIG. 192.—Plan of a portion of Mammoth Cave, Kentucky
(after H. C. Hovey).]


FIG. 193.—Trees and shrubs growing luxuriantly upon the bottoms of
sinks within a limestone country (after a photograph by H. T. A. de L.

Certain of the depressions above caverns are, however, less regular in
outline, and their bottoms are occupied by a mass of limestone rubble.
In some instances, at least, these depressions appear to be the result
of local incaving of the cavern roofs. An incaving of this nature may
close up an earlier gallery in the cavern and divert the cave waters
to a new course. The destruction of the roofs of caverns through this
process of incaving may continue until only relatively small remnants
are left. From long subterranean tunnels the caves are thus transformed
into subaërial rock bridges that have become known as “natural
bridges.” The best-known American example is the Natural Bridge near
Lexington, Virginia. Much grander natural bridges have been formed in
sandstone by a totally different process, and must not be confused with
these limestone remnants of caverns.

=The sinter deposits.=—Just as water can dissolve the calcareous
rocks with the formation of caverns, it can under other conditions
deposit the material which has thus been taken into solution. Its
power to hold carbonate of lime in solution is dependent upon the
presence of carbonic acid gas within the water. Water charged with gas
and dissolved lime carbonate is said to be “hard”, and if the gas be
driven off by boiling or otherwise, the dissolved lime is thrown out of
solution and deposited in a form well known to all housekeepers.

Hard water flowing in a surface stream, if dashed into spray at a
cascade, may deposit its lime carbonate in an ever thickening veneer
wherever the spray is dashed about the falls. This material, when cut
in section, has waving parallel layers and is known as _travertine_
or _calcareous sinter_. Some of the most remarkable deposits of this
nature may be seen at the cascade of Tivoli near Rome, and most of the
Roman buildings have been constructed from travertine that has been
quarried in the vicinity.

=The growth of stalactites.=—Water, after percolating slowly through
the crevices of limestone, where it becomes charged with the carbonic
acid gas and with dissolved carbonate of lime, may trickle from the
roof of a cavern. Emerging from the narrow crevice, it may give off
some of its contained gas and is usually subject to evaporation,
with the result that the lime carbonate is left adhering to the rock
surface from which evaporation took place. If the water collects upon
the cavern roof so slowly that it can entirely evaporate before a drop
can form, the entire content of carbonate will be left adhering to the
roof. Evaporation is most rapid near the margins and over the center of
each drop as it develops, and the deposit which is left thus takes the
form of tiny white rings at those points upon the crevice where there
is the easiest passage for the trickling water. To the outer surface
of these rings water will first adhere and then evaporate, as it will
also slowly ooze through the passage in the ring, but here without
evaporation until it reaches the lower surface. A pendant structure
will, therefore, develop, growing outward in all directions by the
deposition of concentric layers which are thickest near the roof, and
downward into the form of a rock “icicle” through evaporation of the
water which collects near the tip. These pendant sinter formations are
known as stalactites and are thus formed of concentric layers arranged
like a series of nested cornucopias with a perforation of nearly
uniform caliber along the axis of the structure (Fig. 194).


FIG. 194.—Diagrams to show the manner of formation of stalactites,
stalagmites, and sinter columns beneath parallel crevices upon the
roofs of caverns (in part after von Knebel).]

=Formation of stalagmites.=—Wherever the water percolates through
the roof of the cavern so rapidly that it cannot entirely evaporate
upon the roof, a portion falls to the floor, and, spattering as it
strikes, builds up a relatively thick cone of sinter known as a
stalagmite, and this is accurately centered beneath a stalactite
upon the roof. In proportion as the cavern is high, the dropping
water is widely dispersed as it strikes the floor, with the formation
of a correspondingly thick and blunt stalagmite. As this rises by
growth toward the roof, it often develops upon its summit a distinct
crater-like depression (Fig. 194, lower figure). When the process is
long continued, stalactites and stalagmites may grow together to form
columns which may be ranged with their neighbors like the pipes of an
organ, and like them they give out clear tones when struck lightly with
a mallet. At other times the columns are joined to their neighbors to
form hangings and draperies of the most fantastic and beautiful design
(Fig. 195).

[Illustration: FIG. 195.—Sinter formations in the Luray caverns,

In remote antiquity limestone caverns afforded a refuge to many species
of predatory birds and animals as well as to our earliest ancestors.
The bones of all these denizens of the caves lie entombed within the
clays and the sinter formations upon the cavern floors, and they tell
the story of a fierce and long-continued warfare for the possession of
these natural strongholds. The evidence is clear that these cave men
with their primitive weapons were able at times to drive away the cave
bears, lions, and hyenas, and to set up in the cavern their simple
hearths, only in their turn to be conquered by the ferocity of their
enemies. Some of the European caves have yielded many wagonloads of the
skeletons of these fierce predatory animals, together with the simple
weapons of the primitive man.

=The Karst and its features.=—Most so-called limestones have a large
admixture of argillaceous materials (clays) and of siliceous or sandy
particles. Such impurities make up the bulk of the clays and muds which
are left behind when the soluble portions of the limestone have been

[Illustration: FIG. 196.—Map of the dolines of the Karst region near

Swallow holes we have found to be characteristic features within such
districts. When limestones are more nearly pure, as in the Karst
region east of the Adriatic Sea, similar features are developed, but
upon a grander scale, and certain additional forms are encountered. In
place of the sink or swallow hole, there appears the “karst funnel” or
_doline_, a deep, bowl-shaped depression having a flat bottom. Such
funnels may be 30 to 3000 feet across and from 6 to 300 feet in depth
(Fig. 196). Though in one or two instances known to be the result of
the break down of cavern roofs (Fig. 197), yet like the swallow holes
of other regions these larger funnels appear generally to be the work
of solution by the descending waters. Where they have been opened in
artificial cuttings along railroads or in mines, the original rock is
found intact at the bottom, with small crevices only going down to
lower levels. Over the bottoms of the dolines there is spread a layer
of fertile red clay, the _terra rossa_, like that which is obtained
as a residue when a fragment of the limestone has been dissolved in
laboratory experiments.


FIG. 197.—Cross section of the doline formed by inbreak of a cavern
roof. The Stara Apnenka doline in Carinthia (after Martel).]

=A desert from the destruction of forests.=—Between the dolines is
found a veritable desert with jutting limestone angles and little if
any vegetation. The water which falls upon the surface either runs off
quickly or goes down to the subterranean caverns by which so much of
the country is undermined. Hence it is that the gardens which furnish
the sustenance for the scattered population are all included within
the narrow limits of the doline bottoms. Although to-day so largely a
barren waste, we know that the Karst upon the Adriatic was in remote
antiquity a heavily forested region and that it supplied the myriads of
wooden piles upon which the city of Venice is supported. The vessels
which brought to this port upon the Adriatic its ancient prosperity
were built from wood brought from this tract of modern desert. In the
days of Venetian grandeur the fertile terra rossa formed a veneer upon
the rock surface of the Karst and so retained the surface waters for
the support of the luxuriant forest cover. After deforestation this
veneer of rich soil was washed by the rains into the dolines or into
the few stream courses of the region, thus leaving a barren tract which
it will be all but impossible to reclaim (plate 6 A).

[Illustration: FIG. 198.—Sharp _Karren_ of the Ifenplatte Allgäu
(after Eckert).]

Upon the steeper slopes over the purer limestones, the rain water runs
away, guided by the joints within the rock. There is thus etched out a
more or less complete network of narrow channels (Fig. 190, p. 181),
between which the remnants rise in sharp blades to produce a structure
often simulated upon the fissured surface of a glacier that has been
melted in the sun’s rays (Fig. 401). These almost impassable areas of
karst country are described as _Schratten_ or _Karrenfelder_ (Fig. 198).

=The ponore and the polje.=—To-day large areas of the Karst are
devoid of surface streams, nearly all the surface water finding its
way down the crevices of the limestone into caverns, and there flowing
in subterranean courses. The foot traveler in the Karst country is
sometimes suddenly arrested to find a precipice yawning at his feet,
and looking down a well-like opening to the depth of a hundred feet or
more, he may see at the bottom a large river which emerges from beneath
the one wall to disappear beneath the other. These well-like shafts are
in the Austrian Karst known as _Ponores_, while to the southward in
Greece they are called _Katavothren_.

│  PLATE 6.                                                            │
│                                                                      │
│ [Illustration: _A._ Barren Karst landscape near the famous Adelsberg │
│ grottoes.                                                            │
│ (_Photograph by I. D. Scott._)]                                      │
│                                                                      │
│ [Illustration: _B._ Surface of a limestone ledge where joints have   │
│ been widened through solution.                                       │
│ Syracuse, N.Y.                                                       │
│ (_Photograph by I. D. Scott._)]                                      │

Elsewhere the karst river may emerge from its subterranean course in
a broader depressed area bounded by vertical cliffs, from which it
later disappears beneath the limestone wall. Such depressions of the
karst are known as _poljen_, and appear in most cases to be above the
downthrown blocks in the intricate fault mosaic of the region. Some of
these steeply walled inclosures have an area of several hundred square
miles, and especially at the time of the spring snow melting they are
flooded with water and so transformed into seasonal lakes (Fig. 199
and p. 422). It appears that at such times the cave galleries of the
region with their local narrows are not able to carry off all the water
which is conducted to them; and in consequence there is a temporary
impounding of the flood waters in those portions of the river’s course
which are open to the sky and more extended. The rush of water at
such times may bring the red clay into the subterranean channels in
sufficient quantity to clog the passages. The Zirknitz Lake usually has
high water two or three times a year, and exceptionally the flooding
has continued for a number of years. It has thus in some districts been
necessary to afford relief to the population through the construction
of expensive drainage tunnels.


FIG. 199.—The Zirknitz seasonal lake within a polje of the Karst
(after Berghaus).]

The conditions which are typified in the Karst area to the east of the
Adriatic Sea are encountered also in many other lands; as, for example,
in the Vorarlberg and Swiss Alps, in Lebanon, and in Sicily.

=The return of the water to the surface.=—Water which has descended
from the surface and been there held between impervious layers, may be
under the pressure of its own weight or “head”; and will later find its
way upward, it may be to the surface or higher, where a perforation is
discovered in its otherwise impervious cover. Such local perforations
are produced naturally by lines of fracture or faulting (widened at
their intersections), and artificially through the sinking of deep
wells. The water, which at ordinary times reaches the surface upon
fissures, is usually concentrated locally at the intersections of the
fracture network, where it issues in lines of fissure springs (Fig.
200); but at the time of earthquakes the water may rise above the
surface in lines of fountains (p. 83), or occasionally as sheets of
water which may mount some tens of feet into the air.


FIG. 200.—Fissure springs arranged upon lines of rock fracture at
intersections, Pomperaug valley, Connecticut.]

In contrast to the flow of surface springs, which varies with the
season through wide ranges both in its volume and in temperature of
the water, the volume of fissure springs is but slightly affected by
the seasonal precipitation, and the water temperature is maintained
relatively constant. Rock is but a poor heat conductor, and the
seasonal temperature changes descend a few feet only into the ground.
Thus water which rises from depths of a few hundred feet only is
apt to be icy cold, while from greater depths the effect of the
earth’s internal heat is apparent in a uniform but relatively higher
temperature of the water. Such “warm” or _thermal_ springs are apt to
contain considerable mineral matter in solution, both because the water
is far traveled and because its higher temperature has considerably
increased its solvent properties.

It has long been recognized that lines of junction of different rock
formations at the base of mountain ranges are localities favorable for
the occurrence of thermal springs. These junction lines are usually
within zones where by movement upon fractures the widest openings
in the rock have formed, and the catchment area of the neighboring
mountain highland has supplied head for the ground water. A map of the
hot springs within the Great Basin of the western United States would
present in the main a map of its principal faults.

=Artesian wells.=—From the natural fissure spring an artesian
well differs in the artificial character of the perforation of the
impervious cover to the water layer. The water of artesian wells may
flow out at the surface under pressure, or it may require pumping to
raise it from some lower level. Ideal conditions are furnished where
the geological structure of the district is that of a broad basin
or syncline. The water which falls in a neighboring upland is here
impounded between two parallel, saucer-like walls and will flow under
its head if the upper wall be perforated at some low level (Fig. 201,


FIG. 201.—Schematic diagrams to illustrate the different types of
artesian wells, (1) A non-flowing well; (2) flowing wells without basin
structure caused by clogging of the pervious formation; (3) flowing
wells in an artesian basin. The dotted lines are the water levels
within the pervious layers (after Chamberlin).]

A monoclinal structure may furnish artesian conditions when the
generally pervious layer has become clogged at a low level so as
to hold back the water (Fig. 248, 2). Pumping wells may be used
successfully even when such clogging does not exist, for the
slow-moving underground water flows readily in the direction of all
free outlets (Fig. 201, 1).

=Hot springs and geysers.=—Thermal springs whose temperature
approaches the boiling point of water are known as _hot springs_.
A _geyser_ is a hot spring which intermittently ejects a column of
water and steam. Both hot springs and geysers are to be found only in
volcanic regions, and appear to be connected with uncooled masses of
siliceous lava. In two of the three known geyser regions, Iceland and
New Zealand, the volcanoes of the neighborhood are still active, and
the lavas of the Yellowstone National Park date from the quite recent
geological period which immediately preceded the so-called “Ice Age.”

Wherever found, geysers are in the low levels along lines of drainage
where the underground water would most naturally reappear at the
surface. Their water has penetrated to considerable depths below the
surface, but has been chiefly heated by ascending steam or other
vapors. The water journey has been chiefly made along fissures, as is
shown by the cool springs which often issue near them. Though some hot
springs and geysers may disappear from a district, others are found
to be forming, and there is no good reason to think that geysers are
rapidly dying out, as was at one time supposed.

The action of a geyser was first satisfactorily explained by the great
German chemist Bunsen after he had made studies of the Icelandic
geysers, and the mechanics of the eruption was later strikingly
illustrated in the laboratory by an artificial geyser constructed by
the Irish physicist Tyndall. In many respects this action is like that
of the Strombolian eruption within a cinder cone, since it is connected
with the viscosity of the fluid and the resistance which this opposes
to the liberation of the developing vapor. In the case of the geyser, a
column of heated water stands within a vertical tube and is heated near
the bottom of the column.


FIG. 202.—Cross section of Geysir, Iceland, with simultaneously
observed temperatures recorded at the left, and the boiling
temperatures for the same levels at the right (after Campbell).]

Though the water may at its surface have the normal boiling temperature
and be there in quiet ebullition, the boiling point for all lower
levels is raised by the weight of the column of superincumbent liquid,
and so for a time the formation of steam within the mass is prevented.
In Fig. 202 is shown a cross section of the Icelandic _Geysir_ from
which our name for such phenomena has been derived, and to this section
have been added the actual observed temperatures of the water at the
different levels as well as the temperatures at which boiling can take
place at these levels. From this it will be seen that at a depth of 45
feet the water is but 2° Centigrade below its boiling point. A slight
increase of temperature at this level, due to the constantly ascending
steam, will not only carry this layer above the boiling point, but the
expansion of the steam within the mass will elevate the upper layers of
the water into zones where the boiling points are lower, and thus bring
about a sudden and violent ebullition of all these upper portions. Thus
is explained the almost universal observation that just before geysers
erupt the hot water rises in the bowls and generally overflows them.


FIG. 203.—Apparatus for simulating geyser action in the lecture room
(by courtesy of Professor B. W. Snow).]

The water ejected from the geyser is considerably cooled in the air;
and after its return to the tube must be again heated by the ascending
vapors before another eruption can occur. The measure of the cooling,
the time necessary to fill the tube, and the supply of rising steam,
all play a part in fixing the period which separates consecutive
eruptions. If the top of the tube be narrowed from its average
caliber, as is commonly observed to be true of the geysers within the
Yellowstone National Park, the escape of the steam is further hindered,
and frequent geyser eruption promoted.

An artificial geyser for demonstration of the phenomenon in the lecture
room is represented in Fig. 203. The cut has been prepared from a
photograph of an apparatus designed by Professor B. W. Snow of the
University of Wisconsin. In this design the tube is contracted so as
to have a top diameter one fourth only of what it is at the bottom,
where heat is directly applied by multiple Bunsen lamps. The water once
sufficiently heated, this artificial geyser erupts at regular intervals
of time which are dependent upon the dimensions of the apparatus and
the quantity of heat applied.

In case of natural geysers a considerable quantity of heat escapes
between eruptions in steam which issues quietly from the bowl of the
geyser. If this heat be retained by plugging the mouth of the tube with
a barrowful of turf, as is sometimes done with the geyser _Strokr_
in Iceland, eruption is promoted and so takes place earlier. Another
method of securing the same result is to increase the viscosity of the
water through the addition of soap, as was accidentally discovered by
a Chinaman who was utilizing the geyser water in the Yellowstone Park
for laundry operations. After this discovery it became a common custom
to “soap” the Yellowstone geysers in order to make them play; but this
method was prohibited under heavy penalty after the disastrous eruption
of the Excelsior Geyser.

=The deposition of siliceous sinter by plant growth.=—Geysers are
known only from areas of siliceous volcanic lava, and this may perhaps
have its cause in the easier solution of the geyser tube from such
materials. The silica dissolved in the heated waters is _again_
deposited at the surface to form _siliceous sinter_ or _geyserite_.
This material forms terraces surrounding the geysers or is built up
into mounds which are often quite symmetrical, such as those of the Bee
Hive and Lone Star geysers of the Yellowstone Park (Fig. 204).


FIG. 204.—Cone of siliceous sinter built up about the mouth of the
Lone Star Geyser in the Yellowstone National Park.]

The greater part of this separation of silica from the heated geyser
waters is due to the action of plants or algæ that are able to
grow in the boiling waters and which produce the beautiful colors
in the linings to the hot springs. The wonderful variety of the
tints displayed is accounted for by the fact that the algæ take on
different colors at different temperatures. The silica is deposited
from the water in the gelatinous hydrated form, which, however, dries
in the sun to a white sand. The growth within the pools goes on in
a manner similar to that of a coral reef, the algæ dying below and
there becoming encased in the rock lining while still continuing to
grow upon the surface. Whereas sinter of this nature, when deposited
by evaporation alone, can produce a maximum thickness of layer of a
twentieth of an inch each year, the growth from alga deposition within
limited areas may be as much as eight inches during the same period.



 F. H. KING. Principles and Conditions of the Movements of Ground
 Water, 19th Ann. Rept. U. S. Geol. Surv., 1899, Pt. ii, pp. 59-294,
 pls. 6-16.

 C. S. SLICHTER. The Motions of the Underground Waters, Water Supply
 Paper No. 67, U. S. Geol. Surv., 1902, pp. 1-106, pls. 1-8; Field
 Measurements of the Rate of Movement of Underground Waters, _ibid._,
 No. 140, 1905, pp. 1-122, pls. 1-15.

 M. L. FULLER. Occurrence of Underground Water, _ibid._. No. 114, 1905,
 pp. 18-40, pls. 4; Bibliographic review and index of papers relating
 to underground waters published by the United States Geological
 Survey, 1879-1904, _ibid._, No. 120, 1905, pp. 1-128.


 E. A. MARTEL. Les abimes, les eaux souterraines, les cavernes,
 les sources, la spélæologie. Delagrave, Paris, pp. 578. (Lavishly

 H. C. HOVEY. Celebrated American Caverns. Cincinnati, 1896, pp. 228;
 The Mammoth Cave of Kentucky. Louisville, 1897, pp. 111.

 J. W. BEEDE. Cycle of Subterranean Drainage in the Bloomington
 Quadrangle, Proc. Ind. Acad. Sci., 1910, pp. 1-31.

Karst conditions:—

 J. CVIJIC. Das Karstphänomen, Geogr. Abh., vol. 5, 1893.

 ÉMILE CHAIX. La topographie du desert de platé (Hautes Savoie), Le
 Globe, vol. 34, 1895, pp. 1-44, pls. 1-16, pp. 217-330.

 W. V. KNEBEL. Höhlenkunde mit Berücksichtigung der Karstphänomene.
 Vieweg, Braunschweig, 1906, pp. 222.

 A. GRUND. Die Karsthydrographie, Studien aus Westbosnien, Geogr. Abh.,
 vol. 7, No. 3, 1903, pp. 200.

 ÉMILE CHAIX-DU BOIS et ANDRÉ CHAIX. Contribution a l’étude des lapies
 en Carniole et au Steinernes Meer, Le Globe, vol. 46, 1907, pp. 17-56,
 pls. 26.

 P. ARBENZ. Die Karrenbildungen geschildert am Beispiele der
 Karrenfelder bei der Frutt in Kanton Obwalden (Schweiz). Deutsch.
 Alpenzeitung, Munich, 1909, pp. 1-9.

 F. KATZER. Karst und Karsthydrographie. Sarejevo, 1909, pp. 95.

 M. NEUMAYR. Erdgeschichte, vol. 1, pp. 500-510.

 E. DE MARTONNE. Traité de Géographie Physique, pp. 462-472 (excellent
 summaries in this and the last reference).

 E. A. MARTEL. The Land of the Causses, Appalachia, vol. 7, 1893, pp.
 18-149, pls. 4-13.

Fissure springs:—

 A. C. PEALE. Natural Mineral Waters of the United States, 14th Ann.
 Rept. U. S. Geol. Surv., Pt. ii, 1894, pp. 49-88.

 WILLIAM H. HOBBS. The Newark System of the Pomperaug Valley.
 Connecticut, 21st Ann. Rept. U. S. Geol. Surv., Pt. iii, 1901, pp.

Artesian wells:—

 T. C. CHAMBERLIN. Requisite and Qualifying Conditions of Artesian
 Wells, 5th Ann. Rept. U. S. Geol. Surv., 1885, pp. 131-173.

Hot springs and geysers:—

 A. C. PEALE. Yellowstone Park, Thermal Springs, 12th Ann. Rept. Geol.
 and Geogr. Surv. Ter. (Hayden), Pt. ii, Sec. ii, pp. 63-454 (many
 plates and maps).

 W. H. WEED. Geysers, Rept. Smithson. Inst., 1891, pp. 163-178.

 ARNOLD HAGUE and W. H. WEED (on hot springs and geysers of Yellowstone
 National Park), C. R. Cong. Géol. Intern., Washington, 1891, pp.

 W. H. WEED. Formation of Travertine and Siliceous Sinter by the
 Vegetation of Hot Springs, 9th Ann. Rept. U. S. Geol. Surv., 1889, pp.
 613-676, pls. 78-87.

 M. NEUMAYR. Erdgeschichte, vol. 1, pp. 500-510.

 ARNOLD HAGUE. Soaping Geysers, Trans. Am. Inst. Min. Eng., vol. 17,
 1889, pp. 546-553.

 JOHN TYNDALL. Heat as a Mode of Motion, New York, 1873, pp. 115-121
 (artificial geyser).



=The law of the desert.=—It is well to keep ever in mind that there is
no universal law which dominates Nature’s processes in all the sections
of her realm. Those changes which, because often observed, are most
familiar, may not be of general application, for the reason that the
areas habitually occupied by highly civilized races together comprise
but a small portion of the earth’s surface. In the dank tropical
jungle, upon the vast arid sand plains, and in the cold white spaces
near the poles, Nature has instituted peculiar and widely different

The fundamental condition of the desert is aridity, and this
necessitates an exclusion from it of all save the exceptional rain
cloud. Thus deserts are walled in by mountain ranges which serve
as barriers to intercept the moisture-bringing clouds. They are in
consequence saucer-shaped depressions, often with short mountain ranges
rising out of the bottoms, and such rain as falls within the inclosure
is largely upon the borders. Of this rainfall none flows out from the
desert, for the water is largely returned to the atmosphere through

The desert history is thus begun in isolation from the sea from which
the cloud moisture is derived, a balance being struck between inflow
and evaporation. Yet if deserts have no outlets, it is not true that
they have no rivers. These are occasionally permanent, often periodic,
but generally ephemeral and violent. The characteristic drainage of
deserts comes as the immediate result of sudden cloudburst. As a
consequence, the desert stream flows from the mountain wall choked
with sediment, and entering the depressed basin, is for the most part
either sucked down into the floor or evaporated and returned to the
atmosphere. The dissolved material which was carried in the water is
eventually left in saline deposits, and the great burden of sediment
accumulates in thick stratified masses which in magnitude outstrip the
largest deltas in the ocean.

=The self-registering gauge of past climates.=—From the initiation
of the desert in its isolation from the lands tributary to the sea,
its history becomes an individual and independent one. An increasing
quantity of rainfall will be marked by larger inflow to the basin, and
the lakes which form in its lowest depression will, as a consequence,
rise and expand over larger areas. A contrary climatic change will
bring about a lowering of the lakes and leave behind the marks of
former shorelines above the water level (Fig. 205). Deserts are thus
in a sense self-registering climatic gauges whose records go back far
beyond the historic past. From them it is learned that there have been
alternating periods of larger and smaller precipitation, which are
referred to as _pluvial_ and _interpluvial_ periods.

[Illustration: FIG. 205.—Former shore lines on the mountain wall
surrounding the desert of the Great Basin. View from the temple in Salt
Lake City (after Gilbert).]

From such records it is learned that the Great Basin of the western
United States was at one time occupied by two great desert lakes, the
one in the eastern portion being known as Lake Bonneville (Fig. 206).
With the desiccation which followed upon the series of pluvial periods,
which in other latitudes resulted in great continental glaciers and
has become known as the Glacial Period, this former desert lake dried
up to the limits of Great Salt Lake and a few smaller isolated basins.
Between 1850 and 1869 the waters of Great Salt Lake were rising, while
from 1876 to 1890 their level was falling, though subject to periodic
fluctuations, and in recent years the waters of the lake have risen so
high as to pass all records since the occupation of the country. As
a consequence the so-called Salt Lake “cut-off” of the Union Pacific
Railway, constructed at great expense across a shallow portion of the
lake, has been overflowed by its waters. The Sawa Lake in the Persian
Desert, which disappeared some five hundred years ago, again came into
existence in 1888 so as to cover the caravan route to Teheran.


FIG. 206.—Map of the former Lake Bonneville (dotted shores), and the
boundaries of the Great Salt Lake of 1869 (smaller area) and that of
the present (after Berghaus).]

The record in the rocks of the distant past reveals the fact that in
some former deserts barriers were, in the course of time, broken down,
with the result that an invading sea entered through the breached
wall. The result was the sudden destruction of land life, the remains
of which are preserved in “bone beds”, now covered by true marine
deposits. A still later episode of the history was begun when the sea
had disappeared and land animals again roamed above the earlier desert.
Such an alternation of marine deposits with the remains of land plants
and animals in the deposits of the Paris Basin, led the great Cuvier
to his belief that geologic history was comprised of a succession of
cataclysms in which life was alternately destroyed and re-created in
new forms—a view which later, under the powerful influence of Lyell
and Darwin, gave way to that of more gradual changes and the evolution
of life forms.

=Some characteristics of the desert wastes.=—The great stretches of
the arid lands have been often compared to the ocean, and the Bedouin’s
camel is known as “the ship of the desert.” Though a deceptive
resemblance for the most part, the comparison is not without its value.
Both are closed basins, and it is in this respect that the desert and
the ocean may be said to most resemble each other, for none of the
water and none of the sediment is lost to either except as boundaries
are, with the progress of time, transposed or destroyed. Flatness of
surface and monotony of scenery both have in common, and the waters and
the sand are in each case salt; yet the ocean, from the tropics to the
poles, has the same salts in essentially the same proportions, while in
the desert the widest variations are found both in the salts which are
present and in their relative quantities.

Upon the borders of the ocean are found ridges of yellow sand heaped up
by the wind, but these ramparts are small in comparison to those which
in deserts are found upon the borders (plate 7 A).

The desert is a land of geographic paradoxes. As Walther has pointed
out, we have rain in the desert which does not wet, springs which yield
no brooks, rivers without mouths, forests preserved in stone, lakes
without outlets, valleys without streams, lake basins without lakes,
depressions below the level of the sea yet barren of water, intense
weathering with no mantle of disintegrated rock, a decomposition of the
rocks from within instead of from without, and valleys which branch
sometimes upstream and sometimes down.

Within the deserts curious mushroom-like remnants of erosion afford a
local relief from the searching rays of the desert sun. Pocket-like
openings large enough for a hermit’s habitation are hollowed out by the
wind from the disintegrated rock masses. Amphitheaters open out from
little erosion valleys or wadi, and isolated outliers of the mountains
stand like sentinels before their massive fronts.

Because of the general absence of clouds above a desert, no shield
such as is common in humid regions is provided against the blinding
intensity of the sun’s rays. Sun temperatures as high as 180°
Fahrenheit have been registered over the deserts of western Africa.
Every one is familiar with the fact that a blanket of thick clouds is a
prevention of frosts at night, for, with the setting of the sun and the
consequent radiation of heat from the earth, these rays are intercepted
by the clouds, returned and re-returned in many successive exchanges.
Over desert regions the absence of any such blanket of moisture is
responsible for the remarkable falls of temperature at sunset. Though
shortly before temperatures of 100° Fahrenheit or greater may have been
measured, it is not uncommon for water to freeze during the following
night. Much the same conditions of sudden temperature change with
nightfall are experienced in high mountains when one has ascended above
the blanketing clouds.


FIG. 207.—Borax deposits upon the floor of Death valley, California
(after a photograph by Fairbanks).]

=Dry weathering—the red and brown desert varnish.=—In desert lands
the fierce rays of the sun suck up all the available moisture, and
the water table may be hundreds of feet below the surface. Roots of
trees a hundred feet or more in length have been found to testify to
the fierce struggle of the desert plant with the arid conditions. In
humid regions the meteoric water dissolves the more soluble sodium
salts near the surface of the rock and carries them out to the ocean,
where they add to the saltness of the sea. In the desert the rare
precipitations prevent an outflow, but the sun’s strong rays suck out
with the moisture the salts from within the rock, and evaporating upon
the surface, the salts are left as a coat of “alkali”, which is in part
carried away on the wind and in part washed off in one of the rare
cloudbursts. In either case these constituents find their way to the
lowest depressions of the basin, where they contribute to the saline
deposits of the desert lakes (Fig. 207).

[Illustration: FIG. 208.—Hollowed forms of weathered granite in a
desert of central Asia (after Walther).]

Certain of the saline constituents of the rocks, as they are thus
drawn out by the sun’s rays, fuse with the rock at the surface to form
a dense brown substance with smooth surface coat, known as _desert
varnish_. Within the interior a portion of the salts crystallize within
the capillary fissures, and like water freezing within a pipe, they
rend the walls apart. As a direct consequence of this disintegrating
process the interior of rock masses may crumble into sand; and if the
hard shell of varnish be broken at any point, the wind makes its
entrance and removes the interior portion so as to leave a hollow
shell—the characteristic “pocket rock” (Fig. 208) of the desert. The
nummulitic limestone of Mokkatan and many of the great hewn blocks of
Egyptian limestone sound hollow under the tap of the hammer, and when
broken, they reveal a shell a few inches only in thickness (Fig. 209).

[Illustration: FIG. 209.—Hollow hewn blocks in a wall in the Wadi
Guerraui (after Walther).]

The brown desert varnish is one of the most characteristic marks
of an arid country. It is found in all deserts under much the same
conditions, and is especially apt to be present in sandstone. When
scratched, the surface of the rock becomes either cherry-red,
indicating anhydrous ferric oxide, or it is yellowish, due to the
hydrated iron oxide which we know as iron rust. Thus it is seen that
the sands of deserts, in contrast to those yielded by other processes
within humid regions, have a characteristic red color, and this may
vary from brownish red upon the one hand to a rich carmine upon the

=The mechanical breakdown of the desert rocks.=—The chemical changes
of decomposition within desert rocks are, as we have seen, largely due
to the action of concentrated solutions of salts at high temperatures.
That there is a certain mechanical rending of these rocks, due to the
“freezing” of salts within the capillary fissures, has been already
mentioned. A further strain effect arises in rocks like granite, which
are a mixture of different minerals. Heated to a high temperature
during the day and cooled through a considerable range at night, the
different minerals alternately expand and contract at different rates
and by different relative amounts, so that strains are set up, tending
to tear them apart. The effect of these strains is thus a surface
crumbling of rocks.

But rock is, as already pointed out, a relatively poor conductor of
heat, and hence it is a relatively thin skin only which passes through
the daily round of wide temperature range. This outer shell when heated
is expanded, and so tends to peel off, or exfoliate, like the outer
skin of an onion. The process is therefore described as _exfoliation_.
In all rocks of homogeneous texture the continued action of this
process results in convexly spherical surfaces, the material scaled
off in the process remaining as a slope or talus which surrounds the
projecting knob (Fig. 210). Naked, these projecting domes rise above
the rim of débris at their bases. Not a particle of dust adheres to the
fresh rock surface—no dirt interferes with its glaring whiteness. Yet
close at hand lie masses of débris into which wells may be carried to
depths of more than six hundred feet without encountering either solid
rock or ground water. The bare walls of granite sometimes mount upwards
for thousands of feet into the air, as steep and as inaccessible as the
squared towers of the Tyrolean Dolomites.

[Illustration: FIG. 210.—Smooth granite domes shaped by exfoliation
and surrounded by a rim of talus. Gebel Karsala, Nubian Desert (after

Rock is such a poor conductor of heat that special strains are set
up at the margin of sunlight and shade. This localization of the
disintegration on the margin of the shaded portions of rock masses is
known as _shadow weathering_ (see Fig. 215, p. 206).

There is, however, still another mechanical disintegrating process
characteristic of the desert regions, which is likewise dependent
upon the sudden changes of temperature. Rains, though they may not
occur for a year or more, come as sudden downpours of great volume and
violence. Rock masses, which are highly heated beneath the desert sun,
if suddenly dashed with water, may be rent apart by the differential
strains set up near the surface. That rocks may be easily rent as a
result of sudden chilling is well known to our Northern farmers, who
are accustomed to rid themselves of objectionable bowlders by first
building a fire about them and then dashing water upon their surface.
Thus split into fragments, even the larger bowlders may be handled and
so removed from the farming land. The natural process of rock rending
by the occasional cloudburst may be described as _diffission_. Blocks
as much as twenty-five feet in diameter have been observed in the
desert of western Texas, soon after being broken into several fragments
at the time of a downpour of rain (Fig. 211).


FIG. 211.—Granite blocks in the Sierra de los Dolores of Texas, rent
into several fragments by the dash of rain (after Walther).]

=The natural sand blast.=—Because of the saucer-like shape, the vast
expanse, and the absence of wind breaks, the potency of wind as a
geological agent is in desert areas not easily overestimated. While
most of its work is accomplished with the aid of tools, it has been
proven that even without this help, considerable work is done through
the friction of the wind alone, particularly when moving as powerful
eddies in cracks and crannies. This wear of the wind, unaided by
cutting tools, is known as _deflation_.

The greater work of the wind is, however, accomplished with the aid
of larger or smaller rock particles, the sand and dust, with which it
is so generally charged above the deserts. Unprotected by any mat of
vegetation the materials of the desert surface are easily lifted and
are constantly migrating with the wind. The finest dust is raised high
into the air, and is carried beyond the marginal barriers, but none of
the sand or coarser materials ever passes beyond the borders.


FIG. 212.—“Mushroom rock” from a desert in Wyoming (after Fairbanks).]

The efficiency of this sand as a cutting tool when carried by the wind
is directly proportioned to the size of the grain, since with larger
fragments a heavier blow is struck when carried at any given velocity.
These more effective grains are, however, not lifted far above the
ground, but advance with a squirming or hopping motion, much as do
the larger pebbles upon the bottom of a river at the time of a spring
freshet. To quote Professor Walther: “Whoever has had the opportunity
to travel over a surface of dune sand when a strong wind is blowing has
found it easy to convince himself of the grinding action of the wind.
At such times the ground becomes alive, everywhere the sand is creeping
over the surface with snake-like squirmings, and the eye quickly tires
of these writhing movements of the currents of sand and cannot long
endure the scene.”


FIG. 213.—Windkanten shaped by the desert sand blast (after Chamberlin
and Salisbury).]

A direct consequence of this restriction of the more effective cutting
tools to the layer of air just above the ground, is the strong tendency
to cut away all projecting masses near their bases. The “mushroom
rocks”, which are so characteristic of desert landscapes, have been
shaped in this manner (Fig. 212). Another product of the desert
sand blast is the so-called _Windkante_ (wind-edge) or _Dreikante_
(three-edge), a pebble which is usually shaped in the form of a pyramid
(Fig. 213).

Whenever a rock face, open to direct attack by the drifting sand, is
constituted of parts which have different hardness, the blast of sand
pecks away at the softer places and leaves the harder ones in relief.
Thus is produced the well-known “stone lattice” of the desert (Fig.
214). Particularly upon the neck of the great Sphinx have the flying
sand grains, by removing the softer layers, brought the sedimentary
structures of the sandstone into strong relief.


FIG. 214.—The “stone lattice” of the desert, the work of the natural
sand blast (after Walther).]

When guided both by planes of sedimentation and planes of jointing,
forms of a very high degree of ornamentation are developed. Some of
the most remarkable forms are due to the protection afforded to the
sun-exposed surfaces by the shell of desert varnish. In the shaded
portions of projecting masses there is no such protection, and here the
sand blast insinuates itself into every crack and cranny. In this it is
aided by shadow weathering due to the differential strains set up at
the border of the expanded sun-heated surface. As a result, projecting
rock masses are sometimes etched away beneath and give the effect of a
squatting animal. These forms, due to shadow erosion, have also been
likened to projecting faucets. (Fig. 215).


FIG. 215.—Projecting rock carved by the drifting sand into the form of
a couchant animal as a result of shadow weathering and erosion. Cut in
granite on the north Indian Desert (after Walther).]

Worn by its impact upon neighboring sand grains while in transport, but
much more as it is thrown against the ground or hard rock surfaces,
the wind-driven or _eolian_ sand is at last worn into smoothly rounded
granules which approach the form of a sphere. Compared to the surface
which sea sand acquires by attrition, this shaping process is much
the more efficient, since in the water the beach sand is buoyed up
and is more effectively cushioned against its neighboring grains. The
grains of beach sand when examined under a microscope are found to be
much more irregular in form and usually display the original fracture
surfaces only in part abraded.


FIG. 216.—Cliffs in loess 200 feet in height which exhibit the
characteristic vertical jointing (after von Richtofen).]

=The dust carried out of the desert.=—When, standing upon the mountain
wall that surrounds a desert, the traveler gazes out to windward over
the great depression, his field of view is generally obscured by the
yellow haze of the dust clouds moving across the margins. Upon the
mountain flanks and extending far outside the borders, this cloud of
dust settles as a shrouding mantle of impalpable yellow powder, which
is known as _loess_. These deposits are continually deepening, and
have sometimes accumulated until they are hundreds or even thousands
of feet in thickness. Before reaching its final resting place the dust
of this deposit may have settled many times, and has certainly been in
part redistributed by the streams near the desert margin. In it are
the ingredients which are necessary for the nourishment of plants, and
it constitutes the most important of natural soils. Continually fed by
new deposits from the desert, and refertilized from below by a natural
process so soon as the upper layers become impoverished, it requires no
artificial fertilization. Without artificial aids the loess of northern
China has been tilled for thousands of years without any signs of


FIG. 217.—A cañon in loess worn by traffic and wind. A highway in
northern China (after von Richtofen).]

Though easily pulverized between the fingers, loess is none the less
characterized by a perfect vertical jointing and stands on vertical
faces as does the solid rock (Fig. 216), but it is absolutely devoid
of layers or bedding. Its capacity of standing in vertical cliffs the
loess owes to a never failing content of lime carbonate which acts as a
cement, and to a peculiar porous structure caused by capillary canals
that run vertically through the mass, branching like rootlets and lined
with carbonate of lime. This texture once destroyed, loess resolves
itself into a common sticky clay.

By the feet of passing animals or by wheels of vehicles, the loess is
crushed, and a portion is lifted and carried away by the wind. Thus in
the course of time roadways sink deep into the mass as steep-walled
cañons (Fig. 217). A portion of the now structureless clay remaining
upon the roadway is at the time of the rains transformed into a thick
mud which makes traveling all but impossible, though before its
structure has been destroyed the loess is perfectly drained to the
bottom of its deposits.

The particles which compose the loess are sharply angular quartz
fragments, so fine that all but a few grains can be rubbed into the
pores of the skin. Fine scales of mica, such as are easily lifted by
the wind, are disseminated uniformly throughout the mass. The only
inclosures which are arranged in layers consist of irregularly shaped
concretions of clay. These show a striking resemblance to ginger roots
and are called by the Chinese “stone ginger”, though they are elsewhere
more generally known by their German name of _Loessmännchen_, or loess
dolls. These concretions are so disposed in the loess that their longer
axes are vertical, and they were evidently separated from the mass and
not deposited with it.




FIG. 218.—Diagrams to illustrate the effects of obstructions of
different types in arresting wind-driven sand. _a_, An unyielding
obstruction which permits the wind to pass through it; _b_, a flexible
and perforated obstruction; _c_, an unyielding closed barrier (after

=The wandering dunes.=—Over the broad expanse of the desert, sand
and dust, and occasionally gypsum from the saline deposits, are ever
migrating with the wind; on quiet days in the eddying “sand devils”,
but especially during the terrifying sand storms such as in the windy
season darken the air of northern China and southern Manchuria. This
drift of the sand is halted only when an obstruction is encountered—a
projecting rock, a bush, or a bunch of grass, or again the buildings of
a city or a town. The manner in which the sand is arrested by obstacles
of different kinds is of great interest and importance, and is utilized
in raising defenses against its encroachments. If the obstacle is
unyielding but allows some of the wind to pass through it, no eddies
are produced and the sand is deposited both to windward and to leeward
of the obstruction to form a fairly symmetrical mound (Fig. 218 _a_).
An obstruction which yields to the wind causes the sand to deposit
in a mound which is largely to leeward of the obstruction (Fig. 218
_b_). A solid wall, on the other hand, by inducing eddies, is at first
protected from the sand and mounds deposit both to windward and to
leeward (Fig. 218 _c_ and Fig. 219).

Except when held up by an obstruction, the drifting sand travels
to leeward in slowly migrating mounds or ridges which are known as
_dunes_. Their motion is due to the wind lifting the sand from the
windward side and carrying it over the crest, from where it slides down
the leeward slope and assumes a surface which is the angle of repose
of the material. In contrast with this the windward slope is notably
gradual, being shaped in conformity to the wind currents.


FIG. 219.—Sand accumulating both to windward and to leeward of a firm
and impenetrable obstruction. The wind comes from the left (after a
photograph by Bastin).]

The dunes which are raised upon seashores, like those of the desert,
are constantly migrating, those upon the shores of the North Sea at the
average rate of about twenty feet per year. Relentlessly they advance,
and despite all attempts to halt them, have many times overwhelmed the
villages along the coast. Upon the great barrier beach known as the
_Kurische Nehrung_, on the southeastern shore of the Baltic Sea, such
a burial of villages has more than once occurred, but as in the course
of time further migration of the dune has proceeded, the ruins of the
buried villages have been exhumed by this natural excavating process
(Fig. 220).


FIG. 220.—Successive diagrams to show how the town of Kunzen was
buried, and subsequently exhumed in the continued migration of a great
dune upon the Kurische Nehrung (after Behrendt).]

│  PLATE 7.                                                           │
│                                                                     │
│ [Illustration: _A._ Ranges of dunes upon the margin of the Colorado │
│ Desert (after Mendenhall).]                                         │
│                                                                     │
│ [Illustration: _B._ Sand dunes encroaching upon the oasis of Wed    │
│ Souf. Algeria (after T. H. Kearney).]                               │

=The forms of dunes.=—The forms assumed by dunes are dependent to
a very large extent upon the strength of the wind and the available
supply of sand. With small quantities of sand and with moderate winds,
sickle-shaped dunes known as _barchans_ (Fig. 221) are formed, whose
convex and flatter slopes are toward the wind and whose steep concave
leeward slopes are maintained at the angle of repose. The barchan is
shaped by the wind going both over and around the dune, constantly
removing sand from the windward side and depositing it to leeward.
With larger supplies of sand and winds which are not too violent a
series of barchans is built up, and these are arranged transversely to
the wind direction (Fig. 222 _b_). If the winds are more violent, the
minor depressions in the crests of the dunes become wind channels, and
the sand is then trailed out along them until the arrangement of the
ridges is parallel to the wind (Fig. 222 _c_). The surfaces of dunes
are generally marked by beautiful ripples in the sand, which, seen from
a little distance, may give the appearance of watered silk (plate 7 A).

[Illustration: FIG. 221.—View of desert barchans (after Haug).]


FIG. 222.—Diagrams to show the relationships in form and in
orientation of dunes to the supply of sand and to the strength of the
wind. _a_, barchans formed by small supplies of sand and moderate
winds; _b_, transverse dune ridges, formed when supply of sand is large
and winds are moderate; _c_, dune ridges formed with large sand supply
and violent winds (after Walther and Cornish).]

Under normal conditions dunes are not stationary but continue to wander
with the prevailing winds until they have reached the outer edge of
the zone of vegetation near the base of the foothills at the margin of
the desert. Here the grasses and other desert plants arrest the first
sand grains that reach them, and they continue to grow higher as the
sands accumulate. Some of the desert plants, like the yuccas, have so
adapted themselves to desert conditions that they may grow upward with
the sand for many feet and so keep their crowns above its surface.

=The cloudburst in the desert.=—Such clouds as enter the desert
through its mountain ramparts, and those derived from evaporation
from the hot desert soil, usually precipitate their moisture before
passing out of the basin. Above the highly heated floor the heavy
rain clouds are unable to drop their burden. The rain can sometimes
be seen descending, but long before it has reached the ground it has
again passed into vapor, and through repetition of this process the
clouds become so charged with moisture that when they encounter a
mountain wall and are thus forced to rise, there is a sudden downpour
not equaled in the humid regions. Desert rains are rare, but violent
beyond comparison. Often for a year or more there is no rainfall upon
the loose sand or porous clay, and the few plants which survive must
push their roots deep down until they have reached the zone of ground
water. When the clouds burst, each small cañon or _wed_ (pl. _wadi_)
within the mountain wall is quickly occupied by a swollen current
which carries a thick paste of sediment and drowns everything before
it. Ere it has flowed a mile, it may be that the water has disappeared
entirely, leaving a layer of mud and sand which rapidly dries out with
the reappearance of the sun.

[Illustration: FIG. 223.—Ideal section across the rising mountain wall
surrounding a desert and a part of the neighboring slope (after R. W.

As the mountains upon seacoasts are generally rising with reference to
the neighboring sea bottom, so the mountains which hem in the deserts
are generally growing upward with reference to the inclosed desert
floor. The marginal dislocations which separate the two are often in
evidence at the foot of the steep slope (Fig. 223), and these may even
appear as visible earthquake faults to indicate that the uplift is
more accelerated than the deposition along the mountain front.


FIG. 224.—Dry delta or alluvial fan at the foot of a mountain range
upon the borders of a desert.]

=The zone of the dwindling river.=—The rapid uplift so generally
characteristic of desert margins gives to the torrential streams which
develop after each cloudburst such an unusual velocity that when they
emerge from the mountain valleys on to the desert floor, the current
is suddenly checked and the burden of sediment in large part deposited
at the mouth of the valley so as to form a coarse delta deposit which
is described as a _dry delta_ (Fig. 224). Dependent upon its steepness
of slope, this delta is variously referred to as an _alluvial fan_
or _apron_, or as an _alluvial cone_. Over the conical slopes of the
delta surface the stream is broken up into numerous distributaries
which divide and subdivide as do the roots of a tree. In the Mohammedan
countries described as _wadi_, these distributaries upon dry deltas are
on the Pacific coast of the United States referred to as “washes” (Fig.

[Illustration: FIG. 225.—Map of the distributaries of neighboring
streams which emerge at the western base of the Sierra Nevadas in
California (after W. D. Johnson).]

Fast losing their velocity after emerging from the mountains, the
various distributaries drop first of all the heavy bowlders, then
the large pebbles and the sand, so that only the finer sand and the
silt are carried to the margin of the delta. As they enlarge their
boundaries, the neighboring deltas eventually coalesce and so form an
_alluvial bench_ or “gravel piedmont” at the foot of the range. Only
the larger streams are able to entirely cross this bench of parched
deposits with its coarsely porous structure, for the water is soon
sucked up by the thirsty materials. Encountering in its descent more
clayey layers, this water is conducted to the surface near the margin
of the bench and may there appear as a line of springs. At this level
there develops, therefore, a zone of vegetation, though there is no
local rain.

The alluvial bench grows upward by accretion of layers which are
thickest at the mountain end, so that the steepness of the bench
increases with time.

=Erosion in and about the desert.=—The violent cloudburst that is
characteristic of the arid lands is a most potent agent in modeling
the surface of the ground wherever the rock materials are not too
firmly coherent. Under the dash of the rain a peculiar type of “bad
land” topography is developed (plate 5 B and Fig. 226). Such a
rain-cut surface is a veritable maze of alternating gully and ridge, a
country worthless for agricultural purposes and offering the greatest
difficulty in the way of penetrating it. When composed of stiff clay
with scattered pebbles and bowlders, the effect of the “rain erosion”
is to fashion steep clay pillars each capped by a pebble and described
as “demoiselles” (Fig. 226).


FIG. 226.—A group of “demoiselles” in the “bad lands” (after a
photograph by Fairbanks).]

Behind the mountain front the valleys out of which the torrents are
discharged are usually short with steep side walls and a relatively
flat bottom, ending headward in an amphitheater with precipitous walls
(Fig. 227). In the western United States such valleys are referred to
as “box cañons”, but in Mohammedan countries the name “wed” applies to
the river valley within the mountains and to the distributaries as well.

=Characteristic features of the arid lands.=—It is characteristic of
erosion and deposition within humid regions that all outlines become
softened into flowing curves, due to the protective mat of vegetation.
In arid lands those massive rocks which are without structural planes
of separation, partly as a consequence of exfoliation, develop broad
domes which are projected upon the horizon as great semicircles,
broken in half it may be by displacement. The same massive rocks where
intersected by vertical joint planes yield, on the contrary, sharp
granite needles like those of Harney Peak (plate 8 A). Similarly,
schistose or bedded rocks, when tilted at a high angle, may yield forms
which are almost identical. Examples of such needles are found in the
Garden of the Gods in Colorado.

At lower levels, where the flying sand becomes effective as an eroding
agent, flat bedded rocks become etched into shelves and cornices, and
if intersected by joints, the shelves and cornices are transformed
into groups of castellated towers and pinnacles of a high degree of
ornamentation. These fantastic erosion remnants are usually referred to
as “chimneys” and may be seen in numbers in the bad lands of Dakota, as
they may in Colorado either in Monument Park or in the new Monolithic
National Park (plate 8 B).


FIG. 227.—Amphitheater at the head of the Wed Beni Sur (after

Where wind erosion plays a smaller rôle in the sculpture, but where
after an uplift a river has made its way, horizontally bedded rocks
are apt to be carved into broad _rock terraces_, nowhere shown upon so
grand a scale as about the Grand Cañon of the Colorado. Each harder
layer has here produced a floor or terrace which ends in a vertical
escarpment, and this is separated from the next lower layer of more
resistant rock by a slope of talus which largely hides the softer
intermediate beds. The great Desert of Sahara is shaped in a series of
rock terraces or steppes which descend toward the interior of the basin.

A single harder layer of resistant rock comes often to form the
flat capping of a plateau, and is then known as a _mesa_, or table
mountain. Along its front, detached outliers usually stand like
sentinels before the larger mass, and according to their relative
proportions, these are referred to either as small mesas or as the
smaller _buttes_ (Fig. 228).

[Illustration: FIG. 228.—Mesa and outlying butte in the Leucite Hills
of Wyoming (after Whitman Cross, U. S. G. S.).]

=The war of dune and oasis.=—In every desert the deposits are arranged
in consecutive belts or zones which are alternately the work of wind
and water. Surrounding the desert and upon the flanks of the mountain
wall there is found (1) the deposit of loess derived from the dust that
is carried out of the desert by the wind. Immediately within the desert
border at the base of the mountains is (2) the zone of the dwindling
river with its sloping bench of coarse rubble and gravel.

[Illustration: FIG. 229.—Flat-bottomed basin separating dunes—_bajir_
or _takyr_ (after Ellsworth Huntington).]

Next in order is (3) the belt of the flying sand, a zone of dune
ridges often separated by narrow, flat-bottomed basins (Fig. 229) into
which the strongest streams bring the finer sands and silt from the
mountains. Lastly, there is (4) the central sink or sinks, into which
all water not at once absorbed within the zone of alluviation or in
the zone of dunes is finally collected. Here are the true lacustrine
deposits of clay and separated salts (Fig. 230 and Fig. 207, p. 201).
The lake deposits fill in all the original irregularities of the desert
floor, out of which the tops of isolated ranges of mountains now
project like islands out of the surface of the sea. The several zones
of deposits in their order from the margin to the center of the desert
are given schematically in Fig. 231.

│  PLATE 8.                                                           │
│                                                                     │
│ [Illustration: _A._ The granite needles of Harney Peak in the Black │
│ Hills of South Dakota (after Darton).]                              │
│                                                                     │
│ [Illustration: _B._ Castellated erosion chimneys in El Cobra Cañon, │
│ New Mexico.                                                         │
│ (_Photograph by E. C. Case._)]                                      │


FIG. 230.—Billowy surface of the salt crust on the central sink in the
Lop Desert of central Asia (after Ellsworth Huntington).]

The zone of vegetation, as already stated, lies near the foot of
the alluvial bench, so that here are found the oases about which
have clustered the cities of the desert from the earliest records of
antiquity until now. Just without the line of oases is the wall of
dunes held back from further advance only by the vegetation which in
turn is dependent upon the rains in the neighboring mountains. With
every diminution in the water supply, the dunes advance and encroach
upon the oases (plate 7 B); while with every considerable increase in
this supply of moisture the alluvial bench advances over the dunes and
acquires a strip of their territory. Thus with varying fortunes a war
is continually waged between the withering river and the flying sand,
and the alternations of climate are later recorded in the dovetailing
together of the eolian and alluvial deposits at their common junction
(Fig. 231).

[Illustration: FIG. 231.—Schematic diagram to show the zones of
deposition in their order from the margin to the center of a desert.]


FIG. 232.—Mounds upon the site of the buried city of Nippur (after the
cast by Muret).]

In addition to the smaller periodic alternations of pluvial and
interpluvial climate—the “pulse of Asia”—the record of the Asiatic
deserts indicates a progressive desiccation of the entire region, which
has now given the victory to the dune. The ancient history of the
cities of the plains supplies the records of many that have been buried
in the dunes. To-day, where once were prosperous cities, nothing is to
be seen at the surface but a group of mounds (Fig. 232). Exhumed after
much painstaking labor, the walls and palaces of these ancient cities
have once more been brought to the light of day, and much has thus been
learned of the civilization of these early times (Fig. 233). Quite
recently the mounds which cover between one and two hundred buried
villages have been found upon the borders of the Tarim basin of central
Asia, where they were lost to history when they were overwhelmed in the
early centuries of the Christian Era.

[Illustration: FIG. 233.—Exhumed structures in the buried city of
Nippur (after Hilprecht).]

=The origin of the high plains which front the Rocky Mountains.=—To
the eastward of the great backbone of the North American continent
stretches a vast plain gently inclined away from the range and
generally known as the High Plains region (plate 9). The tourist who
travels westward by train ascends this slope so gradually that when he
has reached the mountain front it is difficult to realize that he has
climbed to an altitude of five thousand feet above the level of the
sea. That he has also passed through several climatic zones—a humid, a
semiarid, and an arid—and has now entered a semiarid district, is more
easily appreciated from study of the vegetation (Fig. 234). The surface
of the High Plains, where not cut into by rivers, is remarkably even,
so that it might be compared to the quiet surface of a great lake.

[Illustration: FIG. 234.—Section across the High Plains, showing the
position of the terrace and the climatic zones above it (after W. D.

The materials which compose the surface veneer of these plains are
coarse conglomerates, gravels, and sands, and the so-called “mortar
beds”, which are nothing but sands cemented into sandstone by carbonate
of lime. The pebbles in all these deposits are far-traveled and appear
to have been derived from erosion of those crystalline rocks which
compose the eastern front of the Rocky Mountains. These different
materials are not arranged in strictly parallel beds, as are the
deposits of a lake or sea; but the beds are made up of long threads of
lenticular cross section which are interlaced in the most intricate
fashion and which extend down the slope, or outward from the mountain
front (Fig. 235). It is thus shown that the High Plains are a bench or
plain of alluviation formed at the front of the Rocky Mountains during
an earlier series of pluvial periods, and that subsequent uplift has
produced the modern river valleys which are cut out of the plain. The
plexus of long threads of the coarser materials are the courses of
dwindling rivers which interlaced over the former plain, and which in
time were buried under other channel deposits of the same nature but in
different positions (Fig. 236). The pluvial periods in which this bench
was formed produced in other latitudes the great continental glaciers
which wrought such marvelous changes in northern North America and in
northern Europe.

[Illustration: FIG. 235.—Section across the great lenticular threads
of alluvial deposits which compose the veneer of the High Plains (after
W. D. Johnson).]

[Illustration: FIG. 236.—Distributaries of the foothills superimposed
upon an earlier series (after W. D. Johnson).]

=Character profiles.=—In contrast with the profiles in the landscapes
of humid regions (see Fig. 187, p. 177), those of arid lands are
marked by straighter elements (Fig. 237). Almost the only exception
of importance is furnished by the domes of massive granite monoliths,
which are sometimes broken in half by great displacements. Below
the horizon the secondary lines in the landscape betray the same
straightness of the component elements by the gabled slopes of talus
which are many times repeated so as almost to reproduce the lines in
a house of cards, since the sloping lines are maintained at the angle
of repose of the materials (Fig. 482, p. 443). Wherever the waves of
desert lakes have made an attack upon the rocks and have retired the
projecting spurs, other gables characterized by slightly different
slopes are introduced into the landscape.

[Illustration: FIG. 237.—Character profiles in the landscapes of arid

│   PLATE 9.                      │
│                                 │
│ [Illustration: THE HIGH PLAINS] │



 JOHANNES WALTHER. Das Gesetz der Wüstenbildung in Gegenwart und
 Vorzeit. Berlin, 1900, pp. 175, many plates. (This extremely valuable
 work is now out of print, but both a revised edition and an English
 translation are promised for 1912.)

 SIEGFRIED PASSARGE. Die Kalihari. Berlin, 1904, pp. 662.

 W. M. DAVIS. The Geographic Cycle in an Arid Climate, Jour. Geol.,
 vol. 13, 1905, pp. 381-407.

 ELLSWORTH HUNTINGTON. The Pulse of Asia. New York and Boston, 1907,
 pp. 415.

 SVEN HEDIN. Scientific Results of a Journey through Central Asia,
 1899-1900. Stockholm, 1904-1905, vols. 1 and 2, pp. 523 and 717, pls.
 56 and 76.

 JOSEPH BARRELL. Relative Geological Importance of Continental,
 Littoral and Marine Sedimentation, Jour. Geol., vol. 14, 1906, pp.
 316-356, 429-457, 524-568.

 E. F. GAUTIER. Études sahariennes, Ann. de Géogr., vol. 16, 1907, pp.
 46-69, 117-138.

The self-registering gauge of past climates:—

 G. K. GILBERT. Lake Bonneville, Mon. I, U. S. Geol. Surv., Chapter vi,
 pp. 214-318.

 T. F. JAMIESON. The Inland Seas and Salt Lakes of the Glacial Period,
 Geol. Mag. decade III, vol. 2, 1885, pp. 193-200.

 J. E. TALMAGE. The Great Salt Lake, Present and Past. Salt Lake City,
 1900, pp. 116, plates.

 E. HUNTINGTON. Some Characteristics of the Glacial Period in
 Non-glaciated Regions, Bull. Geol. Soc. Am., vol. 18, 1907, pp.
 351-388, pls. 31-39.

 T. C. CHAMBERLIN. The Future Habitability of the Earth, Rept.
 Smithson. Inst., 1910, pp. 371-389.

The red and brown desert varnish:—

 I. C. RUSSELL. Subaërial Decay of Rocks and Origin of the Red Color of
 Certain Formations. Bull. 52, U. S. Geol. Surv., 1889, pp. 65, pls. 5.

Erosion in the desert:—

 J. A. UDDEN. Erosion, Transportation, and Sedimentation performed by
 the Atmosphere, Jour. Geol., vol. 2, 1894, pp. 318-331.

 S. PASSARGE. Die pfannenförmigen Hohlformen der südafrikanischen
 Steppen, Pet. Mitt., vol. 57, 1911, pp. 57-61, 130-135.

The dust carried out of the desert:—

 F. VON RICHTOFEN. China, Ergebnisse eigene Reisen und darauf
 gegründeten Studien, Berlin, 1877, vol. 1, pp. 56-125.

 E. HILGARD. The Loess of the Mississippi Valley, Am. Jour. Sci., (3),
 vol. 18, 1879, pp. 106-112.

 T. C. CHAMBERLIN and R. D. SALISBURY. Preliminary Paper on the
 Driftless Area of the Upper Mississippi Valley, 6th Ann. Rept. U. S.
 Geol. Surv., 1885, pp. 278-307.

 E. E. FREE. The movement of soil material by the wind, with a
 bibliography of eolian geology by S. C. Stuntz and E. E. Free, Bull.
 68, U. S. Bureau of Soils, 1911, pp. 272, pls. 5.

 M. NEUMAYR. Erdgeschichte, vol. 1, pp. 510-514.

 E. DE MARTONNE. Géographie physique, pp. 663-668.


 VAUGHAN CORNISH. On the Formation of Sand-dunes, Geogr. Jour., vol. 9,
 1897, pp. 278-309 (a most important paper).

 F. SOLGER and Others. Dünenbuch. Enke, Stuttgart, 1910, pp. 373.

The zone of the dwindling river:—

 E. HUNTINGTON. The Border Belts of the Tarim Basin, Bull. Am. Geogr.
 Soc., vol. 38, 1906, pp. 91-96; The Pulse of Asia, pp. 210-222,

The war of dune and oasis:—

 R. PUMPELLY. Explorations in Turkestan, Expedition of 1904, etc., Pub.
 73, Carneg. Inst., Washington, vol. 1, pp. 1-13.

 E. HUNTINGTON. The Oasis of Kharga, Bull. Am. Geogr. Soc., vol. 42.
 1910, pp. 641-661.

 TH. H. KEARNEY. The Country of the Ant Men, Nat. Geogr. Mag., vol. 22,
 1911, pp. 367-382.

Features of the arid lands:—

 C. E. DUTTON. Tertiary History of the Grand Cañon District, with
 Atlas, Mon. II, U. S. Geol. Surv., 1882, pp. 264, pls. 42, maps 23.

 G. SWEINFURTH. Map Sheets of the Eastern Egyptian Desert. Berlin,
 1901-1902, 8 sheets.

The origin of the high plains:—

 W. D. JOHNSON. The High Plains and their Utilization, 21st Ann. Rept.
 U. S. Geol. Surv., Pt. iv, 1901, pp. 601-741.



=The weathering processes under control of the fracture system.=—In
an earlier chapter it was learned that the rocks which compose the
earth’s surface shell are intersected by a system of joint fractures
which in little-disturbed areas divide the surface beds into nearly
square perpendicular prisms (Fig. 36, p. 55), more or less modified
by additional diagonal joints, and often also by more disorderly
fractures. Throughout large areas these fractures may maintain nearly
constant directions, though either one or more of the master series
may be locally absent. This distinctive architecture of the surface
shell of the lithosphere has exercised its influence upon the various
weathering processes, as it has also upon the activities of running
water and of other less common transporting agencies at the surface.

Within high latitudes, where frost action is the dominant weathering
process, the water, by insinuating itself along the joints and
through repeated freezings, has broken down the rock in the immediate
neighborhood of these fractures, and so has impressed upon the surface
an image of the underlying pattern of structure lines (plate 10 A).

In much lower latitudes and in regions of insufficient rainfall, the
same structures are impressed upon the relief, but by other weathering
processes. In the case of the less coherent deposits in these
provinces, the initial forms of their erosional surface have sometimes
been determined by the dash of rain from the sudden cloudburst. Thus
the “bad lands” may have their initial gullies directed and spaced in
conformity with the underlying joint structures (Fig. 238).


FIG. 238.—Rain sculpturing under control by joints. Coast of southern
California (after a photograph by Fairbanks).]

In such portions of the temperate regions as are favored by a humid
climate, the mat of vegetation holds down a layer of soil, and mat
and soil in coöperation are effective in preventing any such large
measure of frostwork as is characteristic of the subpolar regions or
of high levels in the arid lands. In humid regions the rocks become a
prey especially to the processes of solution and accompanying chemical
decomposition, and these processes, although guided by the course of
the percolating ground water along the fracture planes, do not afford
such striking examples of the control of surface relief.

Those limestones which slowly pass into solution in the percolating
water do, however, quite generally indicate a localization of the
solution along the joint channels (Fig. 239 and plate 6 B). Though in
other rocks not so apparent, yet solutions generally take their courses
along the same channels, and upon them is localized the development
of the newly formed hydrated and carbonate minerals, as is well
illustrated by the phenomenon of spheroidal weathering (Fig. 155, p.


FIG. 239.—Outcrop of flaggy limestone which shows the effects of
solution along neighboring joints in a sagging of the upper beds (after
Gilbert, U. S. G. S.).]

=The fracture control of the drainage lines.=—The etching out of
the earth’s architectural plan in the surface relief, which we have
seen begun in the processes of weathering, is continued after the
transporting agents have become effective. It is often easy to see
that a river has taken its course in rectangular zigzags like the
elbows of a jointed stove pipe, and that its walls are formed of joint
planes from which an occasional squared buttress projects into the
channel. This structure is rendered in the plan of the Abisko Cañon of
northern Lapland (Fig. 240). We are later to learn that another great
transporting agent, the water wave, makes a selective attack upon the
lithosphere along the fractures of the joint system (Fig. 250, p. 233
and Fig. 254, p. 235).


FIG. 240.—Map of the joint-controlled Abisko Cañon in northern Lapland
(after Otto Sjögren).]

Where the scale of the example is large, as in the cases which have
been above cited, the actual position and directions of the joint wall
are easily compared with the near-by elements of the river’s course,
so that the connection of the drainage lines with the underlying
structure is at once apparent. In many examples where the scale
is small, the evidence for the controlling influence of the rock
structure in determining the courses of streams may be found in the
peculiar character of the drainage plan. To illustrate: the course of
the Zambesi River, within the gorge below the famous Victoria Falls,
not only makes repeated turnings at a right angle, but its tributary
streams, instead of making the usual sharp angle where they join the
main stream, also affect the right angle in their junctions (Fig. 241).


FIG. 241.—Map of the gorge of the Zambesi River below the Victoria
Falls (after Lamplugh).]

=The repeating pattern in drainage networks.=—It is a characteristic
of the joint system that the fractures within each series are spaced
with approximation to uniformity. If the plan of a drainage system has
been regulated in conformity with the architecture of the underlying
rock basement, the same repeating rectangles of the master joints may
be expected to appear in the lines of drainage—the so-called drainage

Such rectangular patterns do very generally appear in the drainage
network, though they are often masked upon modern maps by what, to
the geologist, seems impertinent intrusion of the black lines of
overprinting which indicate railways, lines of highway, and other
culture elements. On river maps, which are printed without culture, the
pattern is much more easily recognized (Figs. 242 and 243). Wherever
the relief is strong, as is the case in the Adirondack Mountain
province of the State of New York, individual hills may stand in relief
between the bounding streams which compose the rectangular network,
like the squared pedestals of monuments. Such a type of relief carved
in repeating patterns has been described as “checkerboard topography.”


FIG. 242.—Controlled drainage network of the Shepaug River in


FIG. 243.—A river network of repeating rectangular pattern. Near Lake
Temiskaming, Ontario (from the map by the Dominion Government).]

=The dividing lines of the relief patterns—lineaments.=—The repeating
design outlined in the river network of the Temiskaming district
(Fig. 243) would appear in greater perfection if we could reproduce
the relief without at the same time obscuring the lines of drainage;
for where the pattern is not completely closed by the course of the
stream, there is generally found either a dry valley or a ravine to
complete the design. If these are not present, a bit of straight
coast line, a visible line of fracture, a zone of fault breccia, or
the boundary line separating different formations may one or more of
them fill in the gaps of the parallel straight drainage lines which by
their intersection bring out the pattern. These significant lines of
landscapes which reveal the hidden architecture of the rock basement
are described as _lineaments_ (Fig. 82, p. 87). They are the character
lines of the earth’s physiognomy.

It is important to emphasize the essentially composite expression of
the lineament. At one locality it appears as a drainage line, a little
farther on it may be a line of coast; then, again, it is a series of
aligned waterfalls, a visible fault trace, or a rectilinear boundary
between formations; but in every case it is some surface expression
of a buried fracture. Hidden as they so generally are, the fracture
lines must be searched out by every means at our disposal, if we are
not to be misled in accounting for the positions and the attitudes of
disturbed rock masses.

As we have learned, during earthquake shocks, as at no other time,
the surface of the earth is so sensitized as to betray the position
of its buried fractures. As the boundaries of orographic blocks,
certain of the fractures are at such times the seats of especially
heavy vibrations; they are the seismotectonic lines of the earthquake
province. Many lineaments are identical with seismotectonic lines, and
they therefore afford a means of to some extent determining in advance
the lines of greatest danger from earthquake shock.

=The composite repeating patterns of the higher orders.=—Not only
do the larger joint blocks become impressed upon the earth relief as
repeating diaper patterns, but larger and still larger composite units
of the same type may, in favorable districts, be found to present the
same characters. Attention has already been more than once directed
to the fact that the more perfect and prominent fracture planes recur
among the joints of any series at more or less regular intervals (Fig.
40, p. 57, and Fig. 41, p. 58). Nowhere, perhaps, is this larger order
of the repeating pattern more perfectly exemplified than in some recent
deposits in the Syrian desert (plate 10 B). It is usually, however,
in the older sediments that such structures may be recognized; as, for
example, in the squared towers and buttresses of the Tyrolean Dolomites
(Fig. 244). Here the larger blocks appear in the thick bedded lower
formation, the dolomite, divided into subordinate sections of large
dimensions; but in the overlying formations in blocks of relatively
small size, yet with similarly perfect subequal spacing.


FIG. 244.—Squared mountain masses which reveal a distribution of the
joints in block patterns of different orders of magnitude. The Pordoi
range of the Sella group of the Dolomites, seen from the Cima di Rossi
(after Mojsisovics).]

The observing traveler who is privileged to make the journey by
steamer, threading its course in and out among the many islands and
skerries of the Norwegian coast, will hardly fail to be struck by the
remarkable profiles of most of the lower islands (Fig. 245). These
profiles are generally convexly scalloped with a noteworthy regularity,
and not in one unit only, but in at least two with one a multiple of
the other (Fig. 246). As the steamer passes near to the islands, it is
discovered that the smaller recognizable units in the island profiles
are separated by widely gaping joints which do not, however, belong
to the unit series, but to a larger composite group (Fig. 246 _b_).
Frostwork, which depends for its action upon open spaces within the
rocks, has here been the cause of the excessive weathering above the
more widely gaping joints.

│   PLATE 10.                                                          │
│                                                                      │
│ [Illustration: _A._ View in Spitzbergen to illustrate the            │
│ disintegration of rock under the control of joints.                  │
│ (_Photograph by O. Haldin._)]                                        │
│                                                                      │
│ [Illustration: _B._ Composite pattern of the joint structures within │
│ recent alluvial deposits.                                            │
│ (_Photograph by Ellsworth Huntington._)]                             │


FIG. 245.—Island groups of the Lofoten archipelago off the northwest
coast of Norway, which reveal repeating patterns of the relief in two
orders of magnitude (after a photograph by Knudsen).]


FIG. 246.—Diagrams to illustrate the composite profiles of the islands
on the Norwegian coast. _a_, distant view; _b_, near view, showing the
individual joints and the more widely gaping fractures beneath each sag
in the profile.]

High northern latitudes are thus especially favorable for revealing all
the details in the architectural pattern of the lithosphere shell, and
we need not be surprised that when the modern maps of the Norwegian
coast are examined, still larger repeating patterns than any that may
be seen in the field are to be made out. The Norwegian coast was long
ago shown to be a complexly faulted region, and these larger divisions
of the relief pattern, instead of being explained as a consequence
solely of selective weathering, must be regarded as due largely to
fault displacements of the type represented in our model (plate 4 C).
Yet whether due to displacements or to the more numerous joints, all
belong to the same composite system of fractures expressed in the


 WILLIAM H. HOBBS. The River System of Connecticut, Jour. Geol.,
 vol. 9, 1901, pp. 469-485, pl. 1; Lineaments of the Atlantic Border
 Region, Bull. Geol. Soc. Am., vol. 15, 1904, pp. 483-506, pls. 45-47;
 The Correlation of Fracture Systems and the Evidences for Planetary
 Dislocations within the Earth’s Crust, Trans. Wis. Acad. Sci., etc.,
 vol. 15, 1905, pp. 15-29; Repeating Patterns in the Relief and in
 the Structure of the Land, Bull. Geol. Soc. Am., vol. 22, 1911, pp.
 123-176, pls. 7-13.



=The motion of a water wave.=—The motions within a wave upon the
surface of a body of water may be thought of in two different ways.
First of all, there is the motion of each particle of water within
an orbit of its own; and there is, further, the forward motion of
propagation of the wave considered as a whole.

[Illustration: FIG. 247.—Diagram to show the nature of the motions
within a free water wave.]

The water particle in a wave has a continued motion round and round its
orbit like that of a horse circling a race course, only that here the
track is in a vertical plane, directed along the line of propagation of
the wave (Fig. 247). Each particle of water, through its friction upon
neighboring particles, is able to transmit its motion both along the
surface and downward into the water below. The force which starts the
water in motion and develops the wave, is the friction of wind blowing
over the water surface, and the size of the orbit of the water particle
at any point is proportional to the wind’s force and to the stretch
of water over which it has blown. The wind’s effect is, therefore,
cumulative—the wave is proportional to the wind’s effect upon all
water particles in its rear, added to the local wind friction.

The size or _height_ of the wave is measured by the diameter of the
orbit of motion of the surface particle, and this is the difference in
height between trough and crest. The distance from crest to crest, or
from trough to trough, is called the _wave length_. Though the wave
motion is transmitted downward into the water there is a continued
loss of energy which is here not compensated by added wind friction,
and so the orbital motion grows smaller and smaller, until at the
depth of about a wave length it has completely died out. This level
of no motion is called the _wave base_. In quiet weather the level of
no motion is practically at the water’s surface, and inasmuch as the
geological work of waves is in large part accomplished during the great
storms, the term “wave base” refers to the lowest level of wave motion
at the time of the heaviest storms. Upon the ocean the highest waves
that have been measured have an amplitude of about fifty feet and a
wave length of about six hundred feet.

=Free waves and breakers.=—So long as the depth of the water is below
wave base, there is obviously no possibility of interference with the
wave through friction upon the bottom. Under these conditions waves are
described as _free waves_, and their forms are symmetrical except in so
far as their crests are pulled over and more or less dissipated in the
spray of the “white caps” at the time of high winds.

[Illustration: FIG. 248.—Diagram to illustrate the transformation of a
free wave into a breaker as it approaches the shore.]

As a wave approaches a shore, which generally has a gentle outward
sloping surface, there is interposed in the way of a free forward
movement the friction upon the bottom. This friction begins when the
depth of water is less than wave base, and its effect is to hold back
the wave at the bottom. Carried slowly upward in the water by the
friction of particle upon particle, the effect of this holding back
is a piling up of the water, which increases the wave height as it
diminishes the wave length, and also interferes with wave symmetry
(Fig. 248). Moving forward at the top under its inertia of motion and
held back at the bottom by constantly increasing friction, a strong
turning motion or couple is started about a horizontal axis, the
immediate effect of which is to steepen the forward slope of the wave,
and this continues until it overhangs, and, falling, “breaks” into
surf. Such a breaking wave is called a “comber” or “breaker” (plate 11

│   PLATE 11.                                                     │
│                                                                 │
│ [Illustration: _A._ Ripple markings within an ancient sandstone │
│ (courtesy of U. S. Grant).]                                     │
│                                                                 │
│ [Illustration: _B._ A wave breaking as it approaches the shore. │
│ (_Photograph by Fairbanks._)]                                   │


FIG. 249.—Notched rock cliff cut by waves and the fallen blocks
derived from the cliff through undermining. Profile Rock at Farwell’s
Point near Madison, Wisconsin.]

=Effect of the breaking wave upon a steep rocky shore—the notched
cliff.=—If the shore rises abruptly from deeper water, the top of
the breaking wave is hurled against the cliff with the force of a
battering ram. During storms the water of shore waves is charged with
sand, and each sand particle is driven like a stone cutter’s tool
under the stroke of his hammer. The effect is thus both to chip and
to batter away the rock of the shore to the height reached by the
wave, undermining it and notching the rock at its base (Fig. 249).
When the notch has been cut in this manner to a sufficient depth, the
overhanging rock falls by its own weight in blocks which are bounded
by the ever present joints, leaving the upper cliff face essentially


FIG. 250.—A wave-cut chasm under control by joints, coast of Maine
(after Tarr).]

=Coves, sea arches, and stacks.=—It is the headland which is most
exposed to the work of the waves, since with change of wind direction
it is exposed upon more than a single face. The study of headlands
which have been cut by waves shows that the joints within the rock
play a large rôle in the shaping of shore features. The attack of the
waves under the direction of these planes of ready separation opens
out indentations of the shore (Fig. 250) or forms _sea caves_ which, as
they extend to the top of the cliff by the process of sapping, yield
the _coves_ which are so common a feature upon our rock-bound shores
(Fig. 259, p. 238). With continuation of this process, the caves formed
on opposite sides of the headland may be united to form a _sea arch_
(Fig. 251).


FIG. 251.—The sea arch known as the Grand Arch upon one of the
Apostle Islands in Lake Superior (after a photograph by the Detroit
Photographic Company).]

A later stage in this selective wave carving under the control of
joints is reached when the bridge above the arch has fallen in, leaving
a detached rock island with precipitous walls. Such an offshore island
of rock with precipitous sides is known as a _stack_ (Fig. 252), or
sometimes as a “chimney”, though this latter term is best restricted to
other and similar forms which are the product of selective weathering
(p. 300).

[Illustration: FIG. 252.—Stack near the shore of Lake Superior.]

Whenever the rock is less firmly consolidated, and so does not stand
upon such steep planes, the stack is apt to have a more conical form,
and may not be preceded in its formation by the development of the
sea arch (Fig. 260, p. 239). In the reverse case, or where the rock
possesses an unusual tenacity, the stack may be largely undermined
and stand supported like a table upon thick legs or pillars of rock
(Fig. 253). In Fig. 254 is seen a group of stacks upon the coast of
California, which show with clearness the control of the joints in
their formation, but unlike the marble of the South American example
the forms are not rounded, but retain their sharp angles.


FIG. 253.—The Marble Islands, stacks in Lake Buenos Aires, southern
Andes (after F. P. Moreno).]

=The cut rock terrace.=—When waves first begin their attack upon a
steep, rocky shore, the lower limit of the action is near the wave
base. The action at this depth is, however, less efficient, and as the
recession of the cliff is one of the most rapid of erosional processes,
the rock floor outside the receding cliff comes to slope gradually
downward from the cliff to a maximum depth at the edge of the terrace,
approximately equal to wave base (Fig. 255). This cut terrace is
extended seaward or lakeward, as the case may be, in a _built terrace_
constructed from a portion of the rock débris acquired from the cliff.


FIG. 254.—Squared stacks which reveal the position of the joint planes
which have controlled in the process of carving by the waves. Pt.
Buchon, California (after a photograph by Fairbanks).]


FIG. 255.—Ideal section of a steep rocky shore carved by waves into a
notched cliff and cut terrace, and extended by a built terrace.]

The broken wave, after rising upon the terrace under the inertia of
its motion until all its energy has been dissipated, slides outward
by gravity, and though checked and overridden by succeeding breakers,
it continues its outward slide as the “undertow” until it reaches the
end of the terrace. Here it suddenly enters deep water, and losing its
velocity, drops its burden of rock, and builds the terrace seaward
after the manner of construction of an embankment. As we are to see,
the larger portion of the wave-quarried material is diverted to a
different quarter.


FIG. 256.—Map showing the outlines of the Island of Heligoland at
different stages in its recent history. The peripheries given are in

To gain some conception of the importance of wave cutting as an eroding
process, we may consider the late history of Heligoland, a sandstone
island off the mouth of the Elbe in the North Sea (Fig. 256). From a
periphery of 120 miles, which it possessed in the ninth century of the
Christian era, the island has reduced its outline to 45 miles in the
fourteenth century, 8 miles in the seventeenth, and to about 3 miles at
the beginning of the twentieth century. The German government, which
recently acquired this little remnant from England, has expended large
sums of money in an effort to save this last relic.


FIG. 257.—Cut and built terrace with bowlder pavement shaped by waves
on a steep shore formed of loose materials.]

=The cut and built terrace on a steep shore of loose materials.=—In
materials which lack the coherence of firm rock, no vertical cliff can
form; for as fast as undermined by the waves the loose materials slide
down and assume a surface of practically constant slope—the “angle of
repose” of the materials (Fig. 257). The terrace below this sloping
cliff will not differ in shape from that cut upon a rocky shore; but
whenever the materials of the shore include disseminated blocks too
large for the waves to handle, they collect upon the terrace near where
they have been exhumed, thus forming what has been called a “bowlder
pavement” (Fig. 258).


FIG. 258.—Sloping cliff and terrace with bowlder pavement exposed at
low tide upon the shore at Scituate, Massachusetts.]

The edge of the cut and built terrace is, as already mentioned,
maintained at the depth of wave base. If one will study the submerged
contours of any of our inland lakes, it will be found that these
basins are surrounded by a gently sloping marginal shelf,—the cut and
built terrace,—and that the depth of this shelf at its outer edge is
proportioned to the size of the lake. Upon Lake Mendota at Madison,
Wisconsin, the large storm waves have a length of about twenty feet,
which is the depth of the outer edge of the shore terraces (Fig. 267,
p. 242). The shelf surrounding the continents has, with few local
exceptions, a uniform depth of 100 fathoms, or about the wave base of
the heaviest storm waves.

=The work of the shore current.=—In describing the formation of
the built terrace, it was stated that the greater part of the rock
material quarried upon headlands by the waves is diverted from the
offshore terrace. This diversion is the work of the shore current
produced by the wave.

[Illustration: FIG. 259.—Map to show the nature of the shore current
and the forms which are molded by it.]

At but few places upon a shore will the storm waves beat
perpendicularly, and then for but short periods only. The broken wave,
as it crawls ever more slowly up the beach, carries the sand with it in
a sweeping curve, and by the time gravity has put a stop to its forward
movement, it is directed for a brief instant parallel to the shore.
Soon, however, the pull of gravity upon it has started the backward
journey in a more direct course down the slope of the terrace; and
here encountering the next succeeding breaker, a portion of the water
and the coarser sand particles with it are again carried forward for a
repetition of the zigzag journey. This many times interrupted movement
of the sand particles may be observed during a high wind upon any sandy
lee shore. The “set” of the water along the shore as a result of its
zigzag journeyings is described as the _shore current_ (Fig. 259),
and the effect upon sand distribution is the same as though a steady
current moved parallel to the shore in the direction of the average
trend of the moving particles.

=The sand beach.=—The first effect of the shore current is to deposit
some portion of the sand within the first slight recess upon the shore
in the lee of the cliff. The earlier deposits near the cliff gradually
force the shore current farther from the shore and so lay down a sand
selvage to the shore, which is shaped in the form of an arc or crescent
and known as a _beach_ (Fig. 259 and Fig. 260).

[Illustration: FIG. 260.—Crescent-shaped beach formed in the lee of
a headland. Santa Catalina Island, California (after a photograph by

[Illustration: FIG. 261.—Cross section of a beach pebble.]

=The shingle beach.=—With heavy storms and an exceptional reach of the
waves, the shore currents are competent to move, not the sand alone,
but pebbles, the area of whose broader surface may be as great as the
palm of one’s hand. Such rock fragments are shaped by the continued
wear against their neighbors under the restless breakers, until they
have a lenticular or watch-shaped form (Fig. 261). Such beach pebbles
are described as _shingle_, and they are usually built up into distinct
ridges upon the shore, which, under the fury of the high breakers, may
be piled several feet above the level of quiet water (Fig. 262). Such
storm beaches have a gentle forward slope graded by the shore current,
but a steep backward slope on the angle of repose. Most storm beaches
have been largely shaped by the last great storm, such as comes only at
intervals of a number of years.


FIG. 262.—Storm beach of coarse shingle about four feet in height
at the base of Burnt Bluff on the northeast shore of Green Bay, Lake

=Bar, spit, and barrier.=—Wherever the shore upon which a beach is
building makes a sudden landward turn at the entrance to a bay, the
shore currents, by virtue of their inertia of motion, are unable longer
to follow the shore. The débris which they carry is thus transported
into deeper water in a direction corresponding to a continuation of
the shore just before the point of turning (see Fig. 259, p. 238).
The result is the formation of a _bar_, which rises to near the
water surface and is extended across the entrance to the bay through
continued growth at its end, after the manner of constructing a railway
embankment across a valley.

[Illustration: FIG. 263.—Spit of shingle on Au Train Island, Lake
Superior (after Gilbert).]

Over the deeper water near the bar the waves are at first not generally
halted and broken, as they are upon the shore, and so the bar does not
at once build itself to the surface, but remains an invisible bar to
navigation. From its shoreward end, however, the waves of even moderate
storms are broken, and the bar is there built above the water surface,
where it appears as a narrow cape of sand or shingle which gradually
thins in approaching its apex. This feature is the well-known _spit_
(Fig. 263) which, as it grows across the entrance to the bay, becomes a
_barrier_ or _barrier beach_ (Fig. 264).

The continuation of the visible in the usually invisible bar, is at the
time of high winds made strikingly apparent, for the wave base is below
the crest of the bar, and at such times its crescentic course beyond
the spit can be followed by the eye in a white arc of broken water.


FIG. 264.—Barrier beach in front of a lagoon on Lake Mendota at
Madison, Wisconsin. The shallow lagoon behind the barrier is filling up
and is largely hidden in vegetation.]

The construction of a barrier across the entrance to a bay transforms
the latter into a separate body of water, a lagoon, within which
silting up and peat formation usually lead to an early extinction
(p. 429). The formation of barriers thus tends to straighten out
the irregularities of coast lines, and opens the way to a natural
enlargement of the land areas. While the coasts of the United Kingdom
of Great Britain have been losing some four thousand acres through wave
erosion, there has been a gain by growth in quiet lagoons which amounts
to nearly seven times that amount. As evidence of the straightening
of the shore line which results from this process, the coast of the
Carolinas or of Nantucket (Fig. 459) may serve for illustration.

=The land-tied island.=—We have seen that wave erosion operates to
separate small islands from the headlands, but the shore currents
counteract this loss to the continents by throwing out barriers which
join many separated islands to the mainland. Such land-tied islands are
a common feature on many rocky coasts, and upon the New England coast
they usually have been given the name of “neck.” The long arc of Lynn
Beach joins the former island of Nahant, through its smaller neighbor
Little Nahant, to the coast of Massachusetts. A similar land-tied
island is Marblehead Neck. The Rock of Gibraltar, formerly an island,
is now joined to Spain by the low beach known as the “neutral ground.”
The Spanish name, _tombola_, has sometimes been employed to describe an
island thus connected to the shore.

[Illustration: FIG. 265.—Cross section of a barrier beach with lagoon
in its rear.]

=A barrier series.=—The cross section of a barrier beach, like that of
a storm beach upon the shore, slopes gently upon the forward side, and
more steeply at the angle, of repose upon the rear or landward margin
(Fig. 265). The thinning wedge of shore deposits which the barrier
throws out to seaward raises the level of the lake bottom (Fig. 266),
and when coast irregularities are favorable to it, new spits will
develop upon the shore outside the earlier one, and a new bar, and in
its turn a barrier, will be found outside the initial one, taking a
course in a direction more or less parallel to it (Fig. 267).

[Illustration: FIG. 266.—Cross section of a series of barriers and an
outer bar.]


FIG. 267.—Formation of barrier series and an outer bar in University
Bay of Lake Mendota, at Madison, Wisconsin. The water contour interval
is five feet, and the land contour interval ten feet (based on a map by
the Wisconsin Geological Survey).]

[Illustration: FIG. 268.—Series of barriers at the western end of Lake
Superior (after Gilbert).]

So soon as the first barrier is formed, processes are set in operation
which tend to transform the newly formed lagoon into land, and so with
a series of barriers, a zone of water lilies between the outer barrier
and the bar, a bog, and a land platform may represent the successive
stages in this acquisition of territory by the lands. A noteworthy
example of barrier series and extension of the land behind them, is
afforded by the bay at the western end of Lake Superior (Fig. 268).

[Illustration: FIG. 269.—Character profiles resulting from wave action
upon shores.]

=Character profiles.=—The character profiles yielded by the work of
waves are easy of recognition (Fig. 269). The vertical cliff with notch
at its base is varied by the stack of sugar-loaf form carved in softer
rocks, or the steeper notched variety cut from harder masses. Sea caves
and sea arches yield variations of a curve common to the undercut
forms. Wherever the materials of the shore are loosely consolidated
only, the sloping cliff is formed at the angle of repose of the
materials. The barrier beach, though projecting but a short distance
above the waves, shows an unsymmetrical curve of cross section with the
steeper slope toward the land.


 G. K. GILBERT. The Topographic Features of Lake Shores, 5th Ann. Rept.
 U. S. Geol. Surv., 1885, pp. 69-123, pls. 3-20; Lake Bonneville, Mon.
 I, U. S. Geol. Surv., 1890, Chapters ii-iv, pp. 23-187.

 VAUGHAN CORNISH. On Sea Beaches and Sand Banks, Geogr. Jour., vol. 11,
 1898, pp. 528-543, 628-658.

 F. P. GULLIVER. Shore Line Topography, Proc. Am. Acad. Arts and Sci.,
 vol. 34, 1899, pp. 149-258.

 N. S. SHALER. The Geological History of Harbors, 13th Ann. Rept. U. S.
 Geol. Surv., 1893, pp. 93-209.

 SIR A. GEIKIE. The Scenery of Scotland, 1901, pp. 46-89.

 W. H. WHEELER. The Sea Coast. Longmans, London, 1902, pp. 1-78.

 G. W. VON ZAHN. Die zerstörende Arbeit des Meeres an Steilküsten nach
 Beobachtungen in der Bretagne und Normandie in den Jahren 1907 und
 1908, Mitt. d. Geogr. Ges. Hamb., vol. 24, 1910, pp. 193-284, pls.



=The characters in which the record has been preserved.=—The peculiar
forms into which the sea has etched and molded its shores have been
considered in the last chapter. Of these the more significant are
the notched rock cliff, the cut rock terrace, the sea cave, the sea
arch, the stack, and the sloping cliff and terrace, among the carved
features; and the barrier beach and built terrace, among the molded
forms. It is important to remember that the molded forms, by the very
manner of their formation, stand in a definite relationship to the
carved features; so that when either one has been in part effaced and
made difficult of determination, the discovery of the other in its
correct natural position may remove all doubt as to the origin of the


FIG. 270.—The east coast of Florida, with shore line characteristic of
a raised coast.]

In studies of the change of level of the land, it is customary to refer
all variations to the sea level as a zero plane of reference. It is not
on this account necessary to assume that the changes measured from this
arbitrary datum plane are the absolute upward or downward oscillations
which would be measured from the earth’s center; for the sea, like the
land, has been subject to its changes of level. There need, however, be
no apology for the use of the sea surface as a plane of reference; for
it is all that we have available for the purpose, and the changes in
level, even if relative only, are of the greatest significance. It is
probable that in most cases where the coast line is rising from uplift,
some portion of the sea basin not far distant is becoming deepened,
so that the visible change of level is the algebraic sum of the two

=Even coast line the mark of uplift.=—It was early pointed out in
this volume (p. 158) that the floor of the sea in the neighborhood of
the land presents a relatively even surface. The carving by waves,
combined with the process of deposition of sediments, tends to fill up
the minor irregularities of surface and preserve only the features of
larger scale, and these in much softened outlines. Upon the continents,
on the contrary, the running water, taking advantage of every slight
difference in elevation and searching out the hidden structure planes
within the rock, soon etches out a surface of the most intricate
detail. The effect of elevation of the sea floor into the light of day
will therefore be to produce an even shore line devoid of harbors (Fig.
270). If the coast has risen along visible planes of faulting near the
sea margin, the coast line, in addition to being even, will usually be
made up of notably straight elements joined to one another.


FIG. 271.—Ragged coast line of Alaska, the effect of subsidence.]

=A ragged coast line the mark of subsidence.=—When in place of uplift
a subsidence occurs upon the coast, the intricately etched surface,
resulting from erosion beneath the sky, comes to be invaded by the sea
along each trench and furrow, so that a most ragged outline is the
result (Fig. 271). Such a coast has many harbors, while the uplifted
coast is as remarkable for its lack of them.

=Slow uplift of the coast—the coastal plain and cuesta.=—A gradual
uplift of the coast is made apparent in a progressive retirement of the
sea across a portion of its floor, thus exposing this even surface of
recent sediments. The former shore land will be easily recognized by
it’s etched surface, which will now come into sharp contrast with the
new plain. It is therefore referred to as the _oldland_ and the newly
exposed _coastal plain_ as the _newland_ (Fig. 272).


FIG. 272.—Portion of Atlantic coastal plain and neighboring oldland of
the Appalachian Mountains.]

But the near-shore deposits upon the sea floor had an initial dip
or slope to seaward, and this inclination has been increased in the
process of uplift. The streams from the oldland have trenched their way
across these deposits while the shore was rising. But the process being
a slow one, deposits have formed upon the seaward side of the plain
after the landward portion was above tide, and the coastal plain may
come to have a “belted” or zoned character. The streams tributary to
those larger ones which have trenched the plain may encounter in turn
harder and softer layers of the plain deposits, and at each hard layer
will be deflected along its margin so as to enter the main streams
more nearly at right angles. They will also, as time goes on, migrate
laterally seaward through undermining of the harder layers, and thus
will be shaped alternating belts of lowland separated by escarpments
in the harder rock from the residual higher slopes. Belts of upland of
this character upon a coastal plain are called _cuestas_ (Fig. 273).


FIG. 273.—Ideal form of cuestas and intermediate lowlands carved from
a coastal plain (after Davis).]

=The sudden uplifts of the coasts.=—Elevations of the coast which
yield the coastal plain must be accounted among the slower earth
movements that result in changes of level. Such movements, instead of
being accompanied by disastrous earthquakes, were probably marked by
frequent slight shocks only, by subterranean rumblings, or, it may be,
the land rose gradually without manifestations of a sensible character.

Upon those coasts which are often in the throes of seismic disturbance,
a quite different effect is to be observed. Here within the rocks
we will probably find the marks of recent faulting with large
displacements, and the movements have been upon such a scale that shore
features, little modified by subsequent weathering, stand well above
the present level of the seas. Above such coasts, then, we recognize
the characteristic marks of wave action, and the evidence that they
have been suddenly uplifted is beyond question.

[Illustration: FIG. 274.—Uplifted sea cave, ten feet above the water
upon the coast of California; the monument to a former earthquake
(after a photograph by Fairbanks).]

[Illustration: FIG. 275.—Double-notched cliff near Cape Tiro, Celebes
(after a photograph by Sarasin).]

=The upraised cliff.=—Upon the coast of southern California may be
found all the features of wave-cut shores now in perfect preservation,
and in some cases as much as fifteen hundred feet above the level of
the sea. These features are monuments to the grandest of earthquake
disturbances which in recent time have visited the region (Fig.
274). Quite as striking an example of similar movements is afforded
by notched cliffs in hard limestone upon the shore of the Island of
Celebes (Fig. 275). But the coast of California furnishes the other
characteristic coast features in the high sea arch and the stack
as additional monuments to the recent uplift. Let one but imagine
the stacks which now form the Seal Rocks off the Cliff House at San
Francisco to be suddenly raised high above the sea, and the forms which
they would then present would differ but little from those which are
shown in Fig. 276.

[Illustration: FIG. 276.—Jasper rock stacks uplifted on the coast of
California (after a photograph by Fairbanks).]

=The uplifted barrier beach.=—Within the reëntrants of the shore, the
wave-cut cliff is, as we know, replaced by the barrier beach, which
takes its course across the entrance to a bay. After an uplift, such
a barrier composed of sand or shingle should be connected with the
headlands, often with a partially filled lagoon behind it. Its cross
section should be steep in the direction of the lagoon, but quite
gradual in front (Fig. 277).

[Illustration: FIG. 277.—Uplifted shingle beach across the entrance to
a former bay upon the coast of southern California (after a photograph
by Fairbanks).]

[Illustration: FIG. 278.—Raised beach terraces near Elie, Fife,

=Coast terraces.=—Upon those shores where to-day high mountains
front the sea, the coast may generally be seen to rise in a series of
terraces (Fig. 278). This is notably true of those coasts which are
to-day racked by earthquakes, such as is the eastern margin of the
Pacific from Alaska to Patagonia. The traveler by steamer along the
coast from San Francisco to Chili has for weeks almost constantly in
sight these giant steps on which the mountains have been uplifted from
the sea. In Alaska we are fortunate in having the history of the later
stages in this uplift (Fig. 279). As described in a former chapter,
portions of this shore rose in the month of September of the year 1899
in some places as high as forty-seven feet, to the accompaniment of
a terrific earthquake and sea wave. Above the terrace which marks
the beach line of 1899 there is a higher terrace of similar form now
overgrown with trees, but none the less clearly to be recognized as a
shore line of the past century which preceded in the long sequence the
uplift of 1899.

[Illustration: FIG. 279.—Uplifted sea cliffs and terraces on the coast
of Russell Fjord, Alaska (after Tarr and Martin).]

[Illustration: FIG. 280.—Diagrams to show how excessive sinking
upon the sea floor will cause the shore to migrate landward as it is


FIG. 281.—A drowned river mouth, or estuary upon a coastal plain.]

As was noted in our study of earthquakes, the recent instrumental
records of distant earthquakes tell us that the movements upon the sea
floor are many times larger than those upon the continents, and that
while the mountainous coasts are generally rising, the deeps of the sea
are sinking. The effect of this over-balance of sinking, or resultant
shrinking of the earth’s shell, may be to compress the mountain
district and so cause the shore line to move landward at the same time
that it moves upward (Fig. 280).

=The sunk or embayed coast.=—When now, upon the other hand, a section
of the coast line sinks with reference to the sea, the water invades
all the near-shore valleys, thus “drowning” them and yielding the
drowned river mouth or _estuary_. If the relief of the shore was
slight, as it generally is upon a coastal plain, slight depression only
will produce broad estuaries, such as Chesapeake Bay at the drowned
mouth of the Susquehanna (Fig. 281).

If, on the other hand, the relief of the shore is strong and the
subsidence is large, the entire coast line will be transformed into
an archipelago of steep-walled rocky islets which rise abruptly from
the sea (Figs. 282 and 284). A plateau which is intersected by deep
and steep-walled valleys of U-section (p. 341) under large submergence
yields the _fjords_ so characteristic of Scandinavia or Alaska. A
ragged coast line, fringed with islands as a result of submergence, is
described as an _embayed coast_.


FIG. 282.—Archipelago of steep rocky islets due to large submergence
of a coast having strong relief. Entrance to Esquimalt Harbor,
Vancouver Island (after a photograph by Fairbanks).]


FIG. 283.—The submerged Hudsonian channel which continues the Hudson
River across the continental shelf.]

=Submerged river channels.=—The sinking of a coast of small relief be
sufficient to completely submerge river valleys, whose channels then
begin to fill with sediment and whose courses can only be followed in
soundings. One of the most interesting of such channels is that which
continues the Hudson River across the continental shelf into the deeper
sea (Fig. 283).

=Records of an oscillation of movement.=—Because a coast is deeply
embayed is no ground for assuming that a subsidence is now in progress,
or is, in fact, the latest movement recorded upon the coast. In many
cases it is easy to see that such is not the case. The coast of Maine
is perhaps as typical of an embayed shore line as any that might be
cited, but a study of the river valleys in the neighborhood shows
clearly that the present submergence of their mouths is a fraction only
of an earlier one which has left a record of its existence in beds of
marine clay which outline the earlier and far deeper indentations (Fig.


FIG. 284.—Marine clay deposits near the mouths of the rivers of Maine
which preserve a record of earlier subsidence (after Stone).]

If now we give a closer examination to the coast, it is found that
there are marks of recent uplift in an abandoned shore line now far
above the reach of the waves. There is here, then, the record, first of
subsidence and consequent embayment, and, later, of an uplift which has
reduced the raggedness of the coast outline exposed the clay deposits,
and raised the strands of the period of deep subsidence to their
present position.

In countries which possess a more ancient civilization than our
own, the record of such oscillations in the level of the ground has
sometimes been entered upon human monuments, so that it is possible to
date more or less definitely the periods of subsidence or elevation. At
the little town of Pozzuoli, upon the shore of the Bay of Naples, is
found one of the mos instructive of these records.


FIG. 285.—View of the three standing columns of the temple of Jupiter
Serapis at Pozzuoli, showing the dark and rough band nine feet in width
affected by the rock-boring mollusks which now live in the Bay of

In the ruins of the ancient temple of Jupiter Serapis are three marble
monoliths 40 feet in height, curiously marked by a roughened surface
between the heights of 12 and 21 feet above their pedestals (Fig. 285).
Closer inspection shows that this roughened surface has been produced
by a marine, rock-boring mollusk, the _lithodomus_, which lives in
the waters of the Bay of Naples, and the shells of this animal are
still to be found within the cavities upon the surface of the columns.
Without recounting details which have been many times recited since
these interesting monuments were first geologically explored by Babbage
and Lyell, it may be stated that a record is here preserved, first of
subsidence amounting to some 40 feet, and of subsequent elevation, of
the low coast land on which stood the temple in the old Roman city of
Puteoli (Fig. 286).

At the time of deepest submergence the top of the lithodomus zone
upon the column stood at the level of the water in the Bay of Naples,
the smoother lower zone being buried at the time in the sand at the
bottom, and thus made inaccessible for the lithodomi. It is to be added
that studies made in the environs of Pozzuoli have fully confirmed
the changes of level revealed by the columns, through the discovery
of now elevated shore lines which are referable to the period of deep

=Simultaneous contrary movements upon a coast.=—In our study of
the changes in the level of the ground that take place during
earthquakes, it was learned that neighboring sections of the earth’s
crust may be moved at different rates or even in opposite directions,
notwithstanding the fact that the general movement of the province is
one of uplift. Thus during the Alaskan earthquake of 1899, although
portions of the coast line were elevated by as much as forty-seven
feet, neighboring sections were raised by smaller amounts, and some
small sections were sunk and so far submerged that the salt water and
the beach sand were washed about the roots of forest trees.

[Illustration: FIG. 286.—Pozzuoli in the 3rd, 9th, and 20th

A region racked by heavy earthquakes, where the present configuration
of the ground speaks strongly for a movement of somewhat similar
nature, but with average movement of elevation much greater to the
northward than in the opposite direction, is the extended coast line
of Chili. This country is characterized by a great central north and
south valley which separates the coast range from the high chain of the
Cordilleras to the eastward. To the southward the floor of this valley
descends, and has its continuance in the Gulf of Corcovado behind the
island of Chiloe and the Chonos archipelago. The known recent uplift of
the coast of Chili, particularly in the northern sections and during
the earthquakes of the eighteenth, nineteenth, and twentieth centuries,
lends great interest to this topographic peculiarity. Indications are
not lacking that, during the earthquake of Concepcion in 1835, and of
Valparaiso in 1907, the measure of uplift was greater to the north than
it was to the south.

[Illustration: FIG. 287.—Map of San Clemente Island, California,
showing the characteristic topography of recent uplift (after U. S.
Coast and Geodetic Survey).]

=The contrasted islands of San Clemente and Santa Catalina.=—Perhaps
the most striking example of simultaneous opposite movements observable
in neighboring portions of the earth’s crust is furnished by the coast
of southern California. The coast itself at San Pedro and the island
of San Clemente, some fifty miles off this point, in common with most
portions of the neighboring coast land, have been rising in interrupted
movements from the sea, and offer in rare perfection the characteristic
coast terraces (Fig. 287 and Fig. 278, p. 250). Midway between these
two rising sections of the crust, and less than twenty-five miles
distant from either, is the island of Santa Catalina, which has been
sinking beneath the waves, and apparently at a similarly rapid rate
(Fig. 288). The topography of the island shows the intricate detail of
a maturely eroded surface, while that of the neighboring San Clemente
shows only the widely spaced, deep cañons of the infantile stage of
erosion (Fig. 165 and pl. 12 A). While Santa Catalina has been sinking,
San Pedro Hill has risen 1240 feet, and San Clemente, 1500 feet. It
is characteristic of a sinking coast line that the cliff recession is
abnormally rapid, and evidence for this is furnished by the shores of
Santa Catalina, upon which the waves are cutting the cliffs back into
the beds of cañons, and so causing small falls to develop at the cañon

│   PLATE 12.                                                         │
│                                                                     │
│ [Illustration: _A._ V-shaped cañon cut in an upland recently        │
│ elevated from the sea, San Clemente Island, California (after W. S. │
│ Tangier-Smith).]                                                    │
│                                                                     │
│ [Illustration: _B._ A “hogback” at the base of the Bighorn          │
│ Mountains, Wyoming (after Darton).]                                 │


FIG. 288.—Map of Santa Catalina Island, California, showing the
characteristic surface of an area which has long been above the waves,
and the entire absence of coast terraces (after U. S. C. and G. S.).]

=The Blue Grotto of Capri.=—We may now return to the Bay of Naples for
additional evidence that oscillations of level in neighboring portions
of the same coast are not necessarily synchronous, and that near-lying
sections may even move in opposite directions at the same time, as has
already been shown for the islands off the California coast. For the
Pozzuoli shore of the bay it was learned that within the Christian
Era a complete cycle of downward, followed by later upward, movement
has been largely accomplished. Across the bay, and less than 20 miles
distant, is the Blue Grotto of Capri, a sea cave cut in limestone above
an earlier cave of the same nature which is now deep below the water
surface. It is the refracted sunlight which enters the cave through
this lower submerged opening and has been robbed on the way of all but
its blue rays which gives to the famous grotto its special charm (Fig.


FIG. 289.—Cross section of the Blue Grotto on the Island of Capri,
showing the submerged sea cave through which most of the light enters
the grotto, and the higher artificial window now widened by wave action
(after von Knebel).]

It is known that the former, and now submerged, sea cave was in use
by Roman patricians as a cool retreat from the oppressive hot wind
known as the sirocco, and that an artificial entrance or window was
cut where is now the only accessible entrance to the grotto. In the
ancient writings, no mention is made, however, of the remarkable blue
illumination for which it is now famous, and the conditions at the
time, as we may see, were not such as to make this possible. Later
subsidence of the coast has brought the ancient window to the sea
level, where it has been considerably enlarged by the waves. The
earlier grotto, abandoned as its entrance was closed, was rediscovered
in 1826 by the painter and poet, August Kopisch.

A grotto with green illumination (the Grotto Verde) is situated upon
the opposite side of the island, and a blue grotto, having its origin
in similar conditions to those of the famous Blue Grotto, is found upon
the island of Busi off the Dalmatian coast.

=Character profiles.=—In the landscape of a coast which has been
slowly uplifted the characteristic line is the profile of the cuesta,
with short perpendicular element joined to a gently sloping and longer
section and continued in the horizontal portion corresponding to the
lowland (Fig. 290). Rapidly uplifted coasts offer in contrast the
lines characteristic of wave erosion and deposition, but at higher
levels and in repeated sections. Most prominent of all is the staircase
constructed of coast terraces, with either vertical or sloping risers
and with outwardly inclining and gently graded treads. Near the steep
riser in the staircase may sometimes be seen the sugar-loaf outline of
the stack cut in softer material, or the obelisk-like pillar undercut
at its base, which is carved in firmer rock masses. With excessively
rapid uplift, the double-notched cliff or the double sea arch may
appear in the landscape. Upon a submerged coast the most significant
lines in the view are those of the rock islet and the steep-walled

[Illustration: FIG. 290.—Character profiles in coast landscapes where
there has been either elevation or depression.]



 SIR CH. LYELL. Principles of Geology, vol. 2, pp. 180-197.

 ED. SUESS. The Face of the Earth, Clarendon Press, Oxford, 1906, vol.
 2, Chapters i and xiv, pp. 1-29, 535-556.

 ROBERT SIEGER. Seenschwankungen und Strandverschiebungen in
 Scandinavien, Zeit. d. Gesell. f. Erdk., Berlin, vol. 28, 1893, pp.
 1-106, 393-688, pl. 7.

Elevated shore lines:—

 F. B. TAYLOR. The Highest Old Shore Line of Mackinac Island, Am. Jour.
 Sci., vol. 43, 1892, pp. 210-218.

 THOMAS L. WATSON. Evidences of Recent Elevation of the Southern Coast
 of Baffins Land, Jour. Geol., vol. 5, 1897, pp. 17-33.

 J. W. GOLDTHWAIT. The Abandoned Shore Lines of Eastern Wisconsin.
 Bull. 17, Wis. Geol. and Nat. Hist. Surv., 1907, pp. 134, pls. 1-37.

Evidences of depression:—

 W. B. SCOTT. Introduction to Geology, New York, 1907, pp. 33-36.

 W. J. MCGEE. The Gulf of Mexico as a Measure of Isostacy, Am. Jour.
 Sci. (3), vol. 44, 1892, pp. 177-192.

 A. LINDENKOHL. Notes on the Submarine Channel of the Hudson River,
 etc., Am. Jour. Sci. (3), vol. 41, 1891, pp. 489-499, pl. 18.

 J. W. SPENCER. The Submarine Great Cañon of the Hudson River, _ibid._
 (4), vol. 19, 1905, pp. 1-15; Submarine Valleys off the American Coast
 and in the North Atlantic, Bull. Geol. Soc. Am., vol. 14, 1903, pp.
 207-226, pls. 19-20.

 F. NANSEN. The Bathymetrical Features of the North Polar Sea, with
 a Discussion of the Continental Shelves and Previous Oscillations
 of Shore Line, Norwegian North Polar Expedition, vol. 4, 1904, pp.
 99-231, pl. 1.

 W. V. KNEBEL. Höhlenkunde, etc., Braunschweig, 1906, pp. 175-177 (the
 blue grotto of Capri).

Oscillation of movement:—

 C. LYELL. Principles of Geology, vol. 2, pp. 164-176 (Temple of
 Jupiter Serapis).

 E. RAY LANKESTER. Extinct Animals, New York, 1905, pp. 31-42.

 H. W. FAIRBANKS. Oscillations of the Coast of California during the
 Pliocene and Pleistocene, Am. Geol., vol. 20, 1897, pp. 213-245.

 G. H. STONE. Mon. 34, U. S. Geol. Surv., 1899, pp. 56-58, pl. 2.

 BAILEY WILLIS. Ames Knob, North Haven, Maine. Bull. Geol. Soc. Am.,
 vol. 14, 1903, pp. 201-206, pls. 17-18.

Simultaneous contrary movements on a coast:—

 A. C. LAWSON. The Post-Pliocene Diastrophism of the Coast of Southern
 California, Bull. Univ. Calif. Dept. Geol., vol. 1, 1893, pp. 115-160,
 pls. 8-9.

 W. S. TANGIER-SMITH. A Geological Sketch of San Clemente Island, 18th
 Ann. Rept. U. S. Geol. Surv., Pt. ii, 1898, pp. 459-496, pls. 84-96.

 R. S. TARR and L. MARTIN. Recent Changes of Level in the Yakutat Bay
 Region, Alaska, Bull. Geol. Soc. Am., vol. 17, 1906, pp. 29-64, pls.



=Conditions essential to glaciation.=—Wherever for a sufficiently
protracted period the annual snowfall of a district is in excess of
the snow that is melted, a residue must remain from each season to be
added to that of succeeding ones. Eventually so much snow will have
accumulated that under its own weight and in obedience to its peculiar
properties, a movement will begin within the mass tending to spread
it and so to reduce the slope of its upper surface (Frontispiece
plate). The conditions favorable to glaciation are, therefore, heavy
precipitation and low annual temperature. If the precipitation is
scanty, the small snowfall is soon melted; and if the temperature be
too high, the moisture is precipitated not in the form of snow but as
rain. It is important here to keep in mind that snow is a poor heat
conductor and itself protects its deeper layers from melting.

=The snow-line.=—Because of the low temperatures glaciers should
be most abundant or most extensive in high latitudes and in high
altitudes. The largest are found in polar and subpolar regions, and
they are elsewhere encountered only at considerable elevations.
The largest glaciers are the vast sheets of ice which inwrap the
continents of Greenland and Antarctica, but glaciers of large size
are to be found upon other large land masses of the Arctic, as well
as in Alaska, in the southern Andes, and in New Zealand. Much smaller
glaciers are characteristic of certain highlands within temperate and
tropical regions, but because of specially favorable conditions both of
altitude and precipitation the Himalayas, although in relatively low
latitudes, nourish glaciers of large proportions. In general, it may
be said that the nourishing grounds of glaciers are largely restricted
to those areas where snow covers the ground throughout the year. The
lower margin of such areas is designated the _snow line_, and varies
but little from the line on which the average summer temperature is at
the freezing point of water—the so-called _summer isotherm of 32°
Fahrenheit_. Within the tropics this line may rise as high as 18,000
feet above the sea, whereas in polar latitudes it descends to sea level.

=Importance of mountain barriers in initiating glaciers.=—The
precipitation within any district depends, however, not alone upon the
amount of moisture which is brought to it in the clouds, but upon the
amount which is abstracted before the clouds have passed over it. The
capacity of space to hold moisture increases with its temperature, and
hence any lowering of this temperature will reduce the capacity. If
lowered sufficiently, the point of complete saturation will be reached
and further cooling must result in precipitation. Hence, anything which
forces an air current to rise into more rarefied zones above, will
lower the pressure upon it and so bring about a cooling effect in which
no heat is abstracted. This so-called _adiabatic refrigeration_ of a
gas may be illustrated by the cool current which issues in a jet from
a warm expanded rubber tire after the cock has been opened; or even
better, by the instant solidification at extreme low temperatures of
such normal gases as carbonic acid when they are allowed to issue under
heavy pressure from a small orifice.

As applied to moisture-laden and near-surface winds, the effective
agents of adiabatic cooling are the upland areas upon the continents,
and especially the ranges of mountains. These barriers force the moving
clouds to rise, cool, and deposit their moisture. It is, therefore, the
highland barriers which face the oncoming, moisture-laden winds that
receive the heaviest precipitation. Within temperate regions, because
of the prevalence of westerly winds, those barriers which face the
western shores receive the heaviest fall. Within the tropics, on the
other hand, it is the barriers facing the eastern shores which, because
of the easterly “trades”, are most favorable to precipitation.

Thus it is in the Sierra Nevadas of California, and not in the Rockies
or the Appalachians, that the glaciers of the United States are found.
The highland of the Swiss Alps lying likewise athwart the “westerlies”
of the temperate zone acquires the moisture for nourishment of its
glaciers from the western ocean—here the Atlantic (Fig. 291). Within
the tropics the conditions are reversed, and it is in general the
ranges which lie nearer the eastern coasts that are the more favored.
If no barrier is found upon this coast, the clouds may travel over
vast stretches of country before being arrested by mountains and robbed
of their moisture. Thus in tropical Brazil the glaciers are found in
the Andes upon the Pacific coast though nourished by clouds from the

[Illustration: FIG. 291.—Map showing the distribution of existing
glaciers, and the two important wind poles of the earth.]

=Sensitiveness of glaciers to temperature changes.=—How sensitive
is the adjustment between snow precipitation and temperature may be
strikingly illustrated by the statement on excellent authority that if
the average annual temperature of the air within the Scottish Highlands
should be lowered by only three degrees Fahrenheit, small glaciers
would be the result; and a moderate temperature fall within the region
surrounding the Laurentian lakes of North America would bring on
glaciation, otherwise expressed as a depression of the snow line of the

=The cycle of glaciation.=—Though to-day buried beneath its ice
mantle, it is known that Greenland had more than once in earlier
geological ages a notably mild climate, and in some future age it may
revert to this condition. In other regions, also, we have evidence
that such a rotation of climatic changes has been successively
accomplished, the climate having steadily increased in severity towards
a culminating point, and been followed by a reverse series of changes.
Such a complete period may be called a _cycle of glaciation_. While
the climate is steadily becoming more rigorous, we have to do with
an _advancing hemicycle_ of glaciation, but after the culminating
point has been reached, the period of amelioration of climate is the
_receding hemicycle_.

=The advancing hemicycle.=—There is little reason to doubt that
whatever be the cause of the climatic changes which bring on glacial
conditions, these changes come on by insensible gradations. The
first visible evidence of the increased severity of the climate is
the longer persistence of the winter snows, at first within the more
elevated districts. In such positions drifts must eventually continue
throughout the warm season and so contribute to the snow accumulations
of the succeeding winter. This point once reached, small glaciers are
inevitable, even should the average temperature fall no further, for
the snow left over in each season must steadily increase the depth of
the deposits until the weight brings about an internal motion of the
mass from higher to lower levels.

[Illustration: FIG. 292.—An Alaskan glacier spreading out at the foot
of the range which nourishes it.]

The inherited depressions of the upland—the gentle hollows at the
heads of rivers—will first be filled, and so the valleys below become
the natural channels for the outflow of the early glaciers. With a
continued lowering of the annual temperature and consequent increased
snowfall, the early glaciers become more and more amply nourished. Snow
and ice will, therefore, cover larger areas of the upland, and the
glaciers will push their fronts farther down the valleys before they
are wasted in the warm air of the lower levels. As the valleys become
thus more completely invested by the glacier they are likewise filled
to greater and greater depths, and they may thus submerge portions of
the walls that separate adjacent valleys. Reaching at last the front
of the upland area, the glaciers may now be so well nourished at
their heads that they push out upon the flatter foreland and without
restraint from retaining walls spread broadly upon it (Fig. 292).

[Illustration: FIG. 293.—Surface of a glacier whose upper layers
spread with slight restraint from retaining walls. Surface of the
Folgefond, an ice cap of southern Norway.]

The culmination of the progressive climatic change may ere this have
been reached and milder conditions have ensued. If, however, the
severity of the climate should be still further increased, the expanded
fronts of neighboring glaciers will coalesce to form a common ice fan
or apron along the foot of the upland (Plate 18 B). This could hardly
take place without a still further deepening of the ice within the
valleys above, and, probably, a progressive submergence of the lower
crests in the valley walls. This may even continue until all parts of
the upland area have been buried. The snow and ice now take the form
of a covering cap or carapace, and the upper portions being no longer
restrained at the sides, now spread into a broad dome, as would a
viscous liquid like thick molasses when poured out upon the floor (Fig.
293). The lower zones of the mass and the thinner marginal portions
still have their motion to a greater or less extent controlled by the
irregularity of the rock floor against which they rest.

The reverse series of changes in the glacier is inaugurated by an
amelioration of the climate, and here, therefore, the advancing
hemicycle becomes merged in the receding hemicycle of glaciation.

=Continental and mountain glaciers contrasted.=—The time when the
rock surface becomes submerged beneath the glacier is, as regards
both the surface forms and the erosive work, a critical point of much
significance; for the ice cap and larger continental glacier obviously
protect the rock surface from the action of those chemical and
mechanical processes in which the atmosphere enters as chief agent, and
which are collectively known as weathering processes. Until submergence
is accomplished, larger or smaller portions of the rock surface project
either through or between the ice masses and are, therefore, exposed to
direct attack by the weather (see below, p. 370).

[Illustration: FIG. 294.—Section through a mountain glacier (in solid
black), showing how its surface is determined by the irregularities in
the rock basement (after Hess).]

Snow which falls in the mountains is not allowed to remain long where
it falls. By the first high wind it is swept off the more elevated and
exposed surfaces and collected under eddies in any existing hollows,
but especially those upon the lee slopes of the range. We are to learn
that glaciers carve the mountains by enlarging the hollows which they
find and producing great basins for the collection of their snows; but
with the initiation of glaciation the inherited hollows are in most
cases the unimportant depressions at the heads of streams. Whatever
they may be and however formed, the snow first fills those hollows
which are sheltered from the wind, and as it accumulates and becomes
distributed as ice, assumes a surface of its own that is dependent upon
the form and the position of the basin which it occupies (see Fig. 294).

[Illustration: FIG. 295.—Profile across the largest of the Icelandic
ice caps, with the vertical scale greatly exaggerated (after Thoroddsen
and Spethmann).]

When the quantity of accumulated snow is so great that all hollows of
the rock surface are filled, its own surface is no longer controlled
by retaining rock walls, and it now assumes a form largely independent
of the irregularities in the upland. Experience shows that this surface
is approximately that of a flat dome or shield, and as it covers all
the upland, save where the ice thins upon its margins, this type of
glacier is called an _ice cap_ (Fig. 295). All types of glacier in
which rock projects above the highest levels of the ice and snow are
known as _mountain glaciers_.

[Illustration: FIG. 296.—Ideal section across a continental glacier,
with the vertical scale and the projecting rock masses of the marginal
zone greatly magnified.]

The flat domes of ice which mantle the continents of Greenland and
Antarctica, though resembling in form the smaller ice cap, are yet
because of their vast size so distinct from them, particularly in the
manner of their nourishment, that they belong in a separate class
described as _inland ice_ or _continental glaciers_. Though they have
some affinities with ice caps, they are most sharply differentiated
from all types of mountain glaciers. Of them it is true that the
lithosphere projects through them only in the neighborhood of their
margins (Fig. 296), whereas in the case of mountain glaciers rock may
project at any level but _always above the highest snow surface_. Ice
caps may be regarded as intermediate between the two main classes of
mountain and continental glaciers (Fig. 297). Because of the large rôle
which continental glaciers have played in geological history, it is
thought best to consider them first, leaving for later discussion the
no less interesting but less important mountain glaciers.

[Illustration: FIG. 297.—View of the Eyriks-Jökull, an ice-cap of
Iceland (after Grossman).]

=The nourishment of glaciers.=—The life of a glacier is dependent upon
the continued deposition of snow in aggregate amount in excess of that
which is lost by melting or by other depleting processes. Whenever, on
the other hand, the waste exceeds the precipitation, the glacier is in
a receding condition and must eventually disappear, if such conditions
are sufficiently long continued. The source of the snow is the water
of the ocean evaporated into the atmosphere and transported over the
land in the form of clouds. We are to learn that the changes which
this moisture undergoes before its delivery to the glacier are notably
different for the classes of continental and mountain glacier.

=The upper and lower cloud zones of the atmosphere.=—Before we can
comprehend the nature of the processes by which glaciers are nourished,
it will be necessary to review the results of recent studies made upon
the earth’s atmospheric envelope. It must be kept in mind that the
sun’s rays are chiefly effective in warming the atmosphere through
being first absorbed by some solid body such as rock or water and
their heat then communicated by contact to the immediately adjacent
air layers. The layers thus warmed being now lighter than before,
they rise and are replaced by colder air, which in its turn is warmed
and likewise set in upward motion. Such currents developed in the air
by contact with warmer solid bodies constitute the process known as

[Illustration: FIG. 298.—The zones of the lower atmosphere as revealed
by recent kite and balloon explorations.]

To a relatively small degree the atmosphere is heated by the direct
absorption of the sun’s rays which pass through it. Since air has
weight, it compresses the lower layers near the earth, and hence as we
ascend from the earth’s surface the air becomes continually lighter.
Convection currents must, therefore, adjust themselves by the air
expanding as it rises. But expansion of a gas always results in its
cooling, as every one must have observed who has placed his finger in
the air current which escapes from the open valve of a warm rubber
tire. Dry air is cooled a degree Fahrenheit for every six hundred feet
of ascent in the atmosphere. At a height of about seven miles above
the earth’s surface all rising air currents have cooled to about 68°
below the zero of the Fahrenheit scale, and exploration with balloons
has shown that the currents rise no farther. At this level they move
horizontally, just as rising vapor spreads out in a room beneath the
ceiling. Above this level, as far as exploration has gone, or to a
height of more than twelve miles, the temperature remains nearly
constant, and this upper zone is, therefore, called the _isothermal_
or the _advective zone_—the uniform temperature zone of the lower
atmosphere. Beneath the convective ceiling the process of convection is
characteristic, and this zone is therefore described as the _convective
zone_ (Fig. 298).

A large part of the moisture which rises from the ocean’s surface is
condensed to vapor before it has ascended three miles, and in this form
it makes its transit over land as fleecy or stratiform clouds—the
so-called cumulus and stratus clouds and their many intermediate
varieties (see Frontispiece). This lower layer within the convective
zone is, therefore, a moist one overlaid by a relatively drier middle
layer of the convective zone. That moisture which rises above the lower
cloud layer is congealed by adiabatic cooling to fine ice needles
visible as the so-called cirrus clouds which float as feathery fronds
beneath the convective ceiling (see frontispiece at right upper corner
of picture). Thus we have within the convective zone an upper layer
more or less charged with water in the form of ice needles. It is the
clouds of the lower zone whose moisture in the form of vapor supplies
the nourishment of mountain glaciers, and the high cirrus clouds whose
congealed moisture, after interesting transformations, is responsible
for the continued existence of continental glaciers.

As we are to see, there are other noteworthy differences between
continental and mountain glaciers, in the manner of their sculpture
of the lithosphere, so that long after they have disappeared the
characters of each are easily identified in their handiwork. How the
lower clouds are forced upward and so compelled to give up their
moisture to feed the mountain glaciers, and how the upper clouds are
pulled downward to nourish the glaciers of continents, can be best
understood after the characteristics of each glacier class have been




FIG. 299.—Map of Greenland showing the area of inland-ice and the
routes of different explorers.]

=The inland ice of Greenland.=—In Greenland and in Antarctica the land
is almost or quite buried under a cover of snow and ice—the so-called
“inland ice”—which always assumes the surface of a very flat dome or
shield. In Greenland there is found a marginal ribbon of land generally
from five to twenty-five miles in width (Fig. 299), but in Antarctica
all the land, with the exception of a few mountain peaks, is inwrapped
in a mantle of ice which is also extended upon the sea in a broad shelf
of snow and ice. Neither of these vast glaciers has been explored
except in its marginal portion, yet such is the symmetry of the
profiles along the routes traversed, and such the flatness and monotony
of the snow surface within the margins, that there is little reason to
doubt that the profile made along Nansen’s route in southern Greenland
would, save only for magnitude, fairly represent a section across the
middle of the continent (Fig. 300).

=The mountain rampart and its portals.=—As soon as we examine the
coastal belt we observe that the “Great Ice” of Greenland is held
in by a wall of mountains and so prevented from spreading out to its
natural surface in the marginal portions. Through portals of the
inclosing mountain ranges—the _outlets_—it sends out _tongues_ of ice
which in many respects resemble certain types of mountain glaciers.


FIG. 300.—Profile in natural proportions across the southern end of
the continental glacier of Greenland, constructed upon an arc of the
earth’s surface and based upon Nansen’s profile corrected by Hess. The
marginal portions of the profile are represented below upon a magnified
scale in order to bring out the characters of the marginal slopes.]

Such measurements as have been made upon the inland ice of Greenland
at points back from, but yet comparatively near to, the outlets, show
that it has here a surface rate of motion amounting to less than an
inch per day, and it is highly probable that at moderate distances from
the margin this amount diminishes to zero. Upon the outlets, on the
contrary, surface rates as high as 59 feet per day have been measured,
and even 100 feet per day has been reported. We are thus justified in
saying that glacier flow within the outlets is from 700 to 1000 times
as great as it is upon the near-by inland ice, and that the glacier
is in a measure drained through the portals of the inclosing ranges.
Back from these outlet streams of ice, or tongues, the surface of the
inland ice is depressed to form a dimple or “basin of exudation” as is
the surface of a reservoir above the raceway when the water is being
rapidly drawn away (Fig. 301).

│   PLATE 13.                                                        │
│                                                                    │
│ [Illustration: _A._ Precipitous front of the Bryant glacier outlet │
│ of the Greenland inland-ice (after Chamberlin).]                   │
│                                                                    │
│ [Illustration: _B._ Lateral stream beside the Benedict glacier     │
│ outlet, Greenland (after R. E. Peary).]                            │


FIG. 301.—Map of a glacier tongue, with dimple showing above and due
to indraught of the ice. Umanakfjord, Greenland (after von Drygalski).]

Fissures in the ice, the so-called crevasses, are the recognized
marks of ice movement, and these are always concentrated at the steep
slopes of the ice surface in the neighborhood of its margins. Upon the
Greenland ice, crevasses are restricted in their distribution to a zone
which extends from seven to twenty-five miles within the ice border.

=The marginal rock islands.=—From its margin the ice surface rises
so steeply as to be climbed only with difficulty, but this gradient
steadily diminishes until at a distance of between seventy-five and
a hundred miles its slope is less than two degrees. Where crossed by
Nansen near latitude 64° N. the broad central area of ice was so nearly
level as to appear to be a plain.

As we pass across the irregular ice margin in the direction of the
interior, larger and larger proportions of the land’s surface are
submerged, until only projecting peaks rise above the ice as islands
which are described as _nunataks_ (Fig. 302).

Though not a universal observation, it has been often noted that the
absorption of the sun’s rays by rock masses projecting through the snow
results in a radiation of the heat and a lowering by melting of the
surrounding snow and ice. For this reason nunataks are often surrounded
by a deep trench due to a melting of the snow. Such a depression in
the ice surface about the margin of a nunatak, from its resemblance to
a trench about an ancient castle, has been designated a _moat_ (Fig.
303). For the same reason, the outlet tongues of ice which descend in
deep fjords between walls of rock are melted away from the walls and a
lateral stream of water is sometimes found to flow between ice and rock
(pl. 13 B).


FIG. 302.—Edge of the Greenland inland ice, showing the nunataks
diminishing in size toward the interior. The lines upon the ice are
medial moraines starting from nunataks (after Libbey).]

=Rock fragments which travel with the ice.=—Rock surfaces which are
exposed to the atmosphere are in high latitudes broken down through
the freezing of water within their crevices. The fragments resulting
from this rending process fall upon the glacier surface and are carried
forward as passengers in the direction of the ice margin. They are
either visible as long and narrow ridges or trains following the
directions of the steepest slope (Fig. 302), or they become buried
under fresh falls of snow and only again become visible where summer
melting has lowered the glacier surface in the vicinity of its margin.
These longitudinal trains of rock fragments upon the glacier surface
always have their starting point at the lower margin of one of the
nunataks, and are known as _medial moraines_ (Fig. 301, p. 273, and
Fig. 302). Inside the zone of nunataks the glacier surface is, however,
clear of rock débris except where dust has been blown on by the wind,
and this extends for a few miles only. The material of the medial
moraines is a collection of angular blocks whose surfaces are the
result of frost rending, for in their travel above the ice they are
subjected to no abrading processes.

[Illustration: FIG. 303.—Moat surrounding a nunatak in Victoria Land
(after Scott).]

A contrasted type of surface moraines upon the Greenland glacier,
instead of being parallel to the direction of ice movement, is directed
transversely or parallel to the margins. The materials of these
moraines are more rounded fragments of rock which have come up from
the bottom layers, and we shall again refer to the origin of such
moraines after the subglacial conditions have been considered.

=The grinding mill beneath the ice.=—If, now, we examine the front of
a glacier tongue which goes out from the inland ice, we find that while
the upper portion is white and mainly free from rock débris (plate
13 A), the lower zone is of a dark color and crowded with layers of
pebbles and bowlders which have been planed, polished, and scratched in
a quite remarkable manner. The ice front is itself subject to forward
and retrograde migrations of short period, but it is easily seen that
in the main its larger movement has been a retrograde one. The ground
from which it has lately withdrawn is generally a hard rock floor
unweathered, but smooth, polished, and scratched in the same manner
as the bowlders which are imbedded within the ice. It is perfectly
apparent that the latter have been derived from some portion of the
rock basement upon which the glacier still rests, and that floor and
bowlders have alike been ground smooth by mutual contact under pressure.

This erosion beneath the ice is accomplished by two processes; namely,
_plucking_ and _abrasion_. Wherever the rock over which the glacier
moves has stood up in projecting masses and is riven by fissure planes
of any kind, the ice has found it easy to remove it in larger or
smaller fragments by a quarrying process described as plucking. The
rock may be said to be torn away in blocks which are largely bounded by
the preëxisting fissure planes. Over relatively even surfaces plucking
has little importance, but where there are noteworthy inequalities
of surface upon the glacier bed, those sides which are away from the
oncoming ice (_lee_ side) are degraded by plucking in such a manner as
sometimes to leave steep and ragged fracture surfaces. The tools of
the ice thus acquired in the process of plucking are quickly frozen
into the lowest ice layers, and being now dragged along the floor
they abrade in the same manner as does a rasp or file. These tools of
the ice are themselves worn away in the process and are thus given
their characteristic shapes. Just as the lapidary grinds the surface
of a jewel into facets by imbedding the gem in a matrix, first in one
and then in another position, each time wearing down the projecting
irregularities through contact with the abrading surface; so in like
manner the rock fragment is held fast at the bottom of the glacier
until “soled” or “shod”, first upon one side and then upon another.
Accidental contact with some obstruction upon the floor may suffice
to turn the fragment and so expose a new surface to wear upon the
abrading floor. Minor obstructions coming in contact with one side of
the fragment only, may turn it in its own plane without overturning.
Evidence of such interruptions can be later read in the different
directions of striæ upon the same facet (plate 17 A).


 FIG. 304.—A glacier pavement of Permo-Carboniferous age in South
 Africa. The striæ running in the direction of the observer are
 prominent and a noteworthy gouging of the surface is to be noted to
 the right in the middle distance (after Davis).]

The floor beneath the glacier is reduced by the abrading process to a
more or less smooth and generally flattened or rounded surface—the
so-called _glacier pavement_ (Fig. 304). To accomplish this all former
mantle rock due to weathering processes must first be cleared away, and
the firm unaltered rock beneath is wherever susceptible of it given a
smooth polish although locally scored and scratched by the grinding
bowlders. The earlier projections of the surface of the floor, if not
entirely planed away, are at least transformed into rounded shoulders
of rock, which from their resemblance to closely crowded backs in a
flock of sheep have been called “sheep backs” or “_roches moutonnées_.”
Thus the effect of the combined action of the processes of plucking and
abrasion is to reduce the accent of the relief and to mold the contours
of the rock in smoothly flowing curves, generally of large radius.

=The lifting of the grinding tools and their incorporation within the
ice.=—Wherever the ice is locally held in check by the projecting
nunataks, relief is found between such obstructions, and there the flow
of the ice has a correspondingly increased velocity (Fig. 305 _b_).
If the obstructions are not of large dimensions, the ice which flows
around the outer edges is soon joined to that which passes between the
obstructions and so normal conditions of flow are restored below the
nunataks. The locally rapid flow of the ice is, therefore, restricted
to a relatively short distance, the passageway between the nunataks,
and the conditions are thus to be likened to the fall of water at a
raceway due to the sudden descent of its surface from the level of
the reservoir to the level of the stream in the outlet. As is well
known, there is under these conditions a prodigious scour upon the
bottom which tends to dig a pit just above and below the dam—a _scape
colk_—and carry the materials up to the surface below the pit. Such
a tendency was well illustrated by the behavior of the water at the
opening of the Neu Haufen dam below the city of Vienna (Fig. 305 _a_).
In the case of ice, material from the bottom may by the upward current
be brought up to the surface of the glacier at the lower edge of the
colk and thus produce a type of local surface moraine of horseshoe
form with its direction generally transverse to the direction of ice
movement (Fig. 305 _b_).


FIG. 305.—_a_, Map showing pit excavated by the current below the
opening in a dam. _b_, Nunataks and surface moraines on the Greenland
ice. Dalager’s Nunataks (after Suess).]

Any obstruction upon the pavement of the glacier apparently exerts a
larger or smaller tendency to elevate the bowlders and pebbles and
incorporate them within the ice. Rock débris thus incorporated is
described as _englacial_ drift. In the case of Greenland glaciers this
material seems at the ice front to be largely restricted to the lower
100 feet (plate 13 A).

Near the front of the inland ice the increased slope of the upper
surface greatly increases the flow of the upper ice layers in
comparison with those nearer the bottom, so that the upper layers
override the lower as they would an obstruction. The englacial drift
is either for this reason or because of rock obstructions brought to
the surface, where it yields parallel ridges corresponding in direction
to the glacier margin. Such transverse surface moraines are thus in
many respects analogous to those which appear about the lower margins
of scape colks. In contrast to the longitudinal or medial surface
moraines the materials of the transverse moraines are more faceted and
rounded—they have been abraded upon the glacier pavement.

=Melting upon the glacier margins in Greenland.=—During the short but
warm summer season, the margins of the Greenland ice are subject to
considerable losses through surface melting. When the uppermost ice
layer has attained a temperature of 32° Fahrenheit, melting begins
and moves rapidly inward from the glacier margin. In late spring the
surface of the outer marginal zone is saturated with water, and this
zone of slush advances inward with the season, but apparently never
transgresses the inner border of what we have generally referred to as
the marginal zone of the ice characterized by relatively steep slopes,
crevasses, and nunataks. Upon the ice within this zone are found
streams large enough to be designated as rivers and these are connected
with pools, lakes, and morasses. The dirt and rock fragments imbedded
in the ice are melted out in the lowering of the surface, so that late
in the season the ice presents a most dirty aspect. At the front of the
great mountain glaciers of Alaska, a more vigorous operation of the
same process has yielded a surface soil in which grow such rank forests
as entirely to mask the presence of the ice beneath.

In addition to the visible streams upon the surface of the Greenland
ice, there are others which flow beneath and can be heard by putting
the ear to the surface. All surface streams eventually encounter the
marginal crevasses and plunge down in foaming cascades, producing the
well known “glacier wells” or “glacier mills.” The progress of the
water is now throughout in tunnels within the ice until it again makes
its appearance at the glacier margin.


FIG. 306.—Marginal moraine now forming at the edge of Greenland inland
ice, showing a smooth rock pavement outside it. A small lake with a
partial covering of lake ice occupies a hollow of this pavement (after
von Drygalski).]

=The marginal moraines.=—Study of both the Greenland and Antarctic
glaciers has shown that if we disregard the smaller and short-period
migrations of the ice front, the general later movement has been a
retrograde one—we live in a receding hemicycle of glaciation. The
earlier Greenland glacier has now receded so as to expose large areas
of the former glacier pavement. In places this pavement is largely
bare, indicating a relatively rapid retirement of the ice front, but
at all points at which the ice margin was halted there is now found a
ridge of unassorted rock materials which were dropped by the ice as it
melted (Fig. 306). Such ridges, composed of the unassorted materials
described as _till_, come to have a festooned arrangement largely
concentric to the ice margin, and are the _marginal_ or _terminal
moraines_ (see Fig. 336, p. 312). Marginal moraines, if of large
dimensions, usually have a hummocky surface, and are apt to be composed
of rock fragments of a wide range of size from rock flour (clay) to
large bowlders (plate 17 A), which may represent many types since they
have been plucked by the glacier or gathered in at its surface from
many widely separated localities.


FIG. 307.—Small lake impounded between the ice front and a moraine
which it has recently built. Greenland (after von Drygalski).]

As the glacier front retires from the moraine which it has built up,
the water which emerges from beneath the ice is impounded behind the
new dam so as to form a lake of crescentic outline (Fig. 307). Such
lakes are particularly short-lived, for the reason that the water finds
an outlet over the lowest point in the crest of the moraine and easily
cuts a gorge through the loose materials, thus draining the lake (Fig.
308). Thereafter, the escaping water flows in a braided stream across
the late lake bottom and thence at the bottom of the gorge through the


FIG. 308.—View of a drained lake bottom between the moraine-covered
ice front in the foreground and an abandoned marginal moraine in the
middle distance. The water flows from the ice front in a braided stream
and passes out through the moraine in a narrow gorge. Variegated
glacier, Alaska (after Lawrence Martin).]

=The outwash plain or apron.=—The water which descends from the
glacier surface in the glacier wells or mills, eventually arrives at
the bottom, where it follows a sinuous course within a tunnel melted
out in the ice. Much of this water may issue at the ice front beneath
the coarse rock materials which are found there, and so be discovered
with the ear rather than by the eye. The water within the tunnels not
flowing with a free surface but being confined as though it were in
a pipe, may, however, reach the glacier margin under a hydrostatic
pressure sufficient to carry it up rising grades. Inasmuch as it is
heavily charged with rock débris and is suddenly checked upon arriving
at the front it deposits its burden about the ice margin so as to build
up plains of assorted sands and gravels, and over this surface it flows
in ever shifting serpentine channels of braided type (Fig. 308). Such
plains of glacier outwash are described as _outwash plains_ or _outwash

Rising as it does under hydrostatic pressure the water issuing at the
glacier front may find its way upward in some of the crevasses and so
emerge at a level considerably above the glacial floor. It may thus
come about that the outwash plain is built up about the nose of the
glacier so as partially to bury it from sight. When now the ice front
begins a rapid retirement, a depression or _fosse_ (Fig. 309 and Fig.
339, p. 314) is left behind the outwash plain and in front of the
moraine which is built up at the next halting place.

[Illustration: FIG. 309.—Diagrams to show the manner of formation
and the structure of an outwash plain, and the position of the fosse
between this and the moraine.]

=The continental glacier of Antarctica.=—In Victoria Land, upon the
continent of Antarctica, so far as exploration has yet gone, the
continental glacier is held back by a rampart of mountains, as has been
shown to be true of the inland ice of Greenland. The same flat dome or
shield has likewise been found to characterize its upper surface (Fig.

The most noteworthy differences between the inland ice masses of
Greenland and Antarctica are to be ascribed to the greater severity of
the Antarctic climate and to the more ample nourishment of the southern
glacier measured by the land area which it has submerged. There is here
no marginal land ribbon as in Greenland, but the glacier covers all
the land and is, moreover, extended upon the sea as a broad floating
terrace—the shelf ice (Fig. 311). This barrier at its margin puts a
bar to all further navigation, rising as it does in some cases 280 feet
above the sea and descending to even greater depths below (plate 15 B).


FIG. 310.—Map showing the inland ice of Victoria Land bordered by the
shelf ice of the Great Ross Barrier. The arrows show the direction of
the prevailing winds (based on maps by Scott and Shackleton).]

In that portion of Antarctica which was explored by the German
expedition, the inland ice is not as in Victoria Land restrained
within walls of rock, but is spread out upon the continent so as
to assume its natural ice slopes, which are therefore much flatter
than those examined in Greenland and Victoria Land. Here in Kaiser
Wilhelm Land the ice rises at its sea margin in a cliff which is
from 130 to 165 feet in height, then upon a fairly steeply curving
slope to an elevation of perhaps a thousand feet. Here the grades
have become relatively level, and on ever flatter slopes the surface
appears to continue into the distant interior (plate 14). Near the
ice margin numerous fissures betray a motion within the mass which
exact measurements indicate to be but one foot per day, and at a
distance of a mile and a quarter from the margin even this slight
value has diminished by fully one eighth. It can hardly be doubted
that at moderate distances only within the ice margin, the glacier is
practically without motion.

│   PLATE 14.                                                    │
│                                                                │
│ [Illustration: View of the margin of the Antarctic continental │
│ glacier in Kaiser Wilhelm Land (after E. v. Drygalski).]       │

Rain or general melting conditions being unknown in Antarctica, a
striking contrast is offered to the marginal zone of the Greenland
continent. This is to a large extent explained by the existence upon
the northern land mass of a coastland ribbon which becomes quickly
heated in the sun’s rays, and both by warming the air and by radiating
heat to the ice it causes melting and produces local air temperatures
which in summer may even be described as hot. About Independence Bay in
latitude 82° N. and near the northernmost extremity of Greenland, Peary
descended from the inland ice into a little valley within which musk
oxen were lazily grazing and where bees buzzed from blossom to blossom
over a gorgeous carpet of flowers.

[Illustration: FIG. 311.—Sections across the inland ice of Victoria
Land, Antarctica, with the shelf ice in front (after Shackleton).]

=Nourishment of continental glaciers.=—Explorations upon and about the
glaciers of Greenland and Antarctica have shown that the circulation
of air above these vast ice shields conforms to a quite simple and
symmetrical model subject to spasmodic pulsations of a very pronounced
type. Each great ice mass with its atmospheric cover constitutes a sort
of refrigerating air engine and plays an important part in the wind
system of the globe. (See Fig. 291, p. 263). Both the domed surface and
the low temperature of the glacier are essential to the continuation of
this pulsating movement within the atmosphere (Fig. 312). The air layer
in contact with the ice is during a period of calm cooled, contracted,
and rendered heavier, so that it begins to slide downward and outward
upon the domed surface in all directions. The extreme flatness of
the greater portion of the glacier surface—a fraction only of one
degree—makes the engine extremely slow in starting, but like all
bodies which slide upon inclined planes, the velocity of its movement
is rapidly accelerated, until a blizzard is developed whose vigor is
unsurpassed by any elsewhere experienced.

[Illustration: FIG. 312.—Diagram to show the nature of the fixed
glacial anticyclone above continental glaciers and the process by which
their surface is shaped.]

The effect of such centrifugal air currents above the glacier is to
suck down the air of the upper currents in order to supply the void
which soon tends to develop over the central portion of the glacier
dome. This downward vortex, fed as it is by inward-blowing, high-level
currents, and drained by outwardly directed surface currents, is what
is known as an _anticyclone_, here fixed in position by the central
embossment of the dome.

The air which descends in the central column is warmed by compression,
or adiabatically, just as air is warmed which is forced into a rubber
tire by the use of a pump. The moisture congealed in the cirrus clouds
floating in the uppermost layer of the convective zone, is carried
down in this vortex and first melted and in turn evaporated, due to
the adiabatic effect. This fusion and evaporation of the ice by its
transformation of latent, to sensible, heat, in a measure counteracts,
and so retards, the adiabatic elevation of temperature within the
column. Eventually the warm air now charged with water vapor reaches
the ice surface, is at once chilled, and its burden of moisture
precipitated in the form of fine snow needles, the so-called “frost
snow”, which in accompaniment to the sudden elevation of temperature is
precipitated at the termination of a blizzard.

The warming of the air has, however, had the effect of damping as it
were, the engine stroke, and, as the process is continued, to start a
reverse or upward current within the chimney of the anticyclone. The
blizzard is thus suddenly ended in a precipitation of the snow, which
by changing the latent heat of condensation to sensible heat tends to
increase this counter current.

[Illustration: FIG. 313.—Snow deltas about the margins of the Fan
glacier outlet of Greenland (after Chamberlin).]

=The glacier broom.=—During the calm which succeeds to the blizzard,
heat is once more abstracted from the surface air layer, and a new
outwardly directed engine stroke is begun. The tempest which later
develops acts as a gigantic centrifugal broom which sweeps out to
the margins of the glacier all portions of the latest snowfall which
have not become firmly attached to the ice surface. The sweepings
piled up about the margin of continental glaciers have been described
as fringing glaciers, or the glacial fringe. The northern coast of
Greenland and Grant Land are bordered by a fringe of this nature (plate
14 A, and Fig. 315, p. 288). It is by the operation of the glacier
broom that the inland ice is given its characteristic shield-like shape
(Fig. 312). The granular nature of the snow carried by the wind is well
brought out by the little snow deltas about the margins of Greenland
ice tongues (Fig. 313). Obviously because of the presence of the
vigorous anticyclone, no snows such as nourish mountain glaciers can be
precipitated upon continental glaciers except within a narrow marginal
zone, and, as shown by Nansen rock dust from the coastland ribbon and
from the nunataks of Greenland, is carried by a few miles inside the
western margin, and not at all within the eastern.

[Illustration: FIG. 314.—Sea ice of the Arctic region in lat. 80° 5´
N. and long. 2° 52´ E. (after Duc d’Orleans).]

=Field and pack ice.=—Within polar regions the surface of the sea
freezes during the long winter season, the product being known as
_sea-ice_ or _field-ice_ (Fig. 314). This ice cover may reach a
thickness by direct freezing of eight or more feet, and by breaking up
and being crowded above and below neighboring fragments may increase to
a considerably greater thickness. Ice thus crowded together and more or
less crushed is described as _pack ice_ or _the pack_.

The pack does not remain stationary but is continually drifting with
the wind and tide, first in one direction and then in another, but
with a general drift in the direction of the prevailing winds. Because
of the vast dimensions of the pack, the winds over widely separated
parts may be contrary in direction, and hence when currents blow
toward each other or when the ice is forced against a land area, it
is locally crushed under mighty pressures and forced up into lines of
_hummocks_—the so-called _pressure ridges_. At other times, when the
winds of widely separated areas blow away from each other, the pack is
parted, with the formation of lanes or _leads_ of open water.

If seen in bird’s-eye view the lines of hummocks would according to
Nansen be arranged like the meshes of a net having roughly squared
angles and reaching to heights of 15 to 25, rarely 30, feet above the
general surface of the pack. The ice within each mesh of the network
is a _floe_, which at the times of pressure is ground against its
neighbors and variously shifted in position. At the margin of the pack
these floes become separated and float toward lower latitudes until
they are melted.

=The drift of the pack.=—The discovery of the drift in the Arctic
pack is a romantic chapter in the history of polar exploration, and
has furnished an example of faith in scientific reasoning and judgment
which may well be compared with that of Columbus. The great figure in
this later discovery is the Norwegian explorer Fridtjof Nansen, and to
the final achievement the ill-fated _Jeannette_ expedition contributed
an important part.

The _Jeannette_ carrying the American exploring expedition was in 1879
caught in the pack to the northward of Wrangel Island (Fig. 315), and
two years later was crushed by the ice and sunk to the northward of
the New Siberian Islands. In 1884 various articles, including a list
of stores in the handwriting of the commander of the _Jeannette_, were
picked up at Julianehaab near the southern extremity of Greenland
but upon the western side of Cape Farewell. Nansen, having carefully
verified the facts, concluded that the recovered articles could have
found their way to Julianehaab only by drifting in the pack across the
polar sea, and that at the longest only five years had been consumed
in the transit. After being separated from the pack the articles must
have floated in the current which makes southward along the east coast
of Greenland and after doubling Cape Farewell flows northward upon the
west coast. It was clear that if they had come through Smith Sound they
would inevitably have been found upon the other shore of Baffin Bay. In
confirmation of this view there was found at Godthaab, a short distance
to the northward of Julianehaab (Fig. 315), an ornamented Alaskan
“throwing stick” which probably came by the same route. Moreover,
large quantities of driftwood reach the shores of Greenland which have
clearly come from the Siberian coast, since the Siberian larch has
furnished the larger quantity.

[Illustration: FIG. 315.—Map of the north polar regions, showing the
area of drift ice and the tracks of the _Jeannette_ and the _Fram_
(compiled from various maps).]

Pinning his faith to these indubitable facts, Nansen built the _Fram_
in such a manner as to resist and elude the enormous pressures of
the ice pack, stocked her with provisions sufficient for five years,
and by allowing the vessel to be frozen into the pack north of the
New Siberian Islands, he consigned himself and his companions to the
mercy of the elements. The world knows the result as one of the most
remarkable achievements in the long history of polar exploration. The
track of the _Fram_, charted in Fig. 315, considered in connection with
that of the _Jeannette_, shows that the Arctic pack drifts from Bering
Sea westward until near the northeastern coast of Greenland.

Special casks were for experimental purposes fastened in the ice to
the north of Behring Strait by Melville and Bryant, and two of these
were afterwards recovered, the one near the North Cape in northern
Norway, and the other in northeastern Iceland (see map, Fig. 315).
Peary’s trips northward in 1906 and 1909 from the vicinity of Smith
Sound have indicated that between the Pole and the shores of Greenland
and Grant Land the drift is throughout to the eastward, corresponding
to the westerly wind. Upon this border the great area of Arctic drift
ice is in contact with great continental glaciers bordered by a glacier
fringe. Admiral Peary has shown that instead of consisting of frozen
sea ice, the pack is here made up of great floes from 20 to 100 feet in
thickness and that these have been derived from the glacier fringe.

Whenever the blizzards blow off the inland ice from the south, leads
are opened at the margin of the fringe and may carry strips from the
latter northward across the lead. With favorable conditions these
leads may be closed by thick sea ice so that with the occurrence of
counter winds from the north they do not entirely return to their
original position. A continuance of this process may have resulted in
the heavy floe ice to the northward of Greenland, which, acting as
an obstruction, may have forced the thinner drift ice to keep on the
European side of the Arctic pack.

[Illustration: FIG. 316.—The shelf ice of Coats Land with the
surrounding pack ice showing in the foreground (after Bruce).]

About the Antarctic continent there is a broad girdle of pack ice
which, while more indolent in its movements than the Arctic pack, has
been shown by the expeditions of the _Belgica_ and the _Pourquoi-Pas_
to possess the same kind of shifting movements. In the southern spring
this pack floats northward and is to a large extent broken up and
melted on reaching lower latitudes.

[Illustration: FIG. 317.—Tidewater cliff at the front of a glacier
tongue from which icebergs are born.]

=The Antarctic shelf ice.=—It has been already pointed out that the
inland ice of Antarctica is in part at least surrounded by a thick
snow and ice terrace floating upon the sea and rising to heights of
more than 150 feet above it (plate 15 B and Fig. 316). The visible
portions of this shelf-ice are of stratified compact snow, and the
areas which have thus far been studied are found in bays from which
dislodgment is less easily effected. The origin of the shelf ice is
believed to be a sea-ice which because not easily detached at the time
of the spring “break-up” is thickened in succeeding seasons chiefly by
the deposition of precipitated and drifted snow upon its surface, so
that it is bowed down under the weight and sunk to greater and greater
depths in the water. To some extent, also, it is fed upon its inner
margin by overflow of glacier ice from the inland ice masses.

=Icebergs and snowbergs and the manner of their birth.=—Greenland
reveals in the character of its valleys the marks of a large subsidence
of the continent—the serpentine inlets or fjords by which its coast is
so deeply indented. Into the heads of these fjords the tongues from the
inland ice descend generally to the sea level and below. The glacier
ice is thus directly attacked by the waves as well as melted in the
water, so that it terminates in the fjords in great cliffs of ice (Fig.
317). It is also believed to extend beneath the water surface as a long
toe resting upon the bottom (Fig. 319).

│   PLATE 15.                                                       │
│                                                                   │
│ [Illustration: _A._ An Antarctic ice foot with boat party landing │
│ (after R. F. Scott).]                                             │
│                                                                   │
│ [Illustration: _B._ A near view of the front of the Great Ross    │
│ Barrier, Antarctica (after R. F. Scott).]                         │

[Illustration: FIG. 318.—A Greenlandic iceberg after a long journey in
warm latitudes.]

The exposed cliff is notched and undercut by the waves in the same
manner as a rock cliff, and the upper portions override the lower so
that at frequent intervals small masses of ice from this front separate
on crevasses, and toppling over, fall into the water with picturesque
splashes. Such small bergs, whose birth may be often seen at the cliff
front of both the Greenland and Alaskan glaciers, have little in common
with those great floating islands of ice that are drifted by the winds
until, wasted to a fraction only of their former proportions, they
reach the lanes of transatlantic travel and become a serious menace to
navigation (Fig. 318).

[Illustration: FIG. 319.—Diagram showing one way in which northern
icebergs may be born from the glacier tongue (after Russell).]

Northern icebergs of large dimensions are born either by the lifting of
a separated portion of the extended glacier toe lying upon the bottom
of the fjord, or else they separate bodily from the cliff itself,
apparently where it reaches water sufficiently deep to float it. In
either case the buoyancy of the sea water plays a large rôle in its

If derived from the submerged glacier toe (Fig. 319), a loud noise is
heard before any change is visible, and an instant later the great
mass of ice rises out of the water some distance away from the cliff,
lifting as it does so a great volume of water which pours off on all
sides in thundering cascades and exposes at last a berg of the deepest
sapphire blue. The commotion produced in the fjord is prodigious, and a
vessel in close proximity is placed in jeopardy.

Even larger bergs are sometimes seen to separate from the ice cliff, in
this case an instant before or simultaneously, with a loud report, but
such bergs float away with comparatively little commotion in the water.

[Illustration: FIG. 320.—A northern iceberg surrounded by sea ice.]

The icebergs of the south polar region are usually built upon a far
grander scale than those of the Arctic regions, and are, further, both
distinctly tabular in form and bounded by rectangular outlines (Fig.
321). Whereas the large bergs of Greenlandic origin are of ice and blue
in color, the tabular bergs of Antarctica might better be described
as _snowbergs_, since they are of a blinding whiteness and their
visible portions are either compacted snow or alternating thick layers
of compact snow and thin ribbons of blue ice, the latter thicker and
more abundant toward the base. All such bergs have been derived from
the shelf ice and not from the inland ice itself. Blue icebergs which
have been derived from the inland ice have been described from the
one Antarctic land that has been explored in which that ice descends
directly to the sea.

[Illustration: FIG. 321.—Tabular Antarctic iceberg separating from the
shelf ice (after Shackleton).]

In both the northern and southern hemispheres those bergs which have
floated into lower latitudes have suffered profound transformations.
Their exposed surfaces have been melted in the sun, washed by the
rain, and battered by the waves, so that they lose their relatively
simple forms but acquire rounded surfaces in place of the early angular
ones (Fig. 318, p. 291). Sir John Murray, who had such extended
opportunities of studying the southern icebergs from the deck of the
_Challenger_, has thus described their beauties:

 “Waves dash, against the vertical faces of the floating ice island as
 against a rocky shore, so that at the sea level they are first cut
 into ledges and gullies, and then into caves and caverns of the most
 heavenly blue, from out of which there comes the resounding roar of
 the ocean, and into which the snow-white and other petrels may be seen
 to wing their way through guards of soldier-like penguins stationed
 at the entrances. As these ice islands are slowly drifted by wind and
 current to the north, they tilt, turn and sometimes capsize, and then
 submerged prongs and spits are thrown high into the air, producing
 irregular pinnacled bergs higher, possibly, than the original
 table-shaped mass.”



 I. C. RUSSELL. Glaciers of North America. Ginn, Boston, 1897, pp. 210,
 pls. 22.

 CHAMBERLIN and SALISBURY. Geology, vol. 1, pp. 232-308.

 H. HESS. Die Gletscher, Braunschweig, 1904, pp. 426 (illustrated).

 WILLIAM H. HOBBS. Characteristics of Existing Glaciers. Macmillan,
 1911, pp. 301, pls. 34.

Special districts of mountain glaciers:—

 JAMES D. FORBES. Travels Through the Alps of Savoy and other Parts
 of the Pennine Chain with Observations on the Phenomena of Glaciers.
 Edinburgh, 1845, pp. 456, pls. 9, maps 2.

 A. PENCK, E. BRÜCKNER, et L. DU PASQUIER. Le système glaciare des
 alpes, etc., Bull. Soc. Sc. Nat. Neuchâtel, vol. 22, 1894, pp. 86.

 E. RICHTER. Die Gletscher der Ostalpen. Stuttgart, 1888, pp. 306, 7

 JAMES D. FORBES. Norway and Its Glaciers, etc. Edinburgh, 1853, pp.
 349, pls. 10, map.

 I. C. RUSSELL. Existing Glaciers of the United States, 5th Ann. Rept.
 U. S. Geol. Surv., 1885, pp. 307-355, pls. 32-55; Glaciers of Mt.
 Ranier, 18th Ann. Rept. U. S. Geol. Surv., 1898, pp. 349-423, pls.

 W. H. SHERZER. Glaciers of the Canadian Rockies and Selkirks, Smith.
 Cont. to Knowl. No. 1692, Washington, 1907, pp. 135, pls. 42.

 H. F. REID. Studies of Muir Glacier, Alaska, Nat. Geogr. Mag., vol. 4,
 1892, pp. 19-84, pls. 1-16.

 I. C. RUSSELL. Malaspina Glacier, Jour. Geol., vol. 1, 1893, pp.

 G. K. GILBERT. Harriman Alaska Expedition, vol. 3, Glaciers, 1904, pp.
 231, pls. 37.

 W. M. CONWAY. Climbing and Exploration in the Karakoram Himalayas,
 Maps and Scientific Reports, 1894, map sheets I-II.

 Geogr. Jour., vol. 35, 1910, pp. 105-132, 7 pls. and map.

The cycle of glaciation:—

 WILLIAM H. HOBBS. The Cycle of Mountain Glaciation, Geogr. Jour., vol.
 36, 1910, pp. 146-163, 268-284.

Upper and lower cloud zones of the atmosphere:—

 R. ASSMANN, A. BERSON, and H. GROSS. Wissenschaftliche Luftfahrten
 ausgeführt vom deutschen Verein zur Förderung der Luftschiffahrt in
 Berlin, 1899-1900, 3 vols.

 E. GOLD and W. A. HARWOOD. The Present State of our Knowledge of
 the Upper Atmosphere as Obtained by the Use of Kites, Balloons, and
 Pilot-balloons, Rept. Brit. Assoc. Adv. Sci., 1909, pp. 1-55.

 W. H. MOORE. Descriptive Meteorology, Appleton, New York, 1910, pp.

 WILLIAM H. HOBBS. The Pleistocene Glaciation of North America Viewed
 in the Light of our Knowledge of Existing Continental Glaciers, Bull.
 Am. Geogr. Soc., vol. 42, 1911, pp. 647-650.

The continental glacier of Greenland:—

 F. NANSEN. The First Crossing of Greenland, 2 vols, Longmans, London,
 1890 (the scientific results are contained in an appendix to volume 2,
 pp. 443-497).

 R. E. PEARY. A Reconnaissance of the Greenland Inland Ice, Jour. Am.
 Geogr. Soc., vol. 19, 1887, pp. 261-289; Journeys in North Greenland,
 Geogr. Jour., vol. 11, 1898, pp. 213-240.

 T. C. CHAMBERLIN. Glacier Studies in Greenland, Jour. Geol., vol.
 2, 1894, pp. 649-668, 768-788, vol. 3, pp. 61-69, 198-218, 469-480,
 565-582, 668-681, 833-843, vol. 4, pp. 582-592, 769-810, vol. 5, pp.
 229-245; Recent glacial studies in Greenland (Presidential address),
 Bull. Geol. Soc. Am., vol. 6, 1895, pp. 199-220, pls. 3-10.

 R. S. TARR. The Margin of the Cornell Glacier, Am. Geol., vol. 20,
 1897, pp. 139-156, pls. 6-12.

 R. D. SALISBURY. The Greenland Expedition of 1895, Jour. Geol., vol.
 3, 1895, pp. 875-902.

 E. V. DRYGALSKI. Grönland Expedition der Gesellschaft für Erdkunde zu
 Berlin 1891-1893, Berlin, 1897, 2 vols., pp. 551 and 571, pls. 53,
 maps 10.

 WILLIAM H. HOBBS. Characteristics of the Inland Ice of the Arctic
 Regions, Proc. Am. Phil. Soc., vol. 49, 1910, pp. 57-129, pls. 26-30.

The Antarctic continental glacier:—

 R. F. SCOTT. The Voyage of the _Discovery_. London, 2 vols., 1905.

 E. H. SHACKLETON. The Heart of the Antarctic. London, 2 vols., 1910.

 E. VON DRYGALSKI. Zum Kontinent des eisigen Südens, Deutsche
 Südpolar-Expedition, Fahrten und Forschungen des “Gauss”, 1901-1903,
 Berlin, 1904, pp. 668, pls. 21.

 OTTO NORDENSKIÖLD and J. S. ANDERSSON. Antarctica or Two Years Amongst
 the Ice of the South Pole. London, 1905, pp. 608, illustrated.

 E. PHILIPPI. Ueber die fünf Landeis-Expeditionen, etc., Zeit. f.
 Gletscherk., vol. 2, 1907, pp. 1-21.

Nourishment of continental glaciers:—

 WILLIAM H. HOBBS. Characteristics of the Inland Ice of the Arctic
 Regions, Proc. Am. Phil. Soc., vol. 49, 1910, pp. 96-110; The Ice
 Masses on and about the Antarctic Continent, Zeit. f. Gletscherk.,
 vol. 5, 1910, pp. 107-120; Characteristics of Existing Glaciers. New
 York, 1911, pp. 143-161, 261-289. Pleistocene Glaciation of North
 America Viewed in the Light of our Knowledge of Existing Continental
 Glaciers, Bull. Am. Geogr. Soc., vol. 43, 1911, pp. 641-659.

Field and pack ice:—

 EMMA DE LONG. The Voyage of the _Jeannette_, the ship and ice journals
 of George W. de Long, etc. Berlin, 1884, 2 vols., chart in back of
 vol. 1.

 ROBERT E. PEARY. The Discovery of the North Pole (for further
 references on both sea and pack ice and Antarctic shelf ice, consult
 Hobbs’s Characteristics of Existing Glaciers, pp. 210-213, 242-244).


 WYVILLE THOMSON. Challenger Report, Narrative, vol. 1, 1865, Pt. i,
 pp. 431-432, pls. B-D.

 I. C. RUSSELL. An Expedition to Mt. St. Elias, Nat. Geogr. Mag., vol.
 3, 1891, pp. 101-102, fig. 1.

 H. F. REID. Studies of Muir Glacier, Alaska, _ibid._, vol. 4, 1892,
 pp. 47-48.

 E. VON DRYGALSKI. Grönland-Expedition, etc., vol. 1, pp. 367-404.

 M. C. ENGELL. Ueber die Entstehung der Eisberge, Zeit. f. Gletscherk.,
 vol. 5, 1910, pp. 112-132.



=Earlier cycles of glaciation.=—Our study of the rocks composing the
outermost shell of the lithosphere tells us that in at least three
widely separated periods of its history the earth has passed through
cycles of glaciation during which considerable portions of its surface
have been submerged beneath continental glaciers. The latest of these
occurred in the yesterday of geology and has often been referred to as
the “ice age”, because until quite recently it was supposed to be the
only one of which a record was preserved.

[Illustration: FIG. 322.—Map of the globe showing the areas which were
covered by the continental glaciers of the so-called “ice-age” of the
Pleistocene period. The arrows show the directions of the centrifugal
air currents in the fixed anticyclones above the glaciers.]


FIG. 323.—Glaciated granite bowlder which has weathered out of a
moraine of Permo-Carboniferous age upon which it rests. South Australia
(after Howchin).]

This latest ice age represents four complete cycles of glaciation, for
it is believed that the continental ice developed and then completely
disappeared during a period of mild climate before the next glacier
had formed in its place, and that this alternation of climates was
no less than three times repeated, making four cycles in all. At
nearly or quite the same time ice masses developed in northern North
America and in northern Europe, the embossments of the ice domes being
located in Canada and in Scandinavia respectively (Fig. 322). There
appears to have been at this time no extensive glaciation of the
southern hemisphere, though in the next earlier of the known great
periods of glaciation—the so-called Permo-Carboniferous—it was the
southern hemisphere, and not the northern, that was affected (Fig.
323 and Fig. 304, p. 276). From the still earlier glacial period our
data are naturally much more meager, but it seems probable that it
was characterized by glaciated areas within both the northern and the
southern hemispheres.

[Illustration: FIG. 324.—Map to show the glaciated and nonglaciated
regions of North America (after Salisbury and Atwood).]

=Contrast of the glaciated and nonglaciated regions.=—Since we have
now studied in brief outline the characteristics of the existing
continental glaciers, we are in a position to review the evidences of
former glaciers, the records of which exist in their carvings, their
gravings, and their deposits.


FIG. 325.—Map of the glaciated and nonglaciated areas of northern
Europe. The strongly marked morainal belts respectively south and north
of the Baltic depression represent halting places in the retreat of the
latest continental glacier (compiled from maps by Penck and Leverett).]

An observant person familiar with the aspects of Nature in both the
northern and southern portions of the central and eastern United States
must have noticed that the general courses of the Ohio and Missouri
rivers define a somewhat marked common border of areas which in most
respects are sharply contrasted (Fig. 324). Hardly less striking is the
contrast between the glaciated and the nonglaciated regions upon the
continent of Europe (Fig. 325).

It is the northern of the two areas which in each case reveals
the characteristic evidences of glaciation, while there is entire
absence of such marks to the southward of the common border. Within
the American glaciated region there is, however, an area surrounded
like an island, and within this district (Fig. 324) none of the marks
characteristic of glaciation are to be found. This area is usually
referred to as the “driftless area”, and occupies portions of the
states of Wisconsin, Illinois, Minnesota, and Iowa. Even better than
the area to the southward of the Ohio and Missouri rivers, it permits
of a comparison of the nonglaciated with the drift-covered region.


FIG. 326.—“Stand Rock” near the “Dells” of the Wisconsin river, an
unstable erosion remnant characteristic of the driftless area of North
America (after Salisbury and Atwood).]

=The “driftless area.”=—Within this district, then, we have preserved
for our study a landscape which remains largely as it was before
the several ice invasions had so profoundly transformed the general
surface of the surrounding country. Speaking broadly, we may say that
it represents an uplifted and in part dissected plain, which to the
south and east particularly reveals the character of nearly mature
river erosion (Fig. 177, p. 170). The rock surface is here everywhere
mantled by decomposed and disintegrated rock residues of local origin.
The soluble constituents of the rock, such as the carbonates, have
been removed by the process of leaching, so that the clays no longer
effervesce when treated with dilute mineral acid.

Wherever favored by joints and by an alternation of harder and softer
rock layers, picturesque unstable erosion remnants or “chimneys” may
stand out in relief (Fig. 326). Furthermore, the driftless area is
throughout perfectly drained—it is without lakes or swamps—since
all valleys are characterized throughout by forward grades. The side
valleys enter the main valleys as do the branches a tree trunk; in
other words, the drainage is described as arborescent. In so far as any
portions of a plane surface now remain in the landscape, they are found
at the highest levels (plate 16 A). The topography is thus the result
of a partial removal by erosion of an upland and may be described as
_incised topography_. Nowhere within the area are there found rock
masses foreign to the region, but all mantle rock is the weathered
product of the underlying ledges.

│  PLATE 16.                                                         │
│                                                                    │
│ [Illustration: _A._ Incised topography within the “driftless area” │
│ (U. S. Geol. Survey).]                                             │
│                                                                    │
│ [Illustration: _B._ Built-up topography within glaciated region    │
│ (U. S. Geol. Survey).]                                             │

=Characteristics of the glaciated regions.=—The topography of the
driftless area has been described as _incised_, because due to the
partial destruction of an uplifted plain; and this surface is,
moreover, perfectly drained. The characteristic topography of the
“drift” areas is by contrast _built up_; that is to say, the features
of the region instead of being _carved_ out of a plain are the result
of _molding_ by the process of deposition (plate 16 B). In so far as a
plane is recognizable, it is to be found not at the highest, but at the
lowest level—a surface represented largely by swamps and lakes—and
above this plain rise the characteristic rounded hills of various types
which have been _built up_ through deposition. The process by which
this has been accomplished is one easy to comprehend. As it invaded the
region, the glacier planed away beneath its marginal zone all weathered
mantle rock and deposited the planings within the hollows of the
surface (Fig. 327). The effect has been to flatten out the preëxisting
irregularities of the surface, and to yield at first a gently
undulating plain upon which are many undrained areas and a haphazard
system of drainage (Fig. 328). All unstable erosion remnants, such as
now are to be found within the driftless area, were the first to be
toppled over by the invading glacier, and in their place there is left
at best only rounded and polished “shoulders” of hard and unweathered
rock—the well-known _roches moutonnées_.


FIG. 327.—Diagram showing the manner in which a continental glacier
obliterates existing valleys (after Tarr).]

=The glacier gravings.=—The tools with which the glacier works are
never quite evenly edged, and instead of an in all respects perfect
polish upon the rock pavement, there are left furrowings, gougings, and
scratches. Of whatever sort, these scorings indicate the lines of ice
movement and are thus indubitable records graven upon the rock floor.
When mapped over wide areas, a most interesting picture is presented to
our view, and one which supplements in an important way the studies of
existing continental glaciers (Fig. 334, p. 308, and Fig. 336, p. 312).

[Illustration: FIG. 328.—Lake and marsh district in northern
Wisconsin, the effect of glacial deposition in former valleys (after

It has been customary to think of the glacier as everywhere eroding
its bed, although the only warrant for assuming degradation by flow of
the ice is restricted to the marginal zone, since here only is there
an appreciable surface grade likely to induce flow. Both upon the
advance and again during the retreat of a glacier, all parts of the
area overridden must be subjected to this action. Heretofore pictured
in the imagination as enlarged models of Alpine glaciers, the vast
ice mantles were conceived to have spread out over the country as the
result of a kind of viscous flow like that of molasses poured upon
a flat surface in cold weather. The maximum thickness of the latest
American glacier of the ice age has been assumed to have been perhaps
10,000 feet near the summit of its dome in central Labrador. From this
point it was assumed that the ice traveled southward up the northern
slope of the Laurentian divide in Canada, and thence to the Ohio river,
a distance of over 1300 miles. If such a mantle of ice be represented
in its natural proportions in vertical section, to cover the distance
from center to margin we may use a line six inches in length, and
only 1/100 of an inch thick. Upon a reduced scale these proportions
are given in Fig. 329. Obviously the force of gravity acting within
a viscous mass of such proportions would be incompetent to effect a
transfer of material from the center to the periphery, even though the
thickness should be doubled or trebled. Yet until the fixed glacial
anticyclone above the glacier had been proven and its efficiency as a
broom recognized, no other hypothesis than that of viscous flow had
been offered in explanation. The inherited conception of a universal
plucking and abrasion on the bed of the glacier is thus made untenable
and can be accepted for the marginal portion only.

[Illustration: FIG. 329.—Cross section in approximate natural
proportions of the latest North American continental glacier of
Pleistocene age from its center to its margin.]

Not only do the rock scorings show the lines of ice movement, but
the directions as well may often be read upon the rock. Wherever
there are pronounced irregularities of surface still existing on the
pavement, these are generally found to have gradual slopes upon the
side from which the ice came, and relatively steep falls upon the
lee or “pluck” side. If, however, we consider the irregularities of
smaller size, the unsymmetrical slopes of these protruding portions
of the floor are found to be reversed—it is the steep slope which
faces the oncoming ice and the flatter slope which is upon the lee
side. Such minor projections upon the floor usually have their origin
in some harder nodule which deflects the abrading tools and causes
them to pass, some on the one side and some upon the other. By this
process a staple-shaped groove comes to surround the nodule, leaving an
unsymmetrical elevated ridge within, which is steep upon the stoss side
and slopes gently away to leeward.

[Illustration: FIG. 330.—Limestone surface at Sibley, Michigan.]

=Younger records over older—the glacier palimpsest.=—Many important
historical facts have been recovered from the largely effaced writing
upon ancient palimpsests, or parchments upon which an earlier record
has been intentionally erased to make room for another. In the
gravings upon the glacier pavement, earlier records have been likewise
in large part effaced by later, though in favorable localities the
two may be read together. Thus, as an example, at the great limestone
quarries of Sibley, in southeastern Michigan, the glaciated rock
surface wherever stripped of its drift cover is a smoothly polished and
relatively level floor with striæ which are directed west-northwest.
Beneath this general surface there are, however, a number of elliptical
depressions which have their longer axes directed south-southwest, one
being from twenty-five to thirty feet long and some ten feet in depth
(Fig. 330). These boat-shaped depressions are clearly the remnants of
an earlier more undulating surface which the latest glacier has in
large part planed away, since the bottoms of the depressions are no
less perfectly glaciated but have their striæ directed in general near
the longer axis of the troughs. Palimpsest-like there are here also the
records of more than one graving.


FIG. 331.—Map to show the outcroppings of peculiar rock types in the
region of the Great Lakes, and some of the localities where “float
copper” has been collected (float copper localities after Salisbury).]

=The dispersion of the drift.=—Long before the “ice age” had been
conceived in the minds of Agassiz and his contemporaries, it had been
remarked that scattered over the North German plain were rounded
fragments of rock which could not possibly have been derived from their
own neighborhood but which could be matched with the great masses of
red granite in Sweden well known as the “Swedish granite.” Buckland, an
English geologist, had in 1815 accounted for such “erratic” blocks of
his own country, here of Scotch granite, by calling in the deluge of
Noah; but in the late thirties of the nineteenth century, Sir Charles
Lyell, with the results of English Arctic explorers in mind, claimed
that such traveled blocks had been transported by icebergs emanating
from the polar regions. A relic of Buckland’s earlier view we have
in the word “diluvium” still occasionally used in Germany for glacier
transported materials; while the term “drift” still remains in common
use to recall Lyell’s iceberg hypothesis, even though the original
meaning of the term has been abandoned. Drift is now a generic term
and refers to all deposits directly or indirectly referable to the
continental glaciers.

In general the place of derivation of the glacial drift may be said to
be some point more distant from and within the former ice margin at the
time when it was deposited; in other words, the dispersion of the drift
was centrifugal with reference to the glacier.


FIG. 332.—Map of the “bowlder train” from Iron Hill, R. I. (based upon
Shaler’s map, but with the directions of glacial striæ added).]

Wherever rocks of unusual and therefore easily recognizable character
can be shown to occur in place and with but limited areas, the
dispersion of such material is easy to trace. The areas of red Swedish
and Scotch granite have been used to follow out in a broad way the
dispersion of drift over northern Europe. Within the region of the
Great Lakes of North America are areas of limited size which are
occupied by well marked rock types, so that the journeyings of their
fragments with the continental glacier can be mapped with some care.
Upon the northern shore of Georgian Bay occurs the beautiful jasper
conglomerate, whose bright red pebbles in their white quartz field
attract such general notice. At Ishpeming in the northern peninsula of
Michigan is found the equally beautiful jaspilite composed of puckered
alternating layers of black hematite and red jasper. On Keweenaw
Peninsula, which protrudes into Lake Superior from its southern shore,
is found that remarkable occurrence of native copper within a series
of igneous rocks of varied types and colors. Fragments of this copper,
some weighing several hundreds of pounds each and masked in a coat
of green malachite, have under the name of “drift” or “float” copper
been collected at many localities within a broad “fan” of dispersal
extending almost to the very limits of glaciation (Fig. 331).

Some miles to the north of Providence in Rhode Island there is a hill
known as Iron Hill composed in large part of black magnetite rock,
the so-called Cumberlandite. From this hill as an apex there has
been dispersed a great quantity of the rock distributed as a well
marked “bowlder train” within which the size and the frequency of the
dispersed bowlders is in inverse ratio to the distance from the parent
ledge (Fig. 332). Similar though less perfect trains of bowlders are
found on the lee side of most projecting masses of resistant rocks
within the area of the drift.

Large bowlders when left upon a ledge of notably different appearance
easily attract attention, and have been described as “perched
bowlders.” Resting as they sometimes do upon a relatively small area,
they may be nicely balanced and thus easily given a pendular or rocking
motion. Such “rocking stones” are common enough, especially among the
New England hills (plate 17 B). Many such bowlders have made somewhat
remarkable peregrinations with many interruptions, having been carried
first in one direction by an earlier glacier to be later transported in
wholly different directions at the time of new ice invasions.

│   PLATE 17.                                                       │
│                                                                   │
│ [Illustration: _A._ Soled glacial bowlders which show differently │
│ directed striæ upon the same facet.]                              │
│                                                                   │
│ [Illustration: _B._ Perched bowlder upon a striated ledge of      │
│ different rock type, Bronx Park, New York (after Lungstedt).]     │
│                                                                   │
│ [Illustration: _C._ Characteristic knob and basin surface of a    │
│ moraine.]                                                         │

=The diamonds of the drift.=—Of considerable popular, even if not
economic, interest are the diamonds which have been sown in the drift
after long and interrupted journeyings with the ice from some unknown
home far to the northward in the wilderness of Canada. The first stone
to be discovered was taken by workmen from a well opening near the
little town of Eagle in Wisconsin in the year 1876. Its nature not
being known, it remained where it was found as a curiosity only, and it
was not until 1883 that it was taken to Milwaukee and sold to a jeweler
equally ignorant of its value, and for the merely nominal sum of one
dollar. Later recognized as a diamond of the unusual weight of sixteen
carats, it was sold to the Tiffanys and became the cause of a long
litigation which did not end until the Supreme Court of Wisconsin had
decided that the Milwaukee jeweler, and not the finder, was entitled to
the price of the stone, since he had been ignorant of its value at the
time of purchase.


FIG. 333.—Shapes and approximate natural sizes of some of the more
important diamonds from the Great Lakes region of the United States. In
order from left to right these figures represent the Eagle diamond of
sixteen carats, the Saukville diamond of six and one half carats, the
Milford diamond of six carats, the Oregon diamond of four carats, and
the Burlington diamond of a little over two carats.]

An even larger diamond, of twenty-one carats weight, was found at
Kohlsville, and smaller ones at Oregon, Saukville, Burlington, and Plum
Creek in the state of Wisconsin; at Dowagiac in Michigan; at Milford in
Ohio, and in Morgan and Brown counties in Indiana. The appearance of
some of the larger stones in their natural size and shape may be seen
in Fig. 333.

While the number of the diamonds sown in the drift is undoubtedly
large, their dispersion is such that it is little likely they can be
profitably recovered. The distribution of the localities at which
stones have thus far been found is set forth upon Fig. 334. Obviously
those that have been found are the ones of larger size, since these
only attract attention. In 1893, when the finding of the Oregon stone
drew attention to these denizens of the drift, the writer prophesied
that other stones would occasionally be discovered under essentially
the same conditions, and such discoveries are certain to continue in
the future.


FIG. 334.—Glacial map of a portion of the Great Lakes region,
showing the unglaciated area and the areas of older and newer drift.
The driftless area, the moraines of the later ice invasion, and the
distribution of diamond localities upon the latter are also shown.
With the aid of the directions of striæ some attempt has been made to
indicate the probable tracks of more important diamonds, which tracks
converge in the direction of the Labrador peninsula.]

=Tabulated comparison of the glaciated and nonglaciated regions.=—It
will now be profitable to sum up in parallel columns the contrasted
peculiarities of the glaciated and the unglaciated regions.



 The topography is _destructional_; The topography is _constructional_;
 the remnants of a plain are found  the remnants of a plain are found
 at the highest levels or upon the  at the lowest levels in lakes and
 hill tops; hills are _carved_ out  swamps; hills are _molded_ above a
 of a high plain; unstable erosion  plain in characteristic forms; no
 remnants are characteristic.       unstable erosion remnants, but only
                                    rounded shoulders of rock.


 The area is completely drained,    The area includes undrained
 and the drainage network is        areas,—lakes and swamps,—and
 _arborescent_.                     the drainage system is _haphazard_.


 The exposed rock is decomposed      No decomposed or disintegrated
 and disintegrated to a              rock is “in place”, but only
 considerable depth; it is all of    hard, fresh surface; loose rock
 local derivation and hence of few   material is all foreign and of many
 types—_homogeneous_; the fragments  sizes and types—_heterogeneous_;
 are angular; soils are leached and  rock bowlders and pebbles are
 hence do not contain carbonates.    faceted and polished as well as
                                     striated, usually in several
                                     directions upon each facet; soils
                                     are rock flour—the grist of the
                                     glacial mill.


 Rock surface is rough and           Rock surface is planed or grooved,
 irregular.                          and polished. Shows glacial striæ.

=Unassorted and assorted drift.=—The drift is of two distinct types;
namely, that deposited directly by the glacier, which is without
stratification, or unassorted; and that deposited by water flowing
either beneath or from the ice, and this like most fluid deposited
material is assorted or stratified. The unassorted material is
described as _till_, or sometimes as “bowlder clay”; the assorted
is sand or gravel, sometimes with small included bowlders, and is
described as _kame gravel_. To recall the parts which both the glacier
and the streams have played in its deposition, all water-deposited
materials in connection with glaciers are called _fluvio-glacial_.

[Illustration: FIG. 335.—Section in coarse till. Note the range in
size of the materials, the lack of stratification, and the “soled” form
of the bowlders.]

Till is, then, characterized by a noteworthy lack of homogeneity, both
as regards the size and the composition of its constituent parts. As
many as twenty different rock types of varied textures and colors may
sometimes be found in a single exposure of this material, and the
entire gamut is run from the finest rock flour upon the one hand to
bowlders whose diameter may be measured in feet (Fig. 335).

In contrast with those derived by ordinary stream action, the pebbles
and bowlders of the till are faceted or “soled”, and usually show
striations upon their faces. If a number of pebbles are examined,
some at least are sure to be found with striations in more than one
direction upon a single facet. As a criterion for the discrimination
of the material this may be an important mark to be made use of to
distinguish in special cases from rock fragments derived by brecciation
and slickensiding and distributed by the torrents of arid and semiarid

Inasmuch as the capacity of ice for handling large masses is greater
than that of water, assorted drift is in general less coarse, and, as
its name implies, it is also stratified. From ordinary stream gravels,
the kame gravels are distinguished by the form of their pebbles, which
are generally faceted and in some cases striated. In proportion,
however, as the materials are much worked over by the water, the
angles between pebble faces become rounded and the original shapes
considerably masked.

=Features into which the drift is molded.=—Though the preëxisting
valleys were first filled in by drift materials, thus reducing the
accent of the relief, a continuation of the same process resulted in
the superimposition of features of characteristic shapes upon the
imperfectly evened surface of the earlier stages. These features belong
to several different types, according as they were built up outside
of, at and upon, or within the glacier margin. The extra-marginal
deposits are described as _outwash plains_ or _aprons_, or sometimes as
_valley trains_; the marginal are either _moraines_ or _kames_; while
within the border were formed the _till plain_ or _ground moraine_,
and, locally also, the _drumlin_ and the _esker_ or _os_. These
characteristic features are with few exceptions to be found only within
the area covered by the latest of the ice invasions. For the earlier
ones, so much time has now elapsed that the effect of weathering,
wash, and stream erosion has been such that few of the features are

Marginal and extra-marginal features are extended in the direction
of the margin or, in other words, perpendicular to the local ice
movement; while the intra-marginal deposits are as noteworthy for being
perpendicular to the margin, or in correspondence with the direction
of local ice movement. Each of these features possesses characteristic
marks in its form, its size, proportions, surface molding and
orientation, as well as in its constituent materials. It should perhaps
be pointed out that the existing continental glaciers, being in high
latitudes, work upon rock materials which have been subjected to
different weathering processes from those characteristic of temperate
latitudes. Moreover, the melting of the Pleistocene glaciers having
taken place in relatively low latitudes, larger quantities of rock
débris were probably released from the ice during the time of definite
climatic changes, and hence heavier drift accumulations have for both
of these reasons resulted.

=Marginal or “kettle” moraines.=—Wherever for a protracted period the
margin of the glacier was halted, considerable deposits of drift were
built up at the ice margin. These accumulations form, however, not only
about the margin, but upon the ice surface as well; in part due to
materials collected from melting down of the surface, and in part by
the upturning of ice layers near the margin (see _ante_, p. 277).


FIG. 336.—Sketch map of portions of Michigan, Ohio, and Indiana,
showing the festooned outlines of the moraines about the former ice
lobes, and the directions of ice movement as determined by the striæ
upon the rock pavement (after Leverett).]

An important rôle is played by the thaw water which emerges at the
ice margin, especially within the reëntrants or recesses of the
outline. The materials of moraines are, therefore, till with large
local deposits of kame gravel, and these form in a series of ridges
corresponding to the temporary positions of the ice front. Their width
may range from a few rods to a few miles, their height may reach a
hundred feet or more, and they stretch across the country for distances
of hundreds or even thousands of miles, looped in arcs or scallops
which are always convex outward and which meet in sharp cusps that in
a general way point toward the embossment of the former glacier (Fig.
334, p. 308, and Fig. 336). These festoons of the moraines outline
the ice lobes of the latest ice invasion, which in North America were
centered over the depressions now occupied by the Laurentian lakes.
There was, thus, a Lake Superior lobe, a Lake Michigan lobe, etc. With
the aid of these moraine maps we may thus in imagination picture in
broad lines the frontal contours of the earlier glaciers. At specially
favorable localities where the ice front has crossed a deep valley at
the edge of the Driftless Area, we may, even in a rough way, measure
the slope of the ice face. Thus near Devils Lake in southern Wisconsin
the terminal moraine crosses the former valley of the Wisconsin River,
and in so doing has dropped a distance of about four hundred feet
within the distance of a half mile or thereabouts (Fig. 337).

The characteristic surface of the marginal moraine is responsible for
the name “kettle” moraine so generally applied to it. The “kettles” are
roughly circular, undrained basins which lie among hummocks or knobs,
so that the surface has often been referred to as “knob and basin”
topography (plate 17 C).


FIG. 337.—Map of the vicinity of Devils Lake, Wisconsin, located
within a reëntrant of the “kettle” moraine upon the margin of the
Driftless Area. The lake lies within an earlier channel of the
Wisconsin River which has been blocked at both ends, first by the
glacier and later by its moraine. The stippled area upon the heights
and next the moraine represents the clay deposits of a former lake
(based on map by Salisbury and Atwood).]

[Illustration: FIG. 338.—Moraine with outwash apron in front, the
latter in part eroded by a river. Westergötland, Sweden (after H.


FIG. 339.—Fosse between an outwash plain (in the foreground) and the
moraine, which rises to the left in the middle distance. Ann Arbor,

=Kames.=—Within reëntrants or recesses of the ice margin the drift
deposits were especially heavy, so that high hills of hummocky surface
have been built up, which are described as _kames_. Most of the higher
drift hills have this origin. They rarely have any principal extension
along a single direction, but are composed in large part of assorted
materials. In contrast with other portions of the morainal ridges they
lack the prominent basins known as kettles. Other _kames_ are high
hills of assorted materials not in direct association with moraines and
believed to have been built up beneath glacier wells or mills (p. 278).

=Outwash plains.=—Upon the outer margin of the moraine is generally
to be found a plain of glacial “outwash” composed of sand or gravel
deposited by the braided streams (Fig. 308, p. 280) flowing from the
glacier margin. Such plains, while notably flat (Fig. 338), slope
gently away from the moraine. Between the outwash plain and the moraine
there is sometimes found a pit, or _fosse_ (Fig. 309, p. 281), where a
part of the ice front was in part buried in its own outwash (Fig. 339).

[Illustration: FIG. 340.—View looking along an esker in southern Maine
(after Stone).]

=Pitted plains and interlobate moraines.=—Where glacial outwash is
concentrated within a long and narrow reëntrant, separating glacial
lobes, strips of high plain are sometimes built up which overtop the
other glacial deposits of the district. The sand and gravel which
compose such plains have a surface which is pitted by numerous deep and
more or less circular lakes, so that the term “pitted plain” has been
applied to them. The surface of such a plain steadily rises toward its
highest point in the angle between the ice lobes. Though consisting
almost entirely of assorted materials, and built up largely without
the ice margins, such gently sloping pitted platforms are described as
_interlobate moraines_. Upon a topographic map the course of such an
interlobate moraine may often be followed by the belts of small pit
lakes (see Fig. 336).


FIG. 341.—Outline map showing the eskers of Finland trending
southeasterly toward the festooned moraines at the margin of the ice.
The characteristic lakes of a glaciated region appear behind the
moraines (after J. J. Sederholm).]


FIG. 342.—Small sketch maps showing the relationships in size,
proportions, and orientation of drumlins and eskers in southern
Wisconsin. The eskers are in solid black (after Alden).]

=Eskers.=—Intra-morainal features, or those developed beneath the
glacier but relatively near its margin, include the “serpentine kame”,
_esker_, or, as it is called in Scandinavia, the _os_ (plural _osar_)
(Fig. 340). These diminutive ridges have a width seldom exceeding a
few rods, and a height a few tens of feet at most, but with slightly
sinuous undulations they may be followed for tens or even hundreds of
miles in the general direction of the local ice movement (Fig. 341).
They are composed of poorly stratified, thick-bedded sands, gravels,
and “worked over” materials, and are believed to have been formed by
subglacial rivers which flowed in tunnels beneath the ice. Inasmuch as
the deposits were piled against the ice walls, the beds were disturbed
at the sides when these walls disappeared, and the stratification,
which was somewhat arched in the beginning, has been altered by
sliding at both margins. As already stated, eskers have not a general
distribution within the glaciated area, but are often found in great
numbers at specially favored localities. Formed as they are beneath
the ice, it is believed that many have their materials redistributed
so soon as uncovered at the glacier margin, because of the vigorous
drainage there. They are thus to be found only at those favored
localities where for some reason border drainage is less active, or
where the ice ended in a body of water.

=Drumlins.=—A peculiar type of small hill likewise found behind the
marginal moraine in certain favored districts has the form of an
inverted boat or canoe, the long axis of which is parallel to the
direction of ice movement, as is that of the esker (Fig. 342). Unlike
the esker, this type of hill is composed of till, and from being
found in Ireland it is called a _drumlin_, the Irish word meaning a
little hill (Fig. 343). Drumlins are usually found in groups more or
less radial and not far behind the outermost moraine, to which their
radiating axes are perpendicular. The manner of their formation is
involved in some uncertainty, but it is clear that they have been
formed beneath the margin of the glacier, and have been given their
shape by the last glacier which occupied the district.

The mutual relationships of nearly all the molded features resulting
from continental glaciation may be read from Fig. 344.

[Illustration: FIG. 343.—View of a drumlin, showing an opening in the
till. Near Boston, Massachusetts (after Shaler and Davis).]

=The shelf ice of the ice age.=—Shelf ice, such as we have become
familiar with in Antarctica as a marginal snow-ice terrace floating
upon the sea, no doubt existed during the ice age above the Gulf of
Maine (see Fig. 324, p. 298), and perhaps also over the deep sea to the
westward of Scotland. Though the inland ice probably covered the North
Sea, and upon the American side of the Atlantic the Long Island Sound,
both these basins are so shallow that the ice must have rested upon the
bottom, for neither is of sufficient depth to entirely submerge one of
the higher European cathedrals.


FIG. 344.—Outline map of the front of the Green Bay lobe of the latest
continental glacier of the United States. Drumlins in solid black,
moraines with diagonal hachure, outwash plains and the till plain or
ground moraine in white (after Alden).]

=Character profiles.=—All surface features referable to continental
glaciers, whether carved in rock or molded from loose materials,
present gently flowing outlines which are convex upward (Fig. 345). The
only definite features carved from rock are the _roches moutonnées_,
with their flattened shoulders, while the hillocks upon moraines and
kames, and the drumlins as well, approximate to the same profile. The
esker in its cross sections is much the same, though its serpentine
extension may offer some variety of curvature when viewed from higher

[Illustration: FIG. 345.—Character profiles referable to continental



 JAMES GEIKIE. The Great Ice Age. 3d ed. London, 1894, pp. 850, maps 18.

 CHAMBERLIN and SALISBURY. Geology, vol. 3, 1906, pp. 327-516.

 FRANK LEVERETT. The Illinois Glacial Lobe, Mon. 38, U. S. Geol. Surv.,
 1899, pp. 817, pls. 34; Glacial formations and Drainage Features of
 the Erie and Ohio Basins, Mon. 41, _ibid._, 1902, pp. 802, pls. 25;
 Comparison of North American and European Glacial Deposits, Zeit. f.
 Gletscherk., vol. 4, 1910, pp. 241-315, pls. 1-5.

Former glaciations previous to Ice Age:—

 A. STRAHAN. The Glacial Phenomena of Paleozoic Age in the Varanger
 Fjord, Quart. Jour. Geol. Soc., London, vol. 53, 1897, pp. 137-146,
 pls. 8-10.

 BAILEY WILLIS and ELIOT BLACKWELDER. Research in China, Pub. 54,
 Carnegie Inst. Washington, vol. 1, 1907, pp. 267-269, pls. 37-38.

 A. P. COLEMAN. A Lower Huronian Ice Age, Am. Jour. Sci. (4), vol. 23,
 1907, pp. 187-192.

 W. M. DAVIS. Observations in South Africa, Bull. Geol. Soc. Am., vol.
 17, 1906, pp. 377-450, pls. 47-54.

 DAVID WHITE. Permo-Carboniferous Climatic Changes in South America,
 Jour. Geol., vol. 15, 1907, pp. 615-633.

Driftless and drift areas:—

 T. C. CHAMBERLIN and R. D. SALISBURY. Preliminary Paper on the
 Driftless Areas of the Upper Mississippi Valley, 6th Ann. Rept. U. S.
 Geol. Surv., 1885, pp. 199-322, pls. 23-29.

 R. D. SALISBURY. The Drift, its Characteristics and Relationships,
 Jour. Geol., vol. 2, 1894, pp. 708-724, 837-851.

 R. H. WHITBECK. Contrasts between the Glaciated and the Driftless
 Portions of Wisconsin, Bull. Geogr. Soc., Philadelphia, vol. 9, 1911,
 pp. 114-123.

Glacier gravings:—

 T. C. CHAMBERLIN. The Rock Scorings of the Great Ice Invasions, 7th
 Ann. Rept. U. S. Geol. Surv., 1888, pp. 147-248, pl. 8.

The dispersion of the drift:—

 R. D. SALISBURY. Notes on the Dispersion of Drift Copper, Trans. Wis.
 Acad. Sci., etc., vol. 6, 1886, pp. 42-50, pl.

 N. S. SHALER. The Conditions of Erosion beneath Deep Glaciers, based
 upon a Study of the Bowlder Train from Iron Hill, Cumberland, Rhode
 Island, Bull. Mus. Comp. Zoöl. Harv. Coll., vol. 16, No. 11, 1893, pp.
 185-225, pls. 1-4 and map.

 WILLIAM H. HOBBS. The Diamond Field of the Great Lakes, Jour. Geol.,
 vol. 7, 1899, pp. 375-388, pls. 2 (also Rept. Smithson. Inst., 1901,
 pp. 359-366, pls. 1-3).

Glacial features:—

 T. C. CHAMBERLIN. Preliminary Paper on the Terminal Moraine of the
 Second Glacial Epoch, 3d Ann. Rept. U. S. Geol. Surv., 1883, pp.
 291-402, pls. 26-35.

 G. H. STONE. Glacial Gravels of Maine and their Associated Deposits,
 Mon. 34, U. S. Geol. Surv., 1899, pp. 489, pls. 52.

 W. C. ALDEN. The Delaven Lobe of the Lake Michigan Glacier of the
 Wisconsin Stage of Glaciation and Associated Phenomena. Prof. Pap.
 No. 34, U. S. Geol. Surv., 1904, pp. 106, pls. 15; The Drumlins of
 Southeastern Wisconsin, Bull. 273, U. S. Geol. Surv., 1905, pp. 46,
 pls. 9.

 W. M. DAVIS. Structure and Origin of Glacial Sand Plains, Bull. Geol.
 Soc. Am., vol. 1, 1890, pp. 196-202, pl. 3; The Subglacial Origin
 of Certain Eskers, Proc. Bost. Soc. Nat. Hist., vol. 35, 1892, pp.

 F. P. GULLIVER. The Newtonville Sand Plain, Jour. Geol., vol. 1, 1893,
 pp. 803-812.




FIG. 346.—The Illinois River where it passes through the outer moraine
at Peoria, Illinois, showing the flood plain of the ancient stream as
an elevated terrace into which the modern stream has cut its gorge
(after Goldthwait).]

=Interference of glaciers with drainage.=—Every advance and every
retreat of a continental glacier has been marked by a complex series
of episodes in the history of every river whose territory it has
invaded. Whenever the valley was entered from the direction of its
divide, the effect of the advancing ice front has generally been to
swell the waters of the river into floods to which the present streams
bear little resemblance (Fig. 346). Because of the excessive melting,
this has been even more true of the ice retreat, but here _when the ice
front retired up the valley_ toward the divide. A sufficiently striking
example is furnished by the Wabash, Kaskaskia, Illinois, and other
streams to the southward of the divide which surrounds the basin of the
Great Lakes (Fig. 347).


FIG. 347.—Broadly terraced valleys outside the divide of the St.
Lawrence basin, which remain to mark the floods that issued from the
latest continental glacier during its retreat (after Leverett).]

Wherever the relief was small there occurred in the immediate vicinity
of the ice front a temporary diversion of the streams by the parallel
moraines, so that the currents tended to parallel the ice front. This
temporary diversion known as “border drainage” was brought to a close
when the partially impounded waters had, by cutting their way through
the moraines, established more permanent valleys (Fig. 348).

=Temporary lakes due to ice blocking.=—Whenever, on the contrary, the
advancing ice front entered a valley from the direction of its mouth,
or a _retreating ice front retired down the valley_, quite different
results followed, since the waters were now impounded by the ice front
serving as a dam. Though the histories of such blocking of rivers
are often quite complex, the principles which underlie them are in
reality simple enough. Of the lakes formed during advancing hemicycles
of glaciation, and of all save the latest receding hemicycle, no
satisfactory records are preserved, for the reason that the lake
beaches and the lake deposits were later disturbed and buried by the
overriding ice sheets. We have, however, every reason to suppose that
the histories of each of these hemicycles were in every way as complex
and interesting as that of the one which we are permitted to study.


FIG. 348.—Border drainage about the retreating ice front south of Lake
Erie. The stippled areas are the morainal ridges and the hachured bands
the valleys of border drainage (after Leverett).]

As an introduction to the study of the ice-blocked lakes of North
America, and to set forth as clearly as may be the fundamental
principles upon which such lakes are dependent, we shall consider
in some detail the late glacial history of certain of the Scottish
glens, since their area is so small and the relief so strong that
relationships are more easily seen; it is, so to speak, a pocket
edition of the history of the more extended glacial lakes.


FIG. 349.—The “parallel roads” of Glen Roy in the southern highlands
of Scotland (after Jamieson).]

=The “parallel roads” of the Scottish glens.=—In a number of
neighboring glens within the southern highlands of Scotland there are
found faint terraces upon the glen walls which under the name of the
“parallel roads” (Fig. 349) have offered a vexed problem to scientists.
Of the many scientists who long attempted to explain them, though in
vain, was Charles Darwin, the father of modern evolution. He offered it
as his view that the “roads” were beaches formed at a time when the sea
entered the glens and stood at these levels. When, however, Jamieson’s
studies had discovered their true history, Darwin, with a frankness
characteristic of some of the greatest scientists, admitted how far
astray he had been in his reasoning. Let us, then, first examine the
facts, and later their interpretation. The map of Fig. 350 will suffice
to set forth with sufficient clearness the course of the several
“roads.” These “roads” are found in a number of glens tributary to
Loch Lochy, and of the three neighboring valleys, Glen Roy has three,
Glen Glaster two, and Glen Spean one “road.” The facts of greatest
significance in arriving at their interpretation relate to their
elevations with reference to the passes at the valley heads, their
abrupt terminations down-valleyward, and the morainic accumulations
which are found where they terminate. The single “road” of Glen Spean
is found at an elevation of 898 feet, a height which corresponds to
that of the pass or col at the head of its valley and to the lowest of
the “roads” in both Glens Glaster and Roy. Similarly the upper of the
two “roads” in Glen Glaster is at the height of the pass at its head
(1075 feet) and corresponds in elevation to the middle one of the three
“roads” in Glen Roy. Lastly, the highest of the “roads” in Glen Roy is
found at an elevation of 1151 feet, the height of the col at the head
of the Glen. In the neighboring Glen Gloy is a still higher “road”
corresponding likewise in elevation to that of the pass through which
it connects with Glen Roy.


FIG. 350.—Map of Glen Roy and neighboring valleys of the Scottish
highlands with the so-called “roads” entered in heavy lines. Glens Roy,
Glaster, and Spean have three “roads”, two “roads”, and one “road”,
respectively (after Jamieson).]

To come now to the explanation of the “roads”, it may be said at
the outset that they are, as Darwin supposed, beach terraces cut by
waves, not as he believed of the ocean, but of lakes which once filled
portions of the glens when glaciers proceeding from Ben Nevis to the
southwestward were blocking their lower portions. The several episodes
of this lake history will be clear from a study of the three successive
idealistic diagrams in Fig. 351.

[Illustration: FIG. 351.—Three successive diagrams to set forth in
order the late glacial lake history of the Scottish glens.]

To derive the principles underlying this history, it is at once seen
that _all changes are initiated by the retirement of the ice front to
such a point that it unblocks for the waters of a lake an outlet that
is lower than the one in service at the time_. This is the principle
which explains nearly all episodes of glacial lake history. Thus, when
the ice front had retired so as to open direct connections between Glen
Roy and Glen Glaster, the col at the head of Glen Roy was abandoned as
an outlet, and the waters fell to the level fixed for Glen Glaster.
A still further retirement at last opened direct connection between
Glen Glaster and Glen Spean, so that the lake common to Glens Glaster
and Roy fell to the level of the col which was the outlet of the Spean
valley at the time. This stage continued until the ice front had
retired so far that the waters drained naturally down the river Spean
to Loch Lochy and thence to the ocean.

[Illustration: FIG. 352.—Harvesting time on the fertile floor of the
glacial Lake Agassiz (after Howell).]

Only in their far grander scale and in the lesser relief of the
land over which they formed, do the complex histories of the great
ice-blocked lakes of North America differ from these little valley
lakes whose beaches may be visited and the relationships worked out,
thanks to Jamieson, in a single day’s strolling.

[Illustration: FIG. 353.—Map of Lake Agassiz (after Upham).]

=The glacial Lake Agassiz.=—The grandest of the temporary lakes
referable to blocking by the continental glaciers of the ice age must
be looked for in the largest valleys that lay within the territory
invaded and _which normally drain toward the retiring ice front_.
In North America these rivers are the Red River of the North in
Minnesota, the Dakotas, and Manitoba; and the St. Lawrence River
system. To the ice dam which lay across the Red River valley we owe the
fertility of that vast plain of lake deposits where is to-day the most
intensive wheat farming of the northwest (Fig. 352). Lakes Winnipeg,
Winnipegoosis, and Manitoba, and the Lake of the Woods, are all that
now remain of this greatest of the glacial lakes, which in honor of
the distinguished founder of the glacial theory has been called Lake
Agassiz (Fig. 353). With their natural outlet blocked by the ice in
northern Manitoba and Keewatin, the waters of the Red were swollen by
melting from the retiring glacier and spread over a vast area before
finding a southern outlet along the course of the present Lake Traverse
and the valley of the Minnesota River. Along this route there flowed a
mighty flood which carved out a broad valley many times too large for
the Minnesota, its present occupant, and this giant prehistoric river
has been called the Warren River (Fig. 354).


FIG. 354.—Map of the southern end of the Lake Agassiz basin, showing
the position of some of the beaches and the outlet through the former
Warren River (after Upham).]

[Illustration: FIG. 355.—Narrows of the Warren River below Big Stone
Lake, where it passed between jaws of hard granite and gneiss (after


FIG. 356.—Map of the valley of the Warren River in the vicinity of
Minneapolis, with the young valley of the Mississippi entering it at
Fort Snelling (after Sardeson).]

It is interesting to follow this ancient waterway and to discover
that, like our normal, present-day streams, it was held up in narrows
wherever outcroppings of harder rock had constricted its channel
(Fig. 355). The upper end of the Warren River valley is now occupied
by the long and relatively narrow Lakes Traverse and Big Stone, each
the result of blocking by delta deposits where a tributary stream
has emerged into the valley, but this gigantic channel continues
down to and beyond Minneapolis, occupied as far as Fort Snelling
by the Minnesota River—a mere pygmy compared to its predecessor.
To the earnest student of glacial geology there can be few sights
more impressive than are obtained by standing at Fort Snelling, just
above the confluence of the Minnesota and the Mississippi rivers, and
surveying first the steep and narrow valley of the Mississippi above
the junction,—a stream fitted to its valley for the simple reason that
it has carved it,—and then gazing up and down that broad valley in
which the great Warren River once flowed majestically to the sea, now
the bed of the Minnesota above the Fort and of the Mississippi below it
(Fig. 356).


FIG. 357.—Portion of the Herman quadrangle of Minnesota, showing the
position of the Herman beach on the shore of the former Lake Agassiz.
The lake basin is to the left, and the pitted morainal deposits appear
to the right (U. S. G. S.).]

Just as the “parallel roads” of Glen Roy, roads in name only, are the
beaches of earlier glacial lake stages, so in Lake Agassiz we have
parallel beaches of the barrier type which are often roads in fact as
well as in name, and which mark the stages of successive lakes within
this vast basin. The Herman beach, corresponding to the highest level
of the lake, is thus a sharp topographic boundary between lake deposits
and morainal accumulations, and is further itself a well-marked
topographic feature composed of wave-washed and hence well-drained
materials (Fig. 357). Farmers of the district have been quick to
realize that these level and slightly elevated ridges lack the clay
which would render them muddy in the wet seasons, and are thus ideally
adapted for roads. They have in many sections been thus used over long
stretches and are known as the “ridge roads.”

=Episodes of the glacial lake history within the St. Lawrence
valley.=—Within this great drainage basin it has apparently been
possible to read the records of each stage in the latest lake
history—complex as this has been. We have only to recall the lake
stages cited from the Scottish glens and remember that each new stage
was begun in a retirement of the glacier front which unblocked an
outlet of lower level than the last. This sequence might, however,
have been varied by a temporary readvance of the ice, as indeed once
occurred in the Huron-Erie lobe of the great North American glacier.


FIG. 358.—The continental glacier of North America in an early stage
of its recession, when it covered the entire St. Lawrence drainage
basin. The dashed line is the approximate position of the divide (based
on a map by Goldthwait).]


FIG. 359.—Outline map of the early Lake Maumee, with the bordering
moraine and the water-laid moraine remaining on the site of the former
ice cliff.]

=The crescentic lakes of the earlier stages.=—So long as the glacier
covered the entire drainage basin of the St. Lawrence River system, all
water was freely drained away by streams which flowed _away from_ the
ice front (Fig. 358). So soon, however, as at any point the front had
retired behind the divide, impounding of the waters must locally have
occurred. Lakes of this type are to-day to be seen in Greenland and
in the southern Andes; and though upon a diminutive scale, some idea
of their aspect may be obtained from the appearance of the Märjelen
Lake of Switzerland, here blocked by a mountain glacier (Fig. 446, p.
411). Within all areas of small relief, such as the prairie country
surrounding the present Laurentian lakes, the earlier and smaller
stages of such ice-blocked lakes are generally crescentic in outline.
This is because a moraine in most cases forms the land margin of the
lake, and because the ice cliff upon the opposite border, although
somewhat straightened, as a consequence of wave-cutting and iceberg
formation, still retains the convex outlines characteristic of ice
lobes (Fig. 359).

[Illustration: FIG. 360.—Map to show the first stages of the
ice-dammed lakes within the St. Lawrence basin (after Leverett and

Within each of the Great Lake basins a crescentic lake early appeared
at that end of the depression which was first uncovered by the
glacier: Lake Duluth in the Superior basin, Lake Chicago in the
Michigan basin, and Lake Maumee in the Huron-Erie basin (Fig. 360).

We may now, with profit, trace the successive episodes of the glacial
lake history, considering for the earlier stages those changes which
occurred within the Huron-Erie basin, since, these are in essential
respects like those of the Michigan and Superior basins, although
worked out in greater detail. Lake Chicago must, however, be brought
into consideration, since in all save the earliest and the later
stages, the waters from the Huron-Erie depression were discharged
through the Grand River into this lake and thence by the so-called
“Chicago outlet” into the Mississippi (plate 20 A).

=The early Lake Maumee.=—The area, outline, and outlet of this lake
are indicated upon Fig. 360. Its ancient beaches have been traced, as
well as the water-laid moraine beneath its former ice cliff; and no
observant traveler who should take his way down the ancient outlet
from Fort Wayne, Indiana, past the town of Huntington, could fail to
be impressed by its size, suggesting as it does the great volume of
water which must once have flowed along it. Now a channel a mile or
more in width, its bed for the twenty-five miles between Fort Wayne
and Huntington may be seen from the tracks of the Wabash Railway as a
series of swamps merely, while at Huntington the Wabash river enters by
a young V-shaped valley at the side, much as the Mississippi emerges
into the old channel of the Warren River at Fort Snelling, Minnesota
(see p. 327).

The Huron River of southern Michigan, which now discharges into Lake
Erie, then found its lower course blocked by the glacier and was thus
compelled to find a southerly directed channel now easily followed to
the northern horn of the crescent of Lake Maumee.

=The later Lake Maumee.=—When the ice lobe had retired its front
sufficiently, an outlet lower than that at Fort Wayne was uncovered
past the city of Imlay, Michigan, into the Grand River, and thence
through Lake Chicago and its outlet into the Mississippi. This old
outlet south of Chicago follows the course of the present Drainage
Canal and the line of the Chicago & Alton Railway. The traveler
journeying southward by train from Chicago has thus the opportunity of
observing first the beaches of the former lake, and then the several
channels which were joined in the main outlet at the station of Sag
(plate 20 A).

[Illustration: FIG. 361.—Outline map of the later Lake Maumee and of
its “Imlay outlet” to Lake Chicago (after Leverett).]

In this stage of our history Lake Maumee pushed a shrunk arm up past
the site of Ypsilanti in Michigan (Fig. 361), the well-marked beach
being found on Summit Street opposite the State Normal College. The
Huron River, which in the first lake stage had followed the valley now
occupied by the Raisin River southward into Indiana, now discharged
directly into a bay upon this arm of Lake Maumee, and so formed a delta
at Ann Arbor.

[Illustration: FIG. 362.—Outline map of Lakes Whittlesey and Saginaw
(after Leverett).]


FIG. 363.—Map of the glacial Lake Warren, the last of the lakes in the
Huron-Erie basin, which discharged through the “Grand River outlet”
into the Mississippi (after Leverett).]

=Lakes Arkona and Whittlesey.=—The ice front in the Huron-Erie basin
now retired so far that the impounded waters, instead of following
the more direct “Imlay outlet” to the Grand, passed at a lower level
completely around “the thumb” of Michigan into the Saginaw basin.
Meanwhile a crescent-shaped lake had developed in that basin, so that
now the waters of the Maumee basin were joined to those in the Saginaw
basin as a common lake, just as the lowering of the waters in Glen Roy
caused a union with those of Glen Glaster in the example cited for
illustration. Our records of this third North American lake stage,
referred to as Lake Arkona, are however most imperfect, for the reason
that it was followed by a readvance of the ice front which closed the
passage around “the thumb” and raised the level of the waters until
an outlet was found past the town of Ubly at a lower level than the
“Imlay outlet.” When the waters of a lake are thus rising, strong beach
formations result, and those of this stage, which is known as the Lake
Whittlesey stage, are much the strongest that are found within the
Huron-Erie basin. Traced for some three hundred miles entirely around
the southern and western margins of Lake Erie, this beach is for much
of the distance the famous “ridge road” (Fig. 362).

=Lake Warren.=—As the ice advance which had produced Lake Whittlesey
came to an end, the normal recession was resumed and a lake once more
formed as a body common to the Saginaw and Erie basins. This lake,
known as Lake Warren, extended a shrunk arm far eastward along the ice
front into western New York, though it was still blocked from entering
the great Mohawk valley (Fig. 363).

[Illustration: FIG. 364.—Map of the Glacial Lake Algonquin (after

=Lakes Iroquois and Algonquin.=—It must be evident that toward the
close of the Lake Warren stage a profound change was imminent—a
transfer of the glacial waters from their course to the Mississippi
and the Gulf to the trench which crosses New York State and enters the
Atlantic. So soon as the ice front had retired sufficiently to lay
bare the bed of the Mohawk, an outlet was found by this route and its
continuation down the Hudson valley to the sea. The Lake Ontario basin
now became occupied by a considerably larger water body known as Lake
Iroquois, and the three upper lakes, then joined as Lake Algonquin,
discharged their combined waters into Lake Iroquois at first through a
great channel now strongly marked across Ontario in the course of the
Trent River and Lake Simcoe, the so-called “Trent outlet.” At this time
a smaller Lake Erie probably occupied the basin of that lake, and later
the Trent outlet was abandoned for the Port Huron outlet (Fig. 364).

[Illustration: FIG. 365.—Outline map of the Nipissing Great Lakes with
their outlet past North Bay into the Champlain Sea.]

=The Nipissing Great Lakes.=—We have now followed the ice front
step by step in its retreat across the valley of the St. Lawrence
system. The successive unblocking of outlets offers but one further
possibility—the opening of the French River-Nipissing Lake-Ottawa
River, or “North Bay outlet.” Though not so to-day, the bed of this
ancient channel was then much lower than that of the “Mohawk outlet”,
and so soon as the glacier had in its retreat uncovered this northern
channel, the waters of the upper lakes discharged through it past the
site of Ottawa and into an arm of the sea which then occupied the lower
St. Lawrence valley and has been called the Champlain Gulf or Sea
(Fig. 365). The level of the waters was lowered and the area of the
lakes correspondingly reduced.

The reader who has had no opportunity to observe these ancient channels
which carried the swollen waters of the former glacier lakes, will find
it interesting to consider that every one of them has been fixed upon
by engineers for improvement as artificial waterways. Thus we have the
Illinois Drainage Canal and projected ship canal along the “Chicago
outlet”, the projected Mississippi-Lake Erie Canal along the “Fort
Wayne outlet”, the Grand River canal project to connect Lake Michigan
and Saginaw Bay along the course of the “Grand River outlet”, the
Trent Canal along the “Trent outlet”, the Erie Canal along the “Mohawk
outlet”, and, lastly, the proposed Georgian Bay ship canal to the ocean
along the “North Bay” or “Nipissing outlet.”

=Summary of lake stages.=—We have omitted in this summary of late lake
history in the Laurentian basin all the less important lake stages,
including some of a transitional nature which were represented by
beaches and outlets easily traced to-day. This is because it is an
outline only which it seems best to present, and the episodes of this
abridged history may be tabulated as follows:



  Lake Maumee (early), Fort Wayne outlet.
  Lake Maumee (late), Imlay City outlet.
  Lake Arkona, “thumb” outlet.
  Lake Whittlesey (with readvance of glacier), Ubly outlet.
  Lake Warren, “thumb” outlet.


  Lakes Iroquois and Algonquin (early), Trent and Mohawk outlets.
  Lakes Iroquois and Algonquin (late), Port Huron and Mohawk outlets.
  Nipissing Great Lakes, North Bay outlet.

=Permanent changes of drainage affected by the glacier.=—While the
lake history which we have sketched is made up of episodes which
endured only while the ice front lay between certain stations upon its
retreat, there were none the less brought about the profoundest of
permanent modifications in the drainage of the region. It is possible
to restore upon maps in part only the preglacial drainage of the north
central states, but we know at least that it was as different as may be
from that which we find to-day. The Missouri and the Ohio take their
courses to-day along the margin of the glaciated area as an inheritance
from the border drainage of the ice age. Within the glaciated regions
rivers have in many cases been compelled by morainal obstructions to
enter upon new courses, or even to travel in the opposite direction
along their former channels. In districts of considerable relief these
diversions have sometimes caused the streams to plunge over the walls
of deep valleys, and it may truthfully be said that we owe much of our
most beautiful scenery in part to the carving and molding of glaciers,
but especially to the cascades and waterfalls directly due to their
interference with drainage.


FIG. 366.—Probable preglacial drainage of the upper Ohio region (after
Chamberlin and Leverett).]

Many diversions or reversals of former drainage lines, through the
influence of the continental glacier, are at once suggested by
the abnormal stream courses, which appear upon our maps, and the
correctness of these suggestions may often be confirmed by very simple
observations made upon the ground. The map of Fig. 366 shows how
different was the preglacial drainage of the upper Ohio region from
that of to-day.

An interesting additional example is furnished by the Still River
which in Connecticut is tributary to the Farmington, and is no less
remarkable for its abnormal northerly course and sluggish current
perpetuated in its name, than for the way in which it is joined to the
Farmington system (Fig. 367 _A_). A careful study of the district has
shown that the Still River was once a part of the Naugatuck and flowed
southward toward Long Island Sound like other rivers of the district
(Fig. 367 _B_). It possessed, however, an advantage in a narrow belt
of softer rock along its course, and because of this advantage it
captured a portion of one of the tributaries to the Farmington (Fig.
367 _C_). The continental glacier later covered the region, and on its
retreat laid down morainal obstructions directly across this river and
also at the head of the severed arm of the Farmington tributary (Fig.
367 _D_). The now impounded waters found their lowest outlet near Sandy
Brook, and in waterfalls and cascades the now reversed river falls one
hundred feet to the bed of that stream. With the aid of the excellent
topographic maps which are now supplied by a generous government at a
merely nominal price, such bits of recent history may be read at many
places within the glaciated region.


FIG. 367.—Diagrams to illustrate the episodes in the recent history
of the Still River tributary to the Farmington in Connecticut. _A_,
present drainage; _B_, early stage; _C_, after capture of a tributary
to the Farmington; _D_, after blocking by morainal obstructions of the
ice age.]

=Glacial Lake Ojibway in the Hudson Bay drainage basin.=—When by
passing over the “height of land” in northern Ontario the greatly
reduced continental glacier had vacated the basin of St. Lawrence
drainage, it was in a position to impound those waters which normally
drained to Hudson Bay. The lake which then came into existence has been
called Lake Ojibway and was the latest of the entire series. Though of
but recent discovery in a country till lately a trackless wilderness,
its extension seems to have been that of the clay beds suited for
farming. The beaches and outlets remain to be mapped when the country
has been made more easily accessible.


 Parallel roads of Glen Roy:—

 CHARLES DARWIN. Observations on the Parallel Roads of Glen Roy and of
 Other Parts of Lochaber in Scotland, with an attempt to prove that
 they are of Marine Origin, Phil. Trans., vol. 8, 1839, pp. 39-82.

 LOUIS AGASSIZ. Geological Sketches, Boston, 1876, vol. 2, pp. 32-76.

 T. T. JAMIESON. On the Parallel Roads of Glen Roy and their Place in
 the History of the Glacial Period, Quart. Jour. Geol. Soc. Lond., vol.
 19, 1863, pp. 235-259.

Glacial Lake Agassiz:—

 WARREN UPHAM. The Glacial Lake Agassiz. Mon. 25, U. S. Geol. Surv.,
 pp. 658, pls. 38.

 F. W. SARDESON. Beginning and Recession of St. Anthony’s Falls, Bull.
 Geol. Soc. Am., vol. 19, 1908, pp. 29-36.

Glacial lakes in the St. Lawrence valley:—

 CHAMBERLIN AND SALISBURY. Geology, vol. 3, pp. 394-405.

 FRANK LEVERETT. Outline of the History of the Great Lakes
 (Presidential Address), 12th Rept. Mich. Acad. Sci., 1910, pp. 19-42.
 The Pleistocene Features and Deposits of the Chicago Area. Chicago,
 1897, pp. 86, pls. 8 (Chicago Outlet).

 H. L. FAIRCHILD. Glacial Lakes in Western New York, Bull. Geol. Soc.
 Am., vol. 6, 1895, pp. 353-374, pls. 18-23; Glacial Waters in Central
 New York. Bull. 127, N. Y. State Mus., 1909, pp. 66, pls. 42, and maps
 in cover.

Early lakes in the Erie basin:—

 FRANK LEVERETT. On the Correlation of Moraines with Raised Beaches of
 Lake Erie, Am. Jour. Sci. (3), vol. 43, 1892, pp. 281-301.

 F. B. TAYLOR. The Great Ice Dams of Lakes Maumee, Whittlesey, and
 Warren, Am. Geol., vol. 24, 1899, pp. 6-38, pls. 2-3; Relation of Lake
 Whittlesey to the Arkona Beaches, 7th Rept. Mich. Acad. Sci., 1905,
 pp. 30-36.

 FRANK LEVERETT. The Ann Arbor Folio, Folio No. 155, U. S. Geol. Surv.,
 1908, pp. 10-12.



=The response of the earth’s shell to its ice mantle.=—There is now
good reason to believe that the earth’s outer shell makes a response by
oscillations of level due to the loading by ice, on the one hand, and
to the removal of this burden upon the other. We know, at least, that
both in northern Europe and in North America areas which have undergone
depression during and elevation after the ice age, correspond closely
to the regions which were ice covered. Wherever in these regions there
was high relief before the advent of the ice, river valleys were
drowned at the land margins and were also gouged out into troughs
through erosion by the outlet tongues upon the margin of the ice sheet.
Such furrowed and half-submerged valleys have a characteristic U-shaped
section, so that their walls rise precipitously from the sea. From
their typical occurrence in Scandinavian countries the name _fjord_ has
been applied to them.

It is now no less clear that the removal of the ice blanket brought
from the earth a relatively quick response in uplift, which began
before the ice front had retired across the present international
boundary of the United States, and that this uplift continued until the
final disappearance of the ice. A far slower elevation of a somewhat
different nature has continued, even to the present day.

It is obvious that at the time of their formation all shore lines
referable to the work of waves must have been horizontal, and hence
any variations from a perfect level which they reveal to-day must
indicate that a tilting movement of the ground has occurred since the
waters departed from their basins. We have thus provided for us in the
positions of these ancient water planes, particularly because of their
wide extent, a complete record the refinement of which is not easily
overstated. Interpreting this record, we find that it was the uptilt
of the land to the northward which brought the glacial lake history to
an end and inaugurated the present system of St. Lawrence drainage. The
outlet of the Nipissing Great Lakes is to-day more than a hundred feet
above the level of the outlet at Port Huron, where the upper lakes are
now discharging their waters, and this difference in level can only
be ascribed to an upward tilting of the land since the latest of the
glacial lake stages.

=The abandoned strands as they appear to-day.=—The traveler by steamer
upon the upper lakes, as he comes within view of each rocky headland,
may note how the profile against the horizon is notched by a series of
steps or terraces (Fig. 368), and if he has followed the discussion in
previous chapters, he will suspect that these terraces mark the now
abandoned shore lines which have come to their present position through
a series of uplifts of the ground accompanied by earthquake shocks. As
his steamer skirts the shore he may chance to note a cave within the
rock cliff which represents the now elevated sea-arch of an ancient


FIG. 368.—The notched rock headland of Boyer Bluff between Green Bay
and Lake Michigan (after Goldthwait).]

Disembarking from the steamer and traveling inland at any point where
the shores are high, the traveler is certain to come upon still more
convincing proofs of the ancient strands; perhaps in a storm beach of
the unmistakable “shingle”, half buried though it may be under dunes
of newly drifted sand, or possibly at higher levels the highway has
been cut through a shingle barrier as fresh and unmistakable as though
formed upon the present shore. Sometimes it is the rock cliff and
terrace, at other times barrier ridges of shingle, or, again, it is the
sloping cliff and terrace cut in the drift deposits; but of whatever
sort, if studied with proper regard to the topography of the district,
the evidence is clear and unmistakable.

=The records of uplift about Mackinac Island.=—Nowhere are the records
of the recent uplift of the lake region more easily read than about
Mackinac Island in the straits connecting Lake Michigan with Lake
Huron. Approaching the island by steamer from St. Ignace, its profile
upon the horizon is worthy of remark (Fig. 369). From a central crest
broken by minor irregularities and bounded on all sides by a cliff, the
island profile slopes gently away to a still lower cliff, below which
is another terrace.


FIG. 369.—View of Mackinac Island from the direction of St. Ignace.
The irregular central portion is the only part of the island that was
not submerged in Lake Algonquin. The terrace at its base is the old
shore line of Lake Algonquin, and the lower terrace the strand of Lake
Nipissing (after a photograph by Taylor).]


FIG. 370.—The “Sugar Loaf”, a stack near the shore of Lake Algonquin,
as it is seen from the observatory upon Mackinac Island (after a
photograph by Taylor).]

When we have reached the island and have climbed to the summit, we
there find the surface which is characteristic of erosion by running
water, whereas at lower levels are found the forms carved or molded
by the action of waves. This central “island”, superimposed upon the
larger island, is all that rose above Lake Algonquin, the earliest of
the glacial lakes in this northern district; and as we look out from
the observatory upon the summit, it is easy to call up a picture of the
country when the lake stood at the base of this highest cliff. To the
northward one sees the “Sugar Loaf” rise out of a sea of foliage, as it
formerly did from the waters of Lake Algonquin (Fig. 370). It is a huge
stack near the former island shore. If we turn now to the southward and
direct our gaze toward the Fort, we encounter a veritable succession of
beach ridges formed of shingle and ranged like a series of waves within
the cleared space of the “Short Target Range” (Fig. 371). These ridges
mark each a stage within a series of successive uplifts which have
brought the island to its present height.


FIG. 371.—View from the observatory upon Mackinac Island across the
“Short Target Range” toward the Fort. Beach ridges appear in succession
within the cleared space (after a photograph by Rossiter).]

[Illustration: FIG. 372.—Notched stack of the Nipissing Great Lakes at
St. Ignace (after a photograph by Taylor).]


FIG. 373.—Series of diagrams to illustrate the evolution of ideas
concerning the uplift of the lake region since the ice age. _A_,
simple northerly up-canting (Gilbert): _B_, northerly acceleration of
the up-canting (Spencer and Upham); _C_, northerly “feathering out”
of beaches (Spencer and Upham); _D_, hinge, line of up-canting found
within the lake region (Leverett); _E_, multiple and northwardly
migrating hinge lines of up-canting (Hobbs).]

If now we descend from our position and visit the “battlefield”, we
find there a great ridge of level crest, behind which the British
force was stationed in its defense of the island in 1812. Near by in
the woods is Pulpit Rock, a strikingly perfect stack of the Nipissing
Lake. Across the straits at St. Ignace is an even finer example of the
notched stack (Fig. 372). Other less prominent beaches, but all later
than the Nipissing Lakes, intervene between this level and the present
shore to mark the stages in the continued uplift of the land.

=The present inclinations of the uplifted strands.=—It is not enough
that we should have recognized the marks of former shores now at
considerable elevations above the existing lakes; if we are to know
the nature of the uplift, we must prepare accurate maps based upon
measurements by precise leveling at many localities. Such methods are,
however, of comparatively recent application in this field; and, as in
the investigation of so many other problems, the earlier observations
were largely of the nature of reconnaissances with the elevation of
beaches estimated by comparatively crude methods only. The evolution of
ideas concerning the uptilt has, therefore, been a gradual one.


FIG. 374.—Map of the Great Lakes region to show isobases and hinge
lines of uptilt. _a_, isobase of the Chicago outlet; _b_, main hinge
line of the Lake Whittlesey beach (Leverett); _b^1_, hinge line of the
Lake Warren beach (Taylor); _c_, isobase of the Port Huron outlet; _d_,
main hinge line of highest Algonquin beach (Goldthwait); _e_, _f_,
_g_, _h_, additional hinge lines of Algonquin beaches in Door County
peninsula (Hobbs); _l_, isobase of the Lake Superior outlet for the
Algonquin beaches (Leverett): _m_, isobase of the same outlet for the
Nipissing beaches (Leverett).]

It was early observed that the beaches corresponding to a given
lake stage were higher to the northward and northeastward, and the
natural conclusion from this was that the earth’s crust had here been
canted like a trap door (Fig. 373, _A_). As we are to see, this but
half-correct assumption has led to a striking prophecy relating to
future changes within the lake region which we now know to be without
warrant in the facts. Later it was learned that the uptilt of the lake
beaches is much accelerated to the northward (Fig. 373, _B_), and that
new beaches make their appearance from beneath others as we proceed in
this direction—there is a “feathering out” of beaches to the northward
(Fig. 373, _C_).

=The hinge lines of uptilt.=—Still later in the study of the region,
it was learned that the axis or fulcrum about which the region has been
uptilted, instead of lying to the southward of the lake district, as
had been assumed by Gilbert, lay within the region and about halfway up
the basin of Lake Michigan (Fig. 373, _D_, and Fig. 374). Similarly, in
the uptilt which followed the ice retreat in northern Europe a definite
hinge line of movement has been discovered.

Lastly, it has been shown, as a result of the use of precise leveling
methods, that not one but several hinge lines of movement lie within
the region, and that the separate sections into which they divide the
area are each in turn characterized by increased up-cant as we proceed
to the northward (Fig. 373, _E_ and Fig. 374).


FIG. 375.—Series of idealistic diagrams to indicate the nature of the
quick recovery of the crust by uplift in blocks unloaded of the ice in
succession. A further and slower uptilt, added after the completion of
the first movement, is brought out in the last diagram (_b_´).]

The beaches of Lake Maumee, the earliest of the series of lakes within
the Huron-Erie lobe and within the extreme southern portion of the
Great Lakes area, show only the slightest possible northerly uptilt,
and the well-marked hinge line disclosed in the Whittlesey beach is
evidence that the elastic recoil, as it were, from the weight of
the mantling glacier did not begin until after the draining of Lake
Whittlesey. The determination by Taylor that there is a similar initial
hinge line in the Warren beach—that this strand begins its uptilt some
fifteen miles farther northeast than does the Whittlesey beach—is
one of the greatest importance in obtaining a correct idea of the
recent uplift; for it shows that the draining of Lake Whittlesey was
followed by a period of quick uplift and seismic activity, that the
stage of Lake Warren was one of comparative stability of the land,
and, lastly, that the draining of Lake Warren was followed by a second
period of rapid uplift and earthquake disturbance. The strongly marked
hinge lines, additional to the initial one indicated for the Algonquin
beaches in the profiles by Goldthwait from the west shore of Lake
Michigan, when considered in the light of this northeasterly migration
of the still earlier hinge line in the southern district, are best
explained through the assumption of a succession of quick recoveries of
the crust by uplift, separated by periods of relative stability, and
brought on by the removal in turn of the ice burden from successive
blocks of the shell which are separated by the several hinge lines
(Fig. 375).

The elaborate study of erosion in the outlet of Lake Agassiz had
indicated identical interruptions in the up-canting process for that

=Future consequences of the continued uptilt within the lake
region.=—One of the most distinguished of American geologists, Dr.
G. K. Gilbert, in order to determine whether the uptilt revealed by
canted beach lines is still in progress, carried out an elaborate
study upon the gauge records preserved at the various gauging stations
about the Great Lakes. Upon the basis of these studies, he concluded
that the uplift continues, that the axes of equal uplift (isobases)
take their course about fifteen degrees north of west, so that the
lines of greatest uptilt should be perpendicular to this direction,
or fifteen degrees east of north. He further believed that the basin
was undergoing an up-cant in the simple manner of a trap door, the
hinge of which lay to the southward of Chicago, and the study of the
gauge records led him to believe that “the rate of change is such
that the two ends of a line one hundred miles long and lying in a
south-southwest direction are relatively displaced four tenths of a
foot in one hundred years.”

=Gilbert’s prophecy of a future outlet of the Great Lakes to the
Mississippi.=—The _natural_ rock sill, over which the waters of Lake
Chicago once flowed to the Mississippi, is to-day but eight feet above
the common mean level of Lakes Michigan and Huron, and if the tilting
of the lake region were to continue upon Gilbert’s assumption of a
canting plane with the hinge of the movement to the south of Chicago, a
time must come when the “Chicago outlet” will again come into use and
the lakes once more drain to the Mississippi and the Gulf. Upon the
basis of his measurements, Gilbert ventured the prophecy that the first
high-water discharge into the Mississippi should occur in from five
hundred to six hundred years, and for continuous discharge in fifteen
hundred years. In twenty-five hundred years Niagara Falls should at low
water stages be dry from this cause, and in thirty-five hundred years
it should have become extinct.

This prophecy, emanating from a high scientific authority and relating
to changes of such profound economic and commercial importance,
has been often quoted and has taken a firm hold upon the popular
imagination. Obviously, it depends upon the now exploded theory that
the lake basin has been canted _as a plane_ and that the axis of uptilt
lies somewhere to the southward of the lake region, or, in any event,
to the southward of the present Port Huron outlet. We know to-day that
instead of being uniformly distributed over the entire lake region,
the uptilting goes on at a much higher rate within the northern areas,
and that since the early stage of Lake Whittlesey the hinge line of
uplift has been steadily migrating northward with the retreat of the
ice and is now well to the northward of the present outlet. There is,
therefore, no known uptilt of the district which separates the present
from the former Chicago outlet, and there is no apparent natural
cause which should result in the reoccupation of the old outlet to
the Mississippi. The prophecy must be regarded as one that has been
outgrown with the progress of science.

=Geological evidences of continued uplift.=—It has recently been
claimed, on the basis of a reëxamination of Gilbert’s study of the
lake gauge records, that his methods are open to serious criticism and
that in reality the figures afford no evidence of continued uplift
of the region. However this may be, there are not lacking geological
evidences which do not admit of doubt, and these are in a striking way
confirmatory of the latest conclusions upon the manner of the recent

If our conclusions have been correct, the several lake basins should
now be behaving in different ways as regards the changes upon
their shores. If it is true that the lines of greatest uptilt run
north-northeasterly, there should be, speaking broadly, a “spilling
over” of waters upon the south-southwesterly shores and a laying bare
of the north-northeasterly shore terraces of the basins. This should,
however, be true only of basins whose outlets are to the northeastward
of the existing main hinge line of uptilt. Lake Huron, having its
outlet at the southern margin of its basin, should not have its waters
encroaching upon the southern shore, for the simple reason that any
continued uptilt of the basin can only have the effect of pouring more
water through the outlet. Lake Michigan and Saginaw Bay, which are
arms of the Huron basin, ought, however, to become flooded upon their
southern shores, _were it not that the hinge line of uptilt to-day lies
to the northward of the outlet at Port Huron, and, further, that the
two connecting channels still have their beds lower than the sill of
the outlet channel_. Now the evidence goes to show that no encroachment
of waters is occurring upon the Chicago shore of Lake Michigan, and
although the shores of Saginaw Bay are so excessively flat as to reveal
slight changes of level by large migrations of the strand, yet the
ancient meander posts fixed by the early surveys are still found near
the water’s edge.

=Drowning of southwestern shores of Lakes Superior and Erie.=—Within
the basins occupied by Lakes Superior and Erie, a wholly different
condition is found. In each case the outlet is found to the
northeastward (Fig. 374, p. 345), and the northwesterly trend of the
isobases from these outlets is responsible for a continued elevation
from uptilt of the outlets with reference to the western and southern
shores. In consequence, the waters are encroaching upon these shores,
and rivers which there enter the lake are drowned at their mouths, with
the formation of estuaries. Upon Lake Superior these changes are very
marked near Duluth and particularly in the St. Louis River, within
which, since the early treaty with the Indians, certain rapids have
disappeared and submerged trunks of trees are now found in the channel
of the river. As far east as Ontonagon essentially the same conditions
are found.

Upon the shores within the Porcupine Mountain district, the waters are
clearly rising. Here old cedar trees may be seen, in some cases dead
but still upright and standing in from six to eight inches of water a
number of feet out from the present shore, while others near the shore,
but upon the land and still living, are washed by the waves, and losing
their lower bark in consequence. An old road along the shore has had to
be abandoned because of the encroaching water.

Upon the opposite or northeastern shore of the lake, on the other hand,
the land is everywhere rising out of the water, and the waves are now
building storm beaches well out upon the wave-cut terrace. Here the
streams, instead of forming estuaries by drowning, drop down in rapids
to the level of the lake.


FIG. 376.—Portion of the Inner Sandusky Bay, to afford a comparison of
the shore line of 1820 with that of to-day (after Moseley).]

At the southwestern margin of Lake Erie there is everywhere evidence of
a rapid encroachment by the water. In the caves of South Bass Island
stalactites, which must obviously have formed above the lake level, are
now permanently submerged. It is, however, about Sandusky Bay upon the
southwest shore that the most striking observations have been made.
Moseley has collected historical records of the killing of forest
trees through a submergence which was the result of an advance of the
water upon the shores. It seems to be proven from his studies that the
water is now rising in Sandusky Bay at a rate of about 2.14 feet per
century. In Fig. 376 there is a comparison of the shores of the inner
bay separated by an interval of about ninety years.


 Uptilt in basin of Lake Agassiz:—

 WARREN UPHAM. The Glacial Lake Agassiz, Mon. 25, U. S. Geol. Surv.,
 pp. 474-522.

Uptilt in Laurentian Basin:—

 G. K. GILBERT. Recent Earth Movement in the Great Lakes Region, 18th
 Ann. Rept. U. S. Geol. Surv., 1898, Pt. ii, pp. 595-647.

 J. W. SPENCER. Deformation of the Algonquin Beach, etc., Am. Jour.
 Sci. (3), vol. 41, 1891, pp. 14-16.

 F. B. TAYLOR. The Highest Old Shore Line of Mackinac Island, Am. Jour.
 Sci. (3), vol. 43, 1892, pp. 210-218.

 A. C. LAWSON. Sketch of the Coastal Topography of the North Side of
 Lake Superior, with reference to the abandoned strands, etc., 20th
 Ann. Rept. Geol. and Nat. Hist. Surv. Minn., 1893, pp. 181-289, pls.

 J. B. WOODWORTH. Ancient Water Levels of the Champlain and Hudson
 Valleys, Bull. 84, N.Y. State Mus., 1905, pp. 265, pls. 28.

 E. L. MOSELEY. Formation of Sandusky Bay and Cedar Point, Proc. Ohio
 State Acad. Sci., vol. 4, 1905, Pt. v, pp. 179-238.

 F. E. WRIGHT. Rept. Geol. Surv. Mich. for 1903, 1905, p. 37.

 J. W. GOLDTHWAIT. The Abandoned Shore Lines of Eastern Wisconsin,
 Bull. 17, Wis. Geol. and Nat. Hist. Surv., 1907, pp. 134, pls. 37; A
 Reconstruction of Water Planes of the Extinct Glacial Lakes in the
 Lake Michigan Basin, Jour. Geol., vol. 16, 1908, pp. 459-476; Isobases
 of the Algonquin and Iroquois Beaches and their Significance, Bull.
 Geol. Soc. Am., vol. 21, 1910, pp. 227-248, pl. 5; An Instrumental
 Survey of the Shore Lines of the Extinct Lakes Algonquin and Nipissing
 in Southwestern Ontario, Mem. 10, Dept. of Mines, Canada, 1910, pp.
 57, pls. 4.

 WILLIAM H. HOBBS. The Late Glacial and Post-glacial Uplift of the
 Michigan Basin, Pub. 5, Mich. Geol. and Biol. Surv., 1911, pp. 68,
 pls. 2.

 LAWRENCE MARTIN. [Post-glacial Modifications in and Around the Great
 Lakes], Mon. 52, U. S. Geol. Surv., 1911, pp. 455-459.

Uptilt in northern Europe:—

 G. DE GEER. Quaternary Changes of Level in Scandinavia, Bull. Geol.
 Soc. Am., vol. 3, 1892, pp. 65-68, pl. 2.

 H. MUNTHE. Studies in the Late Quaternary History of Southern Sweden,
 paper No. 25, Livret Guide, Cong. Géol. Intern., 1910, pp. 96, many
 plates and maps.



=Features in and about the Niagara gorge.=—A striking example of those
permanent alterations of drainage which have resulted from the presence
of the late continental glacier in North America is to be found in the
Niagara gorge between Lakes Erie and Ontario. With the aid of borings
many of the now buried channels of the region have been followed out,
and in a later paragraph we shall refer to some of the stronger lines
of the earlier drainage system. Before undertaking the study of Niagara
history, it is essential that one become somewhat familiar with the
present topography in and about the Niagara gorge.

Below the present cataract the river flows through a deep gorge for
about seven miles before issuing at the Lewiston Escarpment (Fig. 381,
p. 355). This gorge has been cut in beds of rock sediments which dip
at a gentle angle southward toward Lake Erie. The capping of the rock
series is a compact and relatively resistant limestone which is known
as the Niagara limestone, beneath which there are alternating beds
of shale with thinner limestone and sandstone. The plain formed by
the upper surface of the limestone capping terminates in the Lewiston
Escarpment, which is transverse to the direction of the gorge and seven
miles distant below the Falls. The depth of the gorge varies markedly,
the above-water portion being represented at the upper end by the
height of the cataract, one hundred and sixty-five feet, while at its
lower end near Lewiston it is twice that amount. Halfway down the gorge
a sharp turn is made at an angle of more than ninety degrees, and the
upstream arm is extended to form a basin which contains the famous
whirlpool. This visible extension of the upper gorge is continued in a
buried channel, the St. Davids Gorge, which extends to the escarpment,
broadening as it does so in the form of a trumpet. The materials which
fill this earlier channel are notably coarse glacial deposits (Fig.


FIG. 377.—Ideal cross section of the Niagara gorge to show the
marginal terrace.]

Directly above the whirlpool the Niagara gorge is first contracted, but
almost immediately swells out into the form of a sausage, which under
the name of the Eddy Basin extends to the constricted channel occupied
by the Whirlpool Rapids. This Gorge of the Whirlpool Rapids extends to
and a little above the railroad bridges, where it again suddenly widens
and deepens and with surprisingly uniform cross section now continues
as far as the cataract. This uppermost section is known as the Upper
Great Gorge. About a mile below the whirlpool is that remarkable
projection into the gorge from the Canadian wall which is known as
Wintergreen Flats, below which and nearer the river are Fosters Flats.
Almost throughout its entire length the Niagara gorge is bordered on
either side by a narrow and gently incurving terrace eroded below the
general level of the plain and meeting the gorge in a sharp angle (Fig.

The features immediately about the cataract show that the Falls are
to-day in a condition which, so far as we know, has occurred but once
before in their entire history—the waters of the river are divided
unequally by an island, and for this reason, as we shall see, the
cataract enters over the _side wall_ of the gorge instead of at its
_end_ (Fig. 381), although the turning of the channel from this cause
is combined with a bend of the river.


FIG. 378.—View of the bed of the Niagara River above the cataract,
where water has been drained off in installing a power plant. Some
separated blocks of limestone are still in place (after J. W. Spencer).]

=The drilling of the gorge.=—There appear to be two important
processes which are responsible for the recession of the Falls, the
rate of which is determined largely by the resistance of the limestone
capping and the tenacity of the looser shale beneath it. One of the
eroding processes operates from below and undermines the cap until
the unsupported cornice falls in blocks to the bottom of the gorge;
the other makes its attack directly from above, selecting for the
purpose the lines of jointing of the rock which it widens by solution
and corrasion until the included blocks are in so far separated that
they are torn out and go over the brink of the Falls (Fig. 378). This
process of overhead attack in the powerful currents just above a
cataract is even better illustrated by the Falls of St. Anthony near
Minneapolis, which have had a similar history of recession to that of
the Niagara Falls (Fig. 379).

[Illustration: FIG. 379.—Falls of St. Anthony, looking westward from
Hennepin Island in 1851 (after N. H. Winchell, daguerreotype by Hessler
of Chicago).]

The blocks of the capping limestone at Niagara Falls are to some
extent fixed in size by the joint planes present in them, and as they
fall to the bottom of the gorge, they promote or retard the further
recession of the Falls according as they can or cannot be moved about
by the churning currents beneath the cataract. Of the retarding effect
there is an illustration in the accumulation of the blocks below the
American and the intermediate Luna Falls (plate 23 A), which the weaker
currents upon the American side find too heavy to handle.


FIG. 380.—Ideal section to show the nature of the drilling process
beneath the cataract.]


FIG. 381.—Plan and section of the Niagara gorge, showing how in each
section the depth is proportional to the width, except in the lowest
section where subsequent river action of the normal type has modified
the bed of the channel (plan after Taylor and section after Gilbert).]

The Canadian Fall, with its much greater power, is an example of the
promotion of recession through the churning about of the blocks at the
base of the cataract. We have here to do with a churn drill which bores
its way into the bottom of the gorge with increasing radius of rotary
motion with each increase in volume of the falling water. Under this
rotary churning the soft shales are torn out near the bottom and in
succession the harder layers above until the capping is reached (Fig.
380). The conditions appear now to be such that the effective work is
largely concentrated, as it usually has been, near the middle of the
channel; and so the gorge recedes with a margin of the earlier river
bed remaining as a terrace on either side and extending to the former
river bank (Fig. 377).

As must have been noted, one peculiarity of the operation of the churn
drill beneath the cataract is that the depth of the gorge will bear
a direct proportion to its width, and if the volume of water has
varied during the process of recession, these changes in volume will be
registered in the width and also in the depth of that section of the
gorge which was drilled at the time—the cross section of the gorge at
any place is proportional to the volume of the water falling in the
cataract which produced it, modified, however, by the competency to
handle the joint blocks of definite size (Fig. 381).


FIG. 382.—Comparison of a sketch of the Canadian Fall made with the
aid of a camera lucida in 1827 with a photograph taken from the same
view point in 1895 (after Gilbert).]

=The present rate of recession.=—There are various sketches, more or
less accurate, made in the early part of the nineteenth century, and
from the later period there are daguerreotypes, photographs, and maps,
which refer especially to the Canadian Fall; and which, taken together,
render possible a comparison of the earlier with the later brinks. By
comparing the earliest with the recent, views it is seen at a glance
that the Falls are receding, and at a quite appreciable rate (Fig.
382). A careful comparison of the maps made in 1842, 1875, 1886, 1890,
and 1905 of the brink of the Canadian Fall (Fig. 383) indicates that
for the period covered the rate of recession has been about five feet
per year, and similar studies made of the American Fall show that it
has been receding at the rate of only three inches per year, or one
twentieth the rate of the recession of the Canadian Fall.


FIG. 383.—Map to show the recession of the brink of the Canadian Fall,
based upon maps of different dates (after Gilbert).]

=Future extinction of the American Fall.=—It is because of this many
times more rapid recession of the Canadian Fall that the Niagara
cataract, instead of lying athwart the gorge, enters it from its side.
The Canadian Fall is thus in reality swinging about the American, and
the time can already be roughly estimated when this more effective
drilling tool will have brought about a capture, so to speak, of the
American Fall through the cutting off of its water supply. It will then
be drained and left literally “high and dry”, an enduring witness to
the geological effect of an island in making an unequal division of the
waters for the work of two cataracts.

As already pointed out, the inefficiency of the American Fall as an
eroding agent is amply attested by the wall of blocks already appearing
above the water below it. The tourist who a thousand years hence pays
a visit to the Niagara cataract, provided the water flow is allowed to
remain as it has been, will find above this rampart of blocks a bare
cliff in part undermined, and surmounted by a nearly flat table surface
which is cut off from the existing cataract by a higher section of the
gorge (Fig. 384). It is quite likely that this table will furnish the
most satisfactory viewpoint of the future cataract of that date.


FIG. 384.—Comparison of the present with the future falls.]

=The captured Canadian Fall at Wintergreen Flats.=—What we have
predicted for the future of the present American Fall will be the
better understood from the study of a monument to earlier capture
made long before the upper section of the gorge had been cut or the
whirlpool had come into existence. The tables were then turned, for it
was a fall upon the Canadian side of the gorge that was captured by
one upon the American. The locality is known as Wintergreen Flats, or
sometimes as Fosters Flats; though the first name properly applies to
a higher surface near the brink of the gorge, and Fosters Flats to a
lower plain near the level of the river (see Fig. 381, p. 355). The
peculiar topographic features at this locality are well brought out in
Gilbert’s bird’s-eye view of the locality (Fig. 385); in fact, in some
respects better than they appear to the tourist upon the ground, for
the reason that the abandoned channel and the Flats on the site of the
since undermined island are both heavily forested and so not easy to
include in a single view. For one who has studied the existing cataract
this early monument is full of meaning. Standing, as one may, upon the
very brink of the former cataract, it is easy to call up in imagination
the grandeur of the earlier surroundings and to hear the thunder of the
falling water. A particularly vivid touch is added when, in digging
over the sand about the great blocks of fallen limestone underneath
the brink, one comes upon the shells of an animal still living in the
Niagara River, though only in the continual spray beneath the cataract.


FIG. 385.—Bird’s-eye view of the captured Canadian Fall at Wintergreen
Flats, showing the section of the river bed above the cliff and the
blocks of fallen Niagara limestone strewn over the abandoned channel
below (after Gilbert).]

=The Whirlpool Basin excavated from the St. Davids Gorge.=—It has
already been pointed out that a rock channel now filled with glacial
deposits extends from the Whirlpool Basin to the edge of the escarpment
at St. Davids (Fig. 389, p. 363). In plan this buried gorge has a
trumpet form, being more than two miles wide at its mouth and narrowing
to the width of the upper gorge before it has reached the Whirlpool.
Near the Whirlpool it has been in part excavated by Bowman Creek,
thus revealing walls that are well glaciated. Different opinions have
been expressed concerning the origin of this channel, one being that
it is the course either of a preglacial river or one incised between
consecutive glacial invasions; and another that it is a cataract gorge
drilled out between glacial invasions after the manner of the later
Niagara gorge. In either case its contours have been much modified by
the later glacier or glaciers, whose work of planing, polishing, and
widening is revealed in the exposed surfaces; and it is not improbable
that a cataract has receded along the course of an earlier river valley.

As we shall see, there are facts which point rather clearly to an
earlier cataract which ended its life immediately above the present
Whirlpool. When the later Niagara cataract had receded to near the
upper end of the Cove section, or near the present Whirlpool, the
falling water must have been separated from this older channel and its
filling of till deposits by only a thin wall of rock, and this must
have been constantly weakened as its thickness was further reduced.

When this weakened dam at last gave way, it must have produced a
debacle grand in the extreme. It is hardly to be conceived that the
“washout” of the ancient channel to form the Whirlpool Basin could
have occupied more than a small fraction of a day, though it is highly
probable that the broken rock partition below the Whirlpool was not
immediately removed entire. The mandible-like termination of the Eddy
Basin immediately above the Whirlpool has led Taylor to believe that
the cataract quickly reëstablished itself at this point upon the last
site of the extinct St. Davids cataract. If reduced in power for a
short interval, as a result of the obstructions still remaining in the
lately broken dam below the Whirlpool, the remarkable narrowing of the
gorge at this point would be sufficiently accounted for.

Being compelled to turn through more than a right angle after it
enters the Whirlpool Basin, the swift current of the Niagara River is
forced to double upon itself against the opposite bank and dive below
the incoming current before emerging into the Cove section below the
Whirlpool (Fig. 386).


FIG. 386.—Map of the Whirlpool Basin, showing the rock side walls
like those of the Niagara Gorge, and the drift bank which forms the
northwest wall (after Gilbert).]

In tearing out the loose deposits which had filled this part of
the buried St. Davids Gorge, many bowlders of great size were left
which slid down the slope and in time produced an armor about the
looser deposits beneath, so as to protect them and prevent continued
excavation. Thus it is found that the submerged northwestern wall
of the basin is sheathed with bowlders large enough to retain their
positions and so stop a natural process of placer outwashing upon a
gigantic scale (Fig. 386).

=The shaping of the Lewiston Escarpment.=—To understand the formation
of the Lewiston Escarpment cut in the hard Niagara limestone, it is
necessary to consider the geology of a much larger area—that of the
Great Lakes region as a whole. To the north of the Lakes in Canada
is found a most ancient continent which was in existence when all
the area to the southward lay below the waters of the ocean. In a
period still very many times as long ago as the events we have under
discussion, there were laid down off the shore of this oldland a series
of unconsolidated deposits which, hardened in the course of time, and
elevated, are now represented by the shales, sandstone, and limestone
which we find, one above the other, in the Niagara gorge in the order
in which they were laid down upon the ocean floor. The formations
represented in the gorge are but a part of the entire series, for
other higher members are represented by rocks about Lake Erie and even
farther to the southward. These strata, having been formed upon an
outward sloping sea floor, had a small initial dip to the southward,
and this has been probably increased by subsequent uptilt, including
the latest which we have so recently had under discussion. At the
present time the beds dip southward by an angle of less than four
degrees, or about thirty-five feet in each mile.


FIG. 387.—Map to show the cuestas which have played so important
a part in fixing the boundaries of the Lake basins, and also the
principal preglacial rivers by which they have been trenched (based
upon a map by Grabau).]

When the elevation of the land in the vicinity of this shore had
caused a recession of the waters, there was formed a coastal plain
on the borders of the oldland like that which is now found upon our
Atlantic border between the Appalachians and the sea (Fig. 272, p.
246). The rivers from the oldland cut their way in narrow trenches
across the newland, and because of the harder limestone formations,
their tributaries gradually became diverted from their earlier courses
until they entered the trunk stream nearly at right angles and produced
the type of drainage network which is called “trellis drainage.” It
is characteristic of this drainage that few tributaries of the second
order will flow up the natural slope of the beds, but on the contrary
these natural slopes are followed in the softer rock nearly at right
angles again to the tributaries of the first order of magnitude (Fig.
387). Thus are produced a series of more or less parallel escarpments
formed in the harder rock and having at their base a lowland which
rises gradually in the direction of the oldland until a new escarpment
is reached in the next lower of the hard formations. Such flat-topped
uplands in series with intermediate lowlands and separated by sharp
escarpments are known as _cuestas_ (see p. 246), and the Lewiston
Escarpment limits that formed in Niagara limestone (Figs. 387 and 388).


FIG. 388.—Bird’s-eye view of the cuestas south of Lakes Ontario and
Erie (after Gilbert).]

=Episodes of Niagara’s history and their correlation with those of
the Glacial Lakes.=—Of the early episodes of Niagara’s history, our
knowledge is not as perfect as we could desire, but the later events
are fully and trustworthily recorded. The birth of the Falls is to
be dated at the time when the ice front had here first retired into
what is now Canadian territory, thus for the first time allowing the
waters from the Erie basin to discharge over the Lewiston Escarpment
into the basin of the newly formed Lake Iroquois (Fig. 364, p. 334).
Since the level of Lake Iroquois was far above that of the present Lake
Ontario, the new-born cataract was not the equivalent in height of
the escarpment to-day. The Iroquois waters then bathed all the lower
portion of the escarpment, so that the foot of the Fall was upon the
borders of the Lake.

In order to interpret the history of the Niagara gorge, we must
remember that the effective drilling of this gorge was in each stage
dependent mainly upon the volume of water discharged from Lake Erie,
a large discharge being recorded by a channel drilled both wide
and deep, while that produced by the discharge of a smaller volume
was correspondingly narrow and shallow. To-day the gorges of large
cross section have, moreover, a relatively placid surface, whereas
through the constricted sections the water of the river is unable to
pass without first raising its level at the upper end and under the
head thus produced rushing through under an increased velocity. The
best illustration of such a constricted section is the Gorge of the
Whirlpool Rapids.


FIG. 389.—Sketch map of the greater portion of the Niagara Gorge to
show the changes in cross section in their relations to Niagara history
(based upon a map by Taylor).]

Our reading of the history should begin at the site of the present
cataract, since the records of later events are so much the more
complete and legible, and it should ever be our plan to proceed from
the clearly written pages to those half effaced and illegible.

As we have learned, the most abrupt change in the cross section of the
gorge is found a little above the railroad bridges, where the Upper
Great Gorge is joined to the Gorge of the Whirlpool Rapids (Fig. 389).
In view of the remarkably uniform cross section of the Upper Great
Gorge, there is no reason to doubt that it has been drilled throughout
under essentially the same volume of water, and that its lower limit
marks the position of the former cataract when the waters from the
upper lakes were transferred from the “North Bay Outlet” into the
present or “Port Huron Outlet” and Lake Erie. As the upper limit of the
Gorge of the Whirlpool Rapids thus corresponds to the closing of the
“North Bay Outlet” and the extinction of the Nipissing Great Lakes, so
its lower limit doubtless corresponds to the opening of that outlet and
the termination of the preceding Algonquin stage; for in the stage of
the Nipissing lakes the water of the upper lakes, as we have learned,
reached the ocean through the northern outlet.

Mr. Frank Taylor, who has given much study to the problem of Niagaran
history, believes that the Middle Great Gorge, comprising the Eddy
Basin and the Cove section, represents the gorge drilling which
occurred during the later stage of Lake Algonquin after the “Trent
Outlet” had been closed and the waters of the upper lakes had been
turned into the Erie Basin.

Summarizing, then, the episodes of the lake and the gorge history are
to be correlated as follows:—

        GLACIAL LAKE                             NIAGARA GORGE

 Early Lakes Iroquois and Algonquin.    Drilling of the gorge from the
                                        Lewiston Escarpment to the Cove
                                        section above the Wintergreen

 Later Lakes Iroquois and Algonquin     Drilling of Middle Great Gorge.
 with upper lakes discharging
 into Erie basin.

 Nipissing Great Lakes with the         Drilling of the narrow Gorge of
 upper lake waters diverted from        the Whirlpool Rapids.
 Lake Erie.

 Recent St. Lawrence drainage           Drilling of Upper Great Gorge to
 since the waters of the upper lakes    the present cataract.
 were discharged into Lake Erie
 through occupation of the Port
 Huron Outlet.

=Time measures of the Niagara clock.=—In primitive civilizations time
has sometimes been measured by the lapse necessary to accomplish a
certain task, such, for example, as walking the distance between two
points; and the natural clock of Niagara has been of this type. But men
possess differences in strength and speed, and the same man is at some
times more vigorous than at others, and so does not work at a uniform
rate. The cataract of Niagara, charged with the pent-up energy of the
waters of all the Great Lakes, can rush its work as it is clearly
unable to do at times when the greater part of this energy has been
diverted. Units of distance measured along the gorge are therefore too
unreliable for our use, with the unique exception of the stretch from
the railroad bridges to the site of the present cataract, within which
stretch the gorge cross sections are so nearly uniform as to indicate
an approximation to continued application of uniform energy. This
energy we may actually measure in the existing cataract, and so fix
upon a unit of time that can be translated into years.

In order to secure the normal rate of recession of this Upper Great
Gorge, we should add to the volume of water in the Canadian Fall that
now passing over the American; and for the reason that the blocks
which fall from the cataract cornice and are the tools of the drilling
instrument approximate to a definite size fixed by their joint planes,
the effect of this added energy it is not easy to estimate. We may be
sure, however, that the drilling action would be somewhat increased by
the junction of the two Falls, and thus are assured that the average
rate of recession within the Upper Great Gorge has been somewhat in
excess of the five feet per year determined by Gilbert for the present
Canadian Fall. The Upper Great Gorge is about two miles in length,
and its beginning may thus be dated near the dawning of the Christian
Era. The Whirlpool Gorge was cut when the ice vacated the North Bay
Outlet in Canada, and still lay as a broad mantle over all northeastern
Canada. For the earlier gorge and lake stages, the time estimates are
hardly more than guesses, and we need not now concern ourselves with

=The horologe of late glacial time in Scandinavia.=—A glacial
timepiece of somewhat different construction and of greater refinement
has been made use of in Scandinavia to derive the “geochronology
of the last 12,000 years.” Instead of retreating over the land and
impounding the drainage as it did so, the latest continental glacier
of Scandinavia ended below sea level, and as it retired, its great
subglacial river laid down a giant esker known as the Stockholm Os,
which was bordered by a delta and fringed on either side by water-laid
moraines of the block type. These recessional moraines are upon the
average less than 1000 feet apart, and are believed to have each been
formed in a single season. The delta deposits which surround the esker
are of thin-banded clay, and as an additional uppermost band is found
outside every moraine, these bands are also believed to represent each
the delta deposit of a single year. In studies extending over many
years, Baron de Geer, with the aid of a large body of student helpers,
has succeeded in completing a count of moraines and clay layers, and so
in determining the time to be 12,000 years since the ice front of the
latest continental glacier lay across southern Sweden. The fertility
of conception and the thoroughness of execution of this epoch-making
investigation recommend its conclusion to the scientific reader.


 G. K. GILBERT. Niagara Falls and their History, Nat. Geogr. Soc. Mon.,
 vol. 1, No. 7, 1895, pp. 203-236.

 F. B. TAYLOR. Origin of the Gorge of the Whirlpool Rapids at Niagara,
 Bull. Geol. Soc. Am., vol. 9, 1898, pp. 59-84.

 A. W. GRABAU. Guide to the Geology and Paleontology of Niagara Falls
 and Vicinity, Bull. N. Y. State Mus., vol. 9, No. 45, 1901, pp. 1-85,
 pls. 1-11.

 J. W. SPENCER. The Falls of Niagara, etc. Dept. of Mines, Geol. Surv.
 Branch, Canada, 1907, pp. 490, pls. 43.

 G. K. GILBERT. Rate of Recession of Niagara Falls, etc. Bull. 306, U.
 S. Geol. Surv., 1907, pp. 31, pls. 11.

 G. DE GEER. Quaternary Sea Bottoms of Western Sweden. Paper 23, Livret
 Guide Cong. Géol. Intern., 1910, pp. 57, pls. 3.



=Contrasted sculpturing of continental and mountain glaciers.=—In
discussing in a previous chapter the rock pavement lately uncovered by
the Greenland glacier, we learned that this surface had been lowered
by the processes of plucking and abrasion, the combined effect of
which is always to reduce the irregularities of the surface, soften
its outlines, and from sharply projecting masses to develop rounded
shoulders of rock—_roches moutonnées_.

Though the same processes act in much the same manner beneath
mountain glaciers, though here upon all parts of the bed, they are,
in the earlier stages at least, subordinated to a third process more
important than the two acting together. Sculpture by mountain glaciers,
instead of reducing surface irregularities and softening outlines,
increases the accent of the relief and produces the most sharply
rugged topography that is known. In nearly all places where Alpinists
resort for difficult rock climbing, mountain glaciers are to be seen,
or the evidence for their former presence may be read in unmistakable

=Wind distribution of the snow which falls in mountains.=—Until quite
recently students of glaciation have concerned themselves but little
with the work of the wind in lifting and redistributing the snow after
it has fallen. We have already seen that, for the continental glaciers,
wind appears to be the chief transporting agent, if we except the
marginal lobes where glacier flow assumes large importance. In the case
of mountain glaciers, also, we are to find that for the earlier stages
particularly wind is of the first importance as a redistributing agent.
In the higher levels snow is swept up from the ground by all high
winds, and does not find a resting place until it is dropped beneath an
eddy in some irregularity of the surface; and if the inherited surface
be relatively smooth, this will be found in most cases upon the lee of
the mountain crest.

In normal cases at least the inherited irregularities of the higher
zones of mountain upland are the gentle depressions which develop at
the heads of streams. These become, then, the sites of snowdrifts that
are augmented in size from year to year, though at first they melt away
in the late summer.

[Illustration: FIG. 390.—Snowdrift hollowing its bed by nivation and
building a delta (at the left). Quadrant Mountain, Yellowstone National

=The niches which form on snowdrift sites.=—Wherever a drift is
formed, a process is set in operation, the effect of which is to hollow
out and lower the ground beneath it, a process which has been called
_nivation_. The drift shown in Fig. 390 was photographed in late summer
at an elevation of some 9000 feet in the Yellowstone National Park.
The very gently sloping surface surrounding the drift is covered with
grass, but within a zone a few feet in width on the borders of the
drift no grass is growing, and in its place is found a fine brown soil
which is fast becoming the prey of the moving water derived by melting
of the drift. This is explained by the water permeating the crevices of
the rock and being rent by the nightly freezing. Farther from the drift
the ground is dry, and no such action is possible. With each succeeding
spring the augmented drift as it melts carries all finely comminuted
rock material down slopes beneath the snow to emerge at the lowest
margin and be there deposited in the form of a delta. By the operation
of this process of nivation the higher parts of the drift site are
lowered as deposition goes on upon the lower. The combined effect is
thus to produce a _niche_ or faintly etched amphitheater upon the slope
of the mountain (Fig. 391).

[Illustration: FIG. 391.—Amphitheater formed on a drift site in
northern Lapland (after a photograph by G. von Zahn).]

=The augmented snowdrift moves down the valley—birth of the
glacier.=—In still lower air temperatures the drifts enlarge with
each succeeding year until they endure throughout the summer season.
From this stage on, an increment of snow is left from each succeeding
season. No longer entirely wasted by melting, the time soon comes when
the upper snow layers will by their weight compress the lower into ice,
and the mass will begin to creep down the slope along the course of the
inherited valley. The enlarged snowdrift which feeds this ice stream is
called the _névé_ or _firn_.

Against the sloping cliff which had been shaped by nivation at the
upper margin of the snowdrift, that snow which is not of sufficient
depth to begin a movement towards the valley separates from the moving
portion, opening as it does so a cleft or crevasse parallel to the
wall. This crack in the snow is called by its German name _Bergschrund_
or _Randspalte_, and may perhaps be referred to as the marginal
crevasse (Fig. 392).


FIG. 392.—The marginal crevasse or Bergschrund on the highest margin
of a glacier (after Gilbert).]

=The excavation of the glacial amphitheater or cirque.=—It has been
found that the marginal crevasse plays a most important rôle in the
sculpture of mountains by glaciers, for the great amphitheater which
is everywhere the collecting basin for the nourishment of mountain
glaciers is not an inherited feature, but the handiwork of the ice
itself. This was the discovery of Mr. W. D. Johnson, an American
topographer and geologist, who, in order to solve the problem of the
amphitheater allowed himself to be lowered into such a crevasse upon
the Mount Lyell glacier of the Sierra Nevadas in California.


FIG. 393.—Niches and cirques in the same vicinity in the Bighorn
Mountains of Wyoming. _A, A_, unmodified valleys; _B, B_, niches on
drift sites; _C, C_, cirques on small glacier sites (after map by F. E.
Mathes, U. S. G. S.).]

Let down a distance of a hundred and fifty feet, he reached the bottom
of the crack, and in a drizzling rain of thaw water stood upon a floor
composed of rock masses in part dislodged from a wall which extended
some twenty feet upwards upon the cliff side of the crevasse. It was
evident that the warm air of the day produced the thaw water which was
constantly dripping and which filled every crack and cranny of the rock
surface. With the sinking of the sun below the peaks the sudden chill,
so characteristic of the end of the day in high mountains, causes this
water to freeze and thus rend the rock along its planes of jointing.
Broad and thin plates of ice, loosened by melting at the walls, could
be extracted from the crevices of the rock as mute witnesses to the
powerful stresses developed by this most vigorous of weathering


FIG. 394.—Subordinate small cirques in the amphitheater on the west
face of the Wannehorn above the Great Aletsch Glacier of Switzerland.]

In short, the rock wall above the glacier, which in its initial stage
was the upper wall of the niche hollowed beneath the snowdrift, is
first steepened and later continually both recessed and deepened by
an intensive frost rending which is in operation at the base of the
marginal crevasse. The same process does not go on as rapidly above the
surface of the névé for the reason that _the necessary wetting of the
rock surface does not there so generally result from the daily summer
thaw_. At the bottom of the marginal crevasse alone is this condition
fully realized. Intensive frost action _where the rock is wet with thaw
water daily_ is thus a fundamental cause, both of the hollowing of
the early drift site to form the niche, and of the later enlargement
of this niche into an amphitheater or cirque when the drift has been
transformed into the névé of a glacier. Inasmuch as the crevasse forms
where the snow and ice pull away from the rock toward the middle of the
depression, the cirque wall in its early stage has the outline of a
semicircle. In the Bighorn Mountains of Wyoming, all stages, from the
unmodified valley heads to the full-formed cirque, may be seen near
one another (Fig. 393). It will be noted that wherever a glacier has
formed, as indicated by the cirque, there is a series of lakes which
have developed in the valley below (see p. 412).

[Illustration: FIG. 395.—“Biscuit cutting” effect of glacial sculpture
in the Uinta Mountains of Wyoming (after Atwood).]


FIG. 396.—Two intersecting inverted cones representing glacial cirques
of different sizes, to show that their intersection is the arc of a
hyperbola, the curve to which the col approximates.]

=Life history of the cirque.=—In its earliest stage the cirque is more
or less uniformly supplied with snow from all sides, and so it enlarges
by recession in a manner to retain its early semicircular outline. In
a later stage a larger proportion of the snow reaches the cirque at
its sides so that its further enlargement causes it to broaden and to
flatten somewhat that part of its outline which represents the head of
the valley (Fig. 389, p. 364). As the territory of the upland is still
further invested by the cirques, their nourishment becomes still more
irregular, and the circular outline gives place to a scalloped border,
as the amphitheater becomes differentiated into subordinate smaller
cirques, each of which corresponds to a scallop of the outline (Fig.
398 and Fig. 394).

=Grooved and fretted uplands.=—The partial investment by cirques of a
mountain upland yields a type of topography quite unlike that produced
by any other geological process. The irregularly connected remnants
of the inherited upland resemble nothing so much as a layer of dough
from which biscuits have been cut (Fig. 395). The surface as a whole,
furrowed as it is below the cirques,
may be described as a _grooved upland_ (plate 19 A). A further
continuation of the process removes all traces of the earlier upland,
for the cirques intersect from opposite sides and thus yield palisades
of sharp rock pinnacles which rise on precipitous walls from a terraced
floor. This ultimate product of cirque sculpture by glaciers is called
a _fretted upland_ (plate 18 A and 19 B).

│   PLATE 18.                                                         │
│                                                                     │
│ [Illustration: _A._ Fretted upland of the Alps seen from the summit │
│ of Mount Blanc.]                                                    │
│                                                                     │
│ [Illustration: _B._ Model of the Malaspina Glacier and the fretted  │
│ upland above it (after model by L. Martin).                         │

│   PLATE 19.                                                      │
│                                                                  │
│ [Illustration: _A._ Contour map of a grooved upland, Bighorn     │
│ Mountains, Wyoming (U. S. Geol. Survey).]                        │
│                                                                  │
│ [Illustration: _B._ Contour map of a fretted upland, Philipsburg │
│ Quadrangle, Montana (U. S. Geol. Survey).]                       │

=The features carved above the glacier.=—The ranges of pinnacles
carved out by mountain glaciers have become known by various names of
foreign derivation, such as _arête_, _grat_, _aiguille_ mountains,
“files of _gendarmes_”, etc. They may, perhaps, be best referred to as
_comb ridges_, and according to their position they are differentiated
into main and lateral comb ridges, as will be clear from the second map
of plate 19.

[Illustration: FIG. 397.—A col shaped like a hyperbola between Mount
Sir Donald and Yogo Peak in the Selkirks (after a plate by the Keystone
Plate Co.).]

With the gradual invasion of the upland upon which the cirques have
made their attack, the area from which winds may gather up the snow
is steadily diminished, and hence cirque recession is correspondingly
retarded. Cirques which have approached each other from opposite sides
of the ridge until they have become tangent at one point may, however,
still receive nourishment at the sides and so continue to cut down the
intervening rock wall to form a pass or _col_. The theoretical curve
which results from this intersection is that known as the hyperbola, of
which an illustration is afforded by Fig. 396. An approximation to this
form is clearly furnished by most of the mountain passes in glaciated
mountain districts, and a particularly good illustration is furnished
from the vicinity of Glacier on the line of the Canadian Pacific
Railway (Fig. 397).


FIG. 398.—Diagrams to illustrate the progressive investment of an
upland by cirques with the formation of comb ridges, cols, and horns.
I, early stage, youth; II, intermediate stage; III, late stage,

Upon either side of the col the land mass is left in high relief,
rising from a more or less triangular base (Fig. 398, III) into a sharp
horn or tooth. An illustration of such a _horn_ is furnished by the
Matterhorn in the Swiss Alps, or by Mount Sir Donald in the Selkirks,
though less noteworthy examples may be found in every maturely
glaciated mountain district.

=The features shaped beneath the glacier.=—Those features which are
carved above the glacier—the comb ridge, the col, and the horn—are
all shaped as a result of intensive weathering upon the cirque wall.
The shaping at lower levels is accomplished by processes in operation
below the glacier surface, where weathering is excluded and where
plucking and abrasion work together to tear away and grind off the rock
surface. By their joint action the valley is both deepened and widened,
directly to the height of the glacier surface, and indirectly through
undermining as far up as rock extends. Thus the valley is transformed
into one of broad and flat bed and precipitous side walls—the U-shaped
section illustrated by valleys of the Swiss Alps and in fact in all
districts which have been strongly glaciated by mountain glaciers (Fig.


FIG. 399.—The U-shaped Kern valley in the Sierra Nevadas of California
(after W. B. Scott).]

As high up in the valley as it has been occupied by the glacier,
the bed is rounded, smoothed, and polished, and marked by the
characteristic glacial scorings or striæ which point down the valley.
Above the level of the glacier’s upper surface, on the other hand,
erosion is accomplished through undermining or sapping, a process which
always leaves precipitous slopes of ragged surface made up of the joint
planes on which the fallen blocks have separated from the cliff. Thus
there is found a sharp line which separates the smoothly rounded rock
surface below from the jagged and precipitous one above (Fig. 400).
Inasmuch as this boundary usually separates the scalable from the
inaccessible slopes above, snow is apt to lodge at this level and make
it strikingly apparent.


FIG. 400.—Glaciated valley wall in the Sierra Nevadas of California,
showing the sharp line which separates the abraded from the undermined
rock surface (after a photograph by Fairbanks).]


FIG. 401.—View of the Vale of Chamonix from the séracs of the Glacier
des Bossons. The alb of the opposite side is well brought out.]

If uplift of the land occurs while glaciers occupy the valleys of
mountains, an increased capacity for deepening the valley is imparted
to these ice streams, and we find, as a result, a deep central valley
of U cross section excavated within a relatively broad trough visible
above the shoulder on either side of the later furrow. Save only for
its characteristic curves, such a valley bears close resemblance to
a mature stream valley which has been rejuvenated (see p. 173). The
remnants of the earlier glacier-carved valley are, as already stated,
gently curving high terraces so common in Switzerland, where they
are known as _albs_ or high mountain meadows. These albs may be seen
to special advantage on the sides of the Chamonix valley (Fig. 401),
the Lauterbrunnen valley, or in fact almost any of the larger Alpine

=The cascade stairway in glacier-carved valleys.=—If now, instead of
giving our attention to the cross section, we follow the course of the
valley that has been occupied by a glacier, we find that it descends
by a series of steps or terraces having many backwardly directed
treads (plate 19), whereas a normal and well-established river valley
has only forward grades. Because of these backward grades the stream
waters are impounded, and so lakes are found strung along the valley
in chains as the larger beads are found in a rosary, and these are the
characteristic _rock basin lakes_ sometimes referred to as “Paternoster
Lakes” (see p. 412 and Fig. 402).

│   PLATE 20.                                                       │
│                                                                   │
│ [Illustration: Map of the surface modeled by mountain glaciers in │
│ the Sierra Nevadas of California (after I. C. Russell).]          │

When the backward grades upon the valley floor are especially steep,
the rock step becomes a _rock bar_, or _Riegel_, of which nearly every
Alpine valley has its examples. In a walk from the Grimsel to Meiringen
many such bars are passed. Carrying in suspension the sharp rock sand
from the glacier deposits along its bed, the stream which succeeds
to the glacier as it vacates its valley saws its way through these
obstructions with a rapidity that is amazing, thus producing narrow
defiles, of which the Gorge of the Aar near Meiringen and that of the
Gorner near Zermatt are such well-known examples (Fig. 403).


FIG. 402.—Map of an area near the continental divide in Colorado,
showing an unglaciated surface to the west of the divide, where the
westerly winds have cleared the ground of snow, and the glacier-carved
country to the eastward. Note the regular forms of the youthful cirque,
the glacier stairway, and the rock basin lakes (U. S. G. S.).]


FIG. 403.—Gorge of the Albula River near Berkum in the Engadine, cut
through a rock bar by the river which has succeeded to the earlier


FIG. 404.—Idealistic sketch showing both glaciated and nonglaciated
side valleys tributary to a glaciated main valley (after Davis).]

It is characteristic of rivers that the tributaries cut their valleys
more rapidly than does the main stream within the neighboring section,
though they cannot cut lower than their outlets—the side streams enter
_accordantly_. This is easily explained because the grades of the
tributary streams are the steeper, and, as we well know, the corrosion
of a valley is augmented at a most amazing rate for each increase of
its grade. No such law controls the processes of plucking and abrasion
by which the glacier lowers its floor, for these processes appear to
depend for their efficiency upon the depth of the ice, and the supply
of cutting tools, quite as much as upon the grade of the bed. To apply
a homely illustration, the hollowing of flagstones upon our walks is
dependent more upon the number of persons that pass over them, and upon
their size and the number of protruding nails in their boot heels, than
upon the grades upon which they are placed. At all events we find that
the main glacier valleys are cut deeper than the side valleys, so that
the latter become _hanging valleys_—they enter the main valley, not
upon its bed, but some distance above it (Fig. 404).

The U-shaped hanging valleys, like the main valley, are much too large
for the streams which now fill them, and these diminutive side streams
plunge over the steep wall of the main valley in ribbon-like falls so
thin that the wind turns them aside and disperses the water in the
spray of a “bridal veil.” Such falls are found by the hundred in every
glaciated mountain district, imparting to it one of the greatest of its
scenic charms.

[Illustration: FIG. 405.—Character profiles in landscapes sculptured
by mountain glaciers.]

[Illustration: FIG. 406.—Flat dome shaped under the margin of
a Norwegian ice cap with projecting rock knobs and moraines in

=The character profiles which result from sculpture by mountain
glaciers.=—The lines which are repeated in landscapes carved by
mountain glaciers are easy to recognize (Fig. 405). The highest
horizon lines are the outlines of horns which are separated by cols.
Minaret-like palisades, or “files of _gendarmes_”, often run for long
distances as the characteristic comb ridges. Lower down and lacking
the lighter background of the sky, we make out with less distinctness
the U-valley, either with or without the albs to show that the
sculpturing process has been interrupted by uplift.

[Illustration: FIG. 407.—Two views illustrating successive stages in
the shaping of tinds or “beehive” mountains.]

=The sculpture accomplished by ice caps.=—In the case of ice caps,
the only rock exposed is found in the neighborhood of the margin—the
projecting islands known as nunataks. It is essential for the existence
of the ice cap that the rock base should have relatively slight
irregularities compared to the dimensions of the cap itself. Except
in very high latitudes this base must be somewhat elevated, for like
mountain glaciers ice caps are nourished by the surface air currents,
and their snows are deposited above the snow line.

=The Norwegian tind or beehive mountain.=—Within temperate or tropical
climes the snow line lies so high that only the loftier mountains
are able to support glaciers. It follows that those which are formed
flow upon relatively high grades with correspondingly high rate of
movement and increased cutting power. Within high latitudes the snow
is found nearer the sea level, and glaciers are for the most part
correspondingly sluggish in their movements as well as less active
denuding agents.

To this condition characteristic of high latitude glaciers, there is
added in Norway another in the peculiar shape of the basement beneath
the recent and the still existing glaciers. The plateau of Norway is
intersected by a network of deep and steep walled fjords, and the
glaciers have developed as small ice caps perched upon veritable
pedestals of rock, over the margins of which their outlet tongues of
ice descend on steep slopes into the fjord. The tops of the pedestals
thus come to be shaped by the plucking and abrading processes into
flat domes (Fig. 406), while the knobs of rock, which as nunataks
reach above the surface of the ice, divide the outflowing ice tongues
at the margin of the pedestal. These tongues being much more active
denuding agents, because of their steep gradients, continually lower
their beds, thus transforming the earlier knobs of rock into high and
steep mountains of more or less circular base. Such “beehive” mountains
upon the margins of the fjords are the characteristic Norwegian _tinds_
(Fig. 407).


 I. C. RUSSELL. Quaternary History of Mono Valley, California, 8th Ann.
 Rept. U. S. Geol. Surv., 1889, pp. 329-371, pls. 27-37.

 F. E. MATTHES. Glacial Sculpture of the Bighorn Mountains, Wyoming,
 21st Ann. Rept. U. S. Geol. Surv., 1900, Pt. ii, pp. 179-185, pl. 23.

 W. D. JOHNSON. Maturity in Alpine Glacial Erosion, Jour. Geol., vol.
 12, 1904, pp. 569-578.

 G. K. GILBERT. Systematic Asymmetry of Crest Lines in the High Sierras
 of California, _ibid._, pp. 579-588.

 EMM. DE MARTONNE. Sur la Formation des Cirques, Ann. de Géogr., vol.
 10, 1901, pp. 10-16.

 W. M. DAVIS. Glacial Erosion in North Wales, Quart. Jour. Geol. Soc.
 Lond., vol. 65, 1909, pp. 281-350, pl. 14.

 ED. BRÜCKNER. Die Glazialen Züge im Antlitz der Alpen, Naturw.
 Wochenschr., N. F., vol. 8, 1909.

 WILLIAM H. HOBBS. Characteristics of Existing Glaciers, pp. 1-96.



=Transition from the ice cap to the mountain glacier.=—A study of
existing glaciers leads inevitably to the conclusion that although
subject to short period advances and retreats, yet, broadly speaking,
glaciers are now gradually wasting away, surrounded by wide areas upon
which are the evidences of their recent occupation. We are thus living
in a receding hemicycle of glaciation.

[Illustration: FIG. 408.—Schematic diagram to show the relationships
of glacier types formed in succession during a receding hemicycle of

Many mountain districts which now support small glaciers only, or
none at all, were once nearly or quite submerged beneath snow and
ice. If once covered by an ice carapace or cap, our present interest
in them begins at that stage of the receding hemicycle _when the rock
surface has made its reappearance above the surface of the snow-ice
mass_. At this stage intensive frostwork, the characteristic high
level weathering, begins, and cirques develop above the scars of those
earlier amphitheaters formed in the advancing hemicycle.

=The piedmont glacier.=—In this early stage of transition from the
ice cap to the mountain glacier, the ice flows outward to the mountain
front in ill-defined streams divided by the projecting ridges, and
upon reaching the mountain front these streams deploy upon it so as
to coalesce in a great stagnant ice apron whose upper surface slopes
gently forward at an angle of a few degrees at the most (Fig. 408,
stage I). This is the _piedmont glacier_, a type found to-day in the
high latitudes of Alaska and in the southern Andes (Fig. 409 and pl. 18

[Illustration: FIG. 409.—Map of the Malaspina glacier of Alaska, the
best known of existing piedmont glaciers (after Russell).]

During this stage the cirques may be but poorly defined, and ice flows
in both directions from rock divides so that the streams transect the
range, and later, after the glaciers have disappeared, may expose a
pass smoothed and polished upon its floor and with striæ directed in
opposite directions from the highest point. The pass of the Grimsel
in Switzerland furnishes an excellent illustration of such earlier
transection of the range.


FIG. 410.—Map of the Baltoro glacier of the Himalayas, a typical
glacier of the dendritic type.]

=The expanded-foot glacier.=—As air temperatures continue to become
milder, the glacier streams within the mountains are less deep and
hence more clearly defined, and instead of coalescing upon the mountain
foreland, they now issue from the mountains to form individual aprons
and are described as _expanded-foot glaciers_ (Fig. 408, stage II, and
Fig. 292, p. 264).


FIG. 411.—The Triest glacier, a hanging glacieret separated from the
Great Aletsch glacier to which it was lately a tributary.]

=The dendritic glacier.=—Still later in the hemicycle nourishment of
the glaciers is diminished as depletion from melting increases, so that
the glacier streams no longer reach to the mountain front. Branches
continue to enter the main valley from the several side valleys like
the short branches of a tall tree, and because of this arrangement such
a glacier may be described as a _dendritic glacier_ (Fig. 408, stage
III, and Fig. 410).

Inasmuch as the depletion from melting increases at a rapid rate in
descending to lower levels, the tributary glacier valleys “hanging”
above the main valley in the lower stretches become separated, and may
continue to exist as series of hanging glacierets upon either side
of the main valley below the glacier front (Fig. 408, stage III, and
Fig. 411). It must be clear from this that any attempt to name each
separated ice stream without regard to its relationship must lead to
endless confusion, for glacier size is in such sensitive adjustment
to air temperature that a fall or rise of a few degrees only in the
average annual temperature of the district may prove sufficient to fuse
many glaciers into one or separate one ice mass into many smaller ones.

When in high latitudes a dendritic glacier descends in fjords to
below the level of the sea, it is attacked by the water in the same
manner as are the outlets of Greenland glaciers, and is then known as
a “tidewater glacier”, which may thus be a subtype or variety of the
dendritic glacier (Fig. 412).


FIG. 412.—The Harriman fjord glacier of Alaska, a tidewater variety of
dendritic glacier (after a map by Gannett).]

=The radiating (Alpine) glacier.=—In the progressive wastings of
dendritic glaciers, there comes a time when their dendritic outlines
give place to radiating ones. Attention has already been called to the
division of the cirque into subordinate basins separated by small rock
arêtes and yielding a markedly scalloped border (Fig. 394, p. 371).
When the ice front retires from the main valley into one of these
mature cirques, the now wasted ice stream is broken up into subordinate
glacierets, each of which occupies one of the basins within the larger
cirque, and these ice streams flow together to produce a glacier whose
component elements radiate like the sticks within a lady’s fan (Fig.
408, stage IV, and Fig. 413).


FIG. 413.—Map of the Rotmoos glacier, a radiating glacier of
Switzerland (after Sonklar).]

=The horseshoe glacier.=—As the glacier draws near to its final
extinction, it is crowded hard against the wall of the amphitheater in
which it has so long been nourished. Up to this stage it has offered
a swelling front outwardly convex as a direct consequence of the laws
controlling its flow. No longer amply nourished, for the first time
its front is hollowed, and it awaits its final dissolution curled up
against the cirque wall (Fig. 408, stage V, and Fig. 414). Practically
all the glaciers of the United States and southern Canada are of this

The above classification is one depending directly upon glacier
nourishment, and hence also upon size, and upon the stage of the
glacial hemicycle. In order to determine the type of any glacier it is
necessary to know the outlines of the mountain valley—its divide—and
those of the glacier or glaciers within it. It is likely that the
types of the advancing hemicycle of glaciation would be much the same,
save only for the _new-born_ or _nivation glacier_, which would be as
different as possible from the horseshoe type, to which in size it
corresponds. Upon the continent of Antarctica, where the absence of any
general melting of the ice, even in the summer season and near the sea
level, introduces special conditions, some additional glacier types are
found, which, however, it is not necessary that we consider here.

[Illustration: FIG. 414.—Outline map of the Asulkan glacier in the
Selkirks, a typical horseshoe glacier.]

=The inherited-basin glacier.=—It may be, however, that glaciers have
developed, not upon mountains shaped in a cycle of river erosion, nor
yet in succession to an ice cap, as in the normal cases which we have
considered. On the contrary, glaciers may develop where basins of
one sort or another have been inherited from the preceding period.
In such cases inherited depressions may become more important than
the auto-sculpture of the glacier. Glaciers which develop under such
conditions may be described as _inherited-basin glaciers_.

[Illustration: FIG. 415.—Outline map of the Illecillewaet glacier, an
inherited-basin glacier in the Selkirks.]

A partly closed basin between ridges may supply a collecting ground for
snows carried from neighboring slopes by the wind, and so may yield a
broad névé, approaching in size a small ice cap, yet without developing
definite ice streams except upon its border. Such a glacier is the
Illecillewaet glacier of the Selkirks (Fig. 415).

Again in low latitudes the high and pointed volcanic peaks may push
up beyond the snow line into the upper atmosphere, and so become
snow-capped. Definite cirques do not develop well under these
circumstances, and the loose materials of which such peaks are always
composed are attacked in somewhat irregular fashion from the different
sides. This is the case of Mount Rainier and similar peaks of the
Cascade range of North America.

=Summary of types of mountain glacier.=—In tabular form the various
types of mountain glacier may be arranged as follows:—


 _Piedmont glacier._ Mountain valleys entirely occupied and largely
 submerged, with overflow upon the foreland to form a common ice apron
 through coalescence of neighboring streams.

 _Expanded-foot glacier._ Valley entirely occupied and an overflow upon
 the foreland sufficient to produce individual ice apron.

 _Dendritic glacier._ Valley not completely occupied but with tributary
 ice streams ranged along the sides of the main stream, and with
 hanging glacierets separated near the glacier foot.

 _Radiating glacier._ Glacier largely included in a cirque with
 subordinate glacierets converging below like the sticks in a lady’s

 _Horseshoe glacier._ Small glacier remnants hugging the cirque wall
 and having an incurving front.

 _Inherited-basin glacier._ Of form dependent on a basin inherited and
 not shaped by the glacier itself.


 WILLIAM H. HOBBS. The Cycle of Mountain Glaciation, Geogr. Jour., vol.
 37, 1910, pp. 268-284.



=The glacier flow.=—The downward flow of the ice within a mountain
glacier has been the subject of many investigations and the topic of
many heated discussions since the time when Louis Agassiz and his
companions set a line of stakes across the Aar glacier and numbered the
surface bowlders in preparation for repeated observations. Their first
observation was that the line of stakes, which had run straight across
the glacier, was distorted into a curve which was convex downstream
(Fig. 416, A´), thus showing that the surface layers have more rapid
motion in proportion as they are distant from the side margins.
Summarizing these and later studies, it may be stated that the glacier
increases its rate of motion from its side margin towards its center
line, from its bed upwards towards its surface, and below the névé the
velocity is greatest where the fall is greatest and also wherever the
cross section diminishes. In all these particulars, then, the ice of
the glacier behaves like a stream of water. The average rate of flow of
Alpine glaciers varies from a few inches to a few feet per day, and is
greater during the warm summer season. The Muir glacier of Alaska has
been shown to move at the rate of about seven feet per day.


FIG. 416.—Diagram to illustrate the migrations of lines of stakes
crossing a glacier, due to its surface movement, _A_, original position
of lines; _A´_, later positions; _a_ and _a´_, original and distorted
forms of a square section of the glacier surface near its margin; _r_,
_r´_, diagonal crevasses.]

In traveling from the névé downward to the glacier foot, the snow not
only changes into ice, but it undergoes a granulating process with
continued increase in the size of the nodules until at the foot of
the glacier these may be picked out of the partially melted ice as
articulating balls the size of the fist or larger. Glacier ice has
therefore a structure quite different from that of lake ice, since the
latter is developed in parallel needles perpendicular to the freezing

=Crevasses and séracs.=—Prominent surface indications of glacier
movement are found in the open cracks or _crevasses_, which are the
marks of its yielding to tensional stresses. Crevasses are apt to run
either directly across the glacier, wherever there is a steep descent
upon its bed, or diagonally, running in from the margin and directed
up-glacier (_r_, _r_, _r_, of Fig. 416), though they occasionally run
longitudinally with the glacier when there is a rock terrace at the
side of the valley beneath the ice. The diagonal crevasses at the
glacier margin are due to the more sluggish movement where the ice is
held back by friction upon the walls of the valley, as will be clear
from Fig. 416. The square _a_ has by this movement been distorted into
the lozenge _a´_, so that the line _xy_ has been extended into _x´y´_,
with the obvious tendency to open cracks in the direction _ss_.

Every glacier surface below its névé is marked by steps or terraces,
which are well understood to overlie corresponding steps of the cascade
stairway to be seen in all vacated glacier valleys (plate 19). The
steep risers of these steps are usually marked by parallel crevasses
which cross the glacier. Under the rays of the sun, which strike them
more from one side than from the other, the slices into which the ice
is divided are transformed into sharpened blades and needles which are
known as _séracs_ (Fig. 401, p. 376, and Fig. 417).


FIG. 417.—Transverse crevasses at the fall below a glacier step
transformed by unsymmetrical melting into séracs.]

The numerous crevasses tell us that the ice is many times wrenched
apart during its journey down the glacier. This has been illustrated by
somewhat grewsome incidents connected with accidents to Alpinists, but
as they illustrate in some measure both the mode and the rate of motion
of Swiss glaciers, they are worthy of our consideration.

=Bodies given up by the Glacier _des Bossons_.=—In the year 1820,
during one of the earlier ascents of Mont Blanc, three guides were
buried beneath an avalanche near the _Rochers Rouges_ in the névé
of the Glacier des Bossons (Fig. 418). In 1858 Dr. Forbes, who had
measured the rate of flow of a number of Alpine glaciers, predicted
that the bodies of the victims of this accident would be given up by
the glacier after being entombed from thirty-five to forty years. In
the year 1861, or forty-one years after the disaster, the heads of
the three guides, separated from their bodies, with some hands and
fragments of clothing, appeared at the foot of the Glacier des Bossons,
and in such a state of preservation that they were easily recognized
by a guide who had known them in life. Inasmuch as these fragments of
the bodies had required forty-one years to travel in the ice the three
thousand meters which separate the place of the accident from the foot
of the glacier, the rate of movement was twenty centimeters, or eight
inches, per day.


FIG. 418.—View of the _Glacier des Bossons_ upon the slopes of Mont
Blanc showing the position of accidents to Alpinists and the place of
reappearance of their bodies.]

[Illustration: FIG. 419.—Lines of flow upon the surface of the
Hintereisferner glacier in the Alps (after Hess).]

Various separated parts of the body of Captain Arkwright, who had been
lost in 1866 upon the névé of the same glacier, reappeared at its foot
after entombment in the ice for a period of thirty-one years. To-day
the time of reappearance of portions of the bodies of persons lost
upon Mont Blanc is rather accurately predicted, so that friends repair
to Chamonix to await the giving up of its victims by the Glacier des


FIG. 420.—Lateral and medial moraines of the _Mer de glace_ and its
tributary ice streams.]

=The moraines.=—The horns and comb ridges which rise above the glacier
surface are continually subject to frost weathering, and from time to
time drop their separated fragments upon the glacier. Falling as these
do from considerable heights, they reach the ice under a high velocity,
and rebounding, sometimes travel well out upon its surface before
coming to a temporary rest. Upon a fresh snow surface of the névé
their tracks may sometimes be followed with the eye for considerable
distances, and their fall is a constant menace to Alpine climbers.
Below the névé the larger number of such fragments remain near the
cliff, and the lines of flow of the ice within the glacier surface are
such that blocks which reach points farther out upon the glacier are
later gathered in beneath the cliff at the side (Fig. 419). The ridge
of angular rock débris which thus forms at the side of the glacier is
called a _lateral moraine_ (see Fig. 411, p. 385, and Fig. 420).

At the junction of two glacier streams, the lateral moraines are
joined, and there move out upon the ice surface of the resultant
glacier as a _medial moraine_. Thus from the number of medial moraines
upon a glacier surface it is possible to say that the important
tributary glaciers number one more (Fig. 420).

[Illustration: FIG. 421.—Ideal cross-section of a mountain glacier to
show the position of moraines and other peculiarities characteristic of
the surface of the bed.]

The plucking and abrading processes in operation beneath the glacier,
quarry the rock upon its bed, and after shaping and smoothing the
separated rock fragments, these are incorporated within the lower
layers of the ice as _englacial_ rock débris. In spaces favorable for
its accumulation, a portion of this material, together with much finer
débris and rock flour, is left behind as a ground moraine upon the bed
of the glacier (see Fig. 421).

[Illustration: FIG. 422.—Fragments of rock of different sizes, to
bring out their different effects upon the melting of the glacier

At the foot of the glacier the relatively angular rock débris, which
has been carried upon the surface, and the soled and polished englacial
material from near the bottom, are alike deposited in a common marginal
ridge known as the _terminal_ or _end moraine_ (plate 21 B).

│   PLATE 21.                                                          │
│                                                                      │
│ [Illustration: _A._ View of the Harvard Glacier, Alaska, showing the │
│ characteristic terraces (after U. S. Grant).]                        │
│                                                                      │
│ [Illustration: _B._ The terminal moraine at the foot of a mountain   │
│ glacier (after George Kinney).]                                      │

=Selective melting upon the glacier surface.=—The white surface of the
glacier generally reflects a large proportion of the sun’s rays which
reach it, and its more rapid melting is largely accomplished through
the agency of rock fragments spread upon its surface. Such fragments,
however, promote or retard the melting process in inverse proportion to
their size up to a certain limit, and above that size their action is
always to protect the glacier from the sun. This nice adjustment to the
size of the rock fragments will be clear from examination of Fig. 422,
for rock is a poor conductor of heat, and in even the longest summer
day a thin outer layer only is appreciably warmed. Large rock blocks,
grouped in the medial and lateral moraines, hold back the process of
lowering the glacier surface during the summer, so that late in the
season these moraines stand fifty feet or more above the glacier as
armored ice ridges.


FIG. 423.—Small glacier table upon the surface of the Great Aletsch
glacier in 1908.]

Isolated and large rock slabs, as the season advances, may come to form
the capping of an ice pedestal which they overhang and are known as
_glacier tables_ (Fig. 423). Such tables the sun attacks more upon one
side than upon the other, so that the slab inclines more and more to
the south and may eventually slip down until its edges rest against the
glacier surface. Rounded bowlders, which less frequently become perched
upon ice pedestals, may, from a similar process, slide down upon the
southern side and leave a pyramid of ice furrowed upon this side and
known as an _ice pyramid_.

Fine dirt when scattered over the glacier surface is, on the other
hand, most effective in lowering its level by melting. Use was made
of this knowledge to lower the great drifts of snow which had to be
removed each season during the construction of the new Bergen railway
of southern Norway. Each dirt particle, being warmed throughout by
the sun’s rays, melts its way rapidly into the glacier surface until
the _dust well_ which it has formed is so deep that the slanting
rays of the sun no longer reach it. When the dirt particles are near
together, the thin walls which separate the dust wells are attacked
from the sides in the warm air of summer days, thus producing from a
patch of dirt upon the glacier surface a _bath tub_ (Fig. 424 _d_). At
night the water which fills these basins is frozen to form a lining
of ice needles projecting inward from the wall, and this, repeated in
succeeding nights, may entirely close the basin with water ice and
produce the familiar _glacier star_ (Fig. 424 _c_).


FIG. 424.—Effects of differential melting and subsequent refreezing
upon the glacier surface. _a_, dust wells; _b_, glacier _tub_ produced
by melting about a group of scattered dust particles; _c_, glacier star
produced when the inclosed water of the glacier well has frozen in
successive nights; _d_, “bath tub.”]

If the dirt upon the glacier surface, instead of being scattered, is
so disposed as to make a patch completely covering the ice to the
thickness of an inch or more, the effect is altogether different.
Protecting as it now does the ice below, a local ice hillock rises upon
its site as the surrounding surface is lowered, and as this grows in
height its declivities increase and a portion of the dirt slides down
the side. The final product of this shaping is an almost perfectly
conical ice hill encased in dirt and known as a _débris_, _sand_, or
_dirt cone_ (Fig. 425). The novice in glacier study is apt to assume
that these black cones contain only dirt, but is rudely awakened to the
reality when he attempts to kick them to pieces. Both glacier tubs and
débris cones may assume large dimensions; as, for example, in Alaska,
where they may be properly described as lakes and hills.

[Illustration: FIG. 425.—Dirt cone and one with its casing in part
removed. Victoria glacier (after Sherzer).]

A patch of hard and dense snow which is less easily melted than that
upon which it rests may lead to the formation of snow cones upon
the glacier surface similar in size and shape to the better known
débris cones. Such cones of snow have, with doubtful propriety, been
designated “penitents”, for it is pretty clear that the interesting
bowed snow figures, which really resemble penitents and which were
first described from the southern Andes under the name of _nieves
penitentes_, are of somewhat different character.


FIG. 426.—Schematic diagram to show the manner of formation of glacier

One further ice feature shaped by differential melting around rock
particles remains to be mentioned. Wherever the seasonal snowfalls
of the névé are exposed in crevasses, they are generally found to be
separated by layers of dirt, and lines of pebbles similarly separate
those ice layers which are revealed at the foot of the glacier. In
either case, if the sun’s rays can reach these layers in an opened
crevasse, the half-buried rock fragments are warmed by the sun upon
their exposed surfaces and slowly melt their way down the ice surface,
thus removing from it a thin layer of snow or ice and causing that part
above the pebble layer to project like a cornice. This process will go
on until the overhanging cornice protects the pebbles from any further
warming by the sun, but each lower pebble layer that is reached by the
sun will produce an additional cornice, so that the original surface
may at the bottom have been retired by the process a number of inches.
These features are described as _glacier cornices_ (Fig. 426).

[Illustration: FIG. 427.—Superglacial stream upon the Great Aletsch

=Glacier drainage.=—Already in the early morning of every warm summer
day, active melting has begun upon the surface of the Swiss glaciers.
Rills of icy water soon make their way along depressions upon the
surface, and are joined to one another so that they sometimes form
brooks of considerable size (Fig. 427). Such streams continue their
serpentine courses until these are intersected by a crevasse down which
the waters plunge in a whirling vortex which soon develops a vertical
shaft of circular section within the ice. Such shafts with their
descending columns of whirling water are the well-known _moulins_, or
“_mills_”, which may be detected from a distance by their gurgling
sounds. The first plunge of the water may not reach to the bottom of
the glacier, in which case the stream finds a passageway below the
surface but above the floor until another crevasse is encountered and
a new plunge made, here perhaps to the bottom. Once upon the valley
floor the stream is joined by others, and pursues its course within an
ice tunnel of its own making (Fig. 421, p. 394) until it issues at the
glacier front.

The coarser of the rock débris which was gathered up by the stream
upon the glacier surface is deposited within the tunnel in imperfect
assortment (gravel and sand), while all finer material and that lifted
from the floor (rock flour) is retained in suspension and gives to the
escaping stream its opaque white appearance. This _glacier milk_ may
generally be traced far down the valleys or out upon the foreland,
and is often the traveler’s first indication that a range which he is
approaching supports glaciers.


FIG. 428.—Ideal form of the surface left on the site of the apron of
a piedmont glacier. _M_, moraine; _T_, outwash; _C_, basin usually
occupied by a lake; _D_, drumlins (after Penck).]

=Deposits within the vacated valley.=—For every excavation of the
higher portions of the upland through glacial sculpture, there is a
corresponding deposit of the excavated materials in lower levels. So
far as these materials are deposited directly by the ice, they form
the lateral, medial, ground, and terminal moraines already described.
A considerable proportion of them are, however, deposited by the
water outside the terminal moraine; but as with the shrinking glacier
the ice front retires in halting movements over the area earlier
ice-covered, the terminal moraines are ranged along the vacated valley
as _recessional moraines_, each with a _valley train_ of outwash below.
About the apron of the piedmont glacier, such deposits are particularly
heavy (Fig. 428). During the “ice age” the Swiss glaciers extended down
the valleys below the existing ice remnants and spread upon the Swiss
foreland as great piedmont glaciers such as may now be seen in Alaska.
To-day we find there moraines and glacial outwash, a lake in the middle
of the apron site, and sometimes a group of radiating drumlins like
those found within the ice lobes of the continental glacier in southern
Wisconsin (Fig. 429, and Fig. 344, p. 317).


FIG. 429.—Moraines and drumlins about Lake Constance upon the site of
the earlier piedmont glacier of the Upper Rhine. The white area outside
the outermost moraine is buried in glacial outwash (after Penck and

Behind the recessional moraines within the glaciated valley are found
the valley moraine lakes (Fig. 448, p. 413), in association with the
rock basin lakes due to glacial sculpture (Fig. 447, p. 412). After
the glacier has vacated its valley, the precipitous side walls become
the prey of frostwork and are the scenes of disastrous avalanches or
landslides. Within the cirques, drifts of snow are nourished long after
the ice has disappeared, and as a consequence the amphitheater walls
succumb to the process of solifluxion (p. 153).

Diversions and reversals of drainage, which are so characteristic of
the work of continental glaciers, are hardly less common to glaciated
mountain districts. Many of our most beautiful waterfalls have resulted
from either the temporary or permanent obstruction of earlier valleys
above the falls. The famous Yosemite Falls offers an interesting
illustration of the shifting of an earlier waterfall, itself no doubt
due to ice blocking in a still earlier glaciation (plate 22 B).

=Marks of the earlier occupation of mountains by glaciers.=—It is
well that we should now bring together within a small compass those
evidences by which the existence of earlier mountain glaciers may be
proven in any district. These marks are so deeply stamped upon the
landscape that no one need err in their interpretation.


 _High-level sculpture._ The grooved upland with its cirques, or the
 fretted upland with its cirques, cols, horns, and comb ridges.

 _Low-level sculpture._ The U-shaped main valley, the hanging side
 valleys with their ribbon falls, the glacier staircase with its rock
 bars and gorges, the rounded, polished, and striated rock floor.

 _Deposits._ The recessional moraines of till and the valley trains of
 sand and gravel, the soled erratic blocks derived always from higher
 levels of the valley.

 _Lakes._ The valley moraine lakes and the chains of rock basin lakes.


 Glacier movement:—

 L. AGASSIZ. Nouvelles Études et Expériences sur les Glaciers Actuels,
 etc., Paris, 1847, pp. 435-539.

 H. HESS. Die Gletscher, Braunschweig, 1904, pp. 115-150.

 H. F. REID. The Mechanics of Glaciers, Jour. Geol., vol. 4, 1896, pp.
 912-928; Glacier Bay and Its Glaciers, 16th Ann. Rept. U. S Geol.
 Surv., Pt. i, 1898, pp. 445-448.

│   PLATE 22.                                                         │
│                                                                     │
│ [Illustration: _A._ Model of the vicinity of Chicago, showing the   │
│ position of the ancient beaches and the outlet of the former Lake   │
│ Chicago.]                                                           │
│                                                                     │
│ [Illustration: _B._ Map of Yosemite Falls and its earlier site near │
│ Eagle Peak (after F. E. Matthes).]                                  │



=Freshwater and saline lakes.=—Lakes require for their existence a
basin within which water may be impounded, and a supply of water more
than sufficient to meet the losses from seepage and evaporation. If
there is a surplus beyond what is needed to meet these losses, lakes
have outlets and remain fresh; their content of mineral matter is
then too slight to be detected by the palate. If, on the other hand,
supply is insufficient for overflow, continued evaporation results in a
concentration of the mineral content of the water, subject as it is to
continual augmentation from the inflowing streams.

As we have seen, there are in areas of small rainfall special
weathering processes which tend to bring out the salts from the
interior of rock masses, these concentrated salts generally first
appearing as a surface efflorescence which is ultimately transferred
through the agency of wind and cloudburst to the characteristically
saline desert lakes.

Lake basins may be formed in many ways. Depressions of the land surface
may result from tectonic movements of the crust; they may be formed by
excavating processes; but in by far the greater number of instances
they result from the obstruction in some manner of valleys which were
before characterized by uniformly forward grades. In relatively few
cases loose materials are heaped up in such a manner as to produce
fairly symmetrical basins.


FIG. 430.—Map and diagram to bring out the characteristics of newland

=Newland lakes.=—On land recently elevated from the sea, basins of
lakes may be merely the inherited slight irregularities of the earlier
sea floor, in which case they may be assumed to be largely the result
of an irregular distribution of deposits derived from the land. Lakes
of this type are especially well exhibited in Florida, and are known as
newland lakes (Fig. 430). Such lakes are exceptionally shallow, and are
apt to have irregular outlines and extremely low banks. Under these
circumstances, they are soon filled with a rank growth of vegetation,
so that it is sometimes difficult to properly distinguish lake and

[Illustration: FIG. 431.—View of the Warner Lakes, Oregon (after

[Illustration: FIG. 432.—Schematic diagrams to illustrate the
characteristics of basin-range lakes.]

=Basin-range lakes.=—Newland lakes may be said to have their origin in
an uplift of the land and sea floor near their common margin. A lake
type dependent upon movements of the earth’s crust but within interior
areas has been described as the basin-range type and is exemplified
by the Warner Lakes of Oregon. In this district great rectangular
blocks of the earth’s crust, which in their upper portions at least
are composed of basaltic lavas, have undergone vertical adjustments
in level and have been tilted so that the corresponding corners of
neighboring blocks have been given a similar degree of down-tilt (Fig.
431). Lakes formed in this way are of triangular outline, are bounded
on the two shorter sides by cliffs, but have extremely flat shores on
their longest side. From this shore the water increases gradually in
depth and attains a maximum depth at or near the opposite angle. Such
lakes naturally betray a tendency to appear in series (Fig. 432), and
are unfortunately much too often illustrated on a small scale after a
shower by the tilted blocks of imperfectly made cement sidewalks.


FIG. 433.—Schematic diagrams of rift-valley lakes, and the rift valley
of the Jordan with the Dead Sea and the Sea of Galilee as remnants of a
larger lake in which their basins were included.]

=Rift-valley lakes.=—Another type of lake basin which has its origin
in faulted block movements is known as the rift-valley lake, and is
best exemplified by the great lakes of east Central Africa. In this
type a strip of crust, many times as long as it is wide, has been
relatively sunk between the blocks on either side so as to produce a
deep rift, or what in Germany is known as a _Graben_ (trench). Such
a basin when occupied by water yields a lake which is long, straight,
deep, and narrow, and is in addition bounded on the sides by steep rock
cliffs. At the ends the shores are generally by contrast decidedly
low. If the hard rock at the bottom of the lake could be examined, it
would be found to be of the same type as that exposed near the top of
the side cliffs. The valley of the Jordan in Palestine is a rift of
this character and was at one time occupied by a long and narrow lake
of which the Dead Sea and the Sea of Galilee are the existing remnants
(Fig. 433).


FIG. 434.—Map showing the rift valley lakes of east Central Africa.]

One of the most striking examples of a rift valley lake is Lake
Tanganyika, while Albert Nyanza, Nyassa, and Rudolf in the same region
are similar (Fig. 434).


FIG. 435.—Earthquake lakes which were formed in the flood plain of the
lower Mississippi during the earthquake of 1811 (after Humphreys).]

=Earthquake lakes.=—The complex adjustments in level of the surface
of the ground at the time of sensible earthquakes are many of them
made apparent in no other way than by the derangements of the surface
water. This is at such times impounded either in pools or in broad
lakes, which inasmuch as they date from known earthquakes have been
called “earthquake lakes”, even though in a strict sense any lake
which has originated in earth movements might properly be regarded as
an earthquake lake. To avoid unnecessary confusion, the term must,
however, be restricted to those lakes which are known to have been
formed at the time of definite earthquakes (Fig. 435). Reelfoot Lake
in Tennessee, which in late years has acquired undesirable notoriety
because of the feuds between the fishermen of the district and the
constituted authorities, is a lake more than twenty miles across
and came into existence during the great earthquake of the lower
Mississippi valley in 1811.

=Crater lakes.=—The craters of volcanic mountains are natural basins
in which surface waters are certain to be collected, provided only
the supply is sufficient and seepage into the loose materials is
not excessive. Some craters, still visibly more or less active, are
occupied by lakes (Fig. 436).


FIG. 436.—View of lake in Poas Crater in Costa Rica, a volcanic crater
more than half a mile across and with walls 800 feet deep. At intervals
there is an ejection of steam mixed with mud and ash after the manner
of a geyser (after H. Pittier).]

In the larger number of cases in which craters become occupied by
lakes, the evidence of continued activity is lacking, and it would
appear in such cases that the lava of the chimney had consolidated into
a volcanic plug, closing the bottom of the crater. Notable groups of
crater lakes are the _Caldera_ of the Roman Campagna (Fig. 437) and
the so-called _maare_ of the Eifel about the Lower Rhine. Crater lakes
are easy to recognize by their circular plan, their steep walls of
volcanic materials, and their considerable depth with a maximum near
the center.

One of the most remarkable of these water-filled basins is Crater Lake
in Oregon, which has a diameter of about six miles and is believed to
have resulted from the incaving of a great volcanic cone in the latest
stage of its activity. This remarkable feature has now been made a
national park and will soon be conveniently reached by tourists and
counted one of the greatest nature wonders of the Pacific slope.


FIG. 437.—Diagrams to illustrate the characteristics of crater lakes.
The Roman Campagna is a plain formed of volcanic ash, with the crater
lakes of Bracciano, Vico, and Bolseno arranged on a line traversing it.]

=Coulée lakes.=—Far more important as lakes are those volcanic basins
which arise from the flow of a stream of lava across the valley of a
river so as to impound its waters (Fig. 438).

At the time of the great eruption under Skaptár Jökull in 1783 the
river Skaptár and many of its tributaries were blocked by the flow of
lava, which it is estimated exceeded in bulk the mass of Mont Blanc.


FIG. 438.—View of Snag Lake, a _coulée_ lake with lava dam shown in
middle distance (after Fairbanks).]

=Morainal lakes.=—As we have learned, the obstruction of drainage,
due to the distribution of rock débris by continental glaciers, has
yielded lakes in almost countless numbers. Probably ninety per cent
or more of the known lakes have had this origin, and the type is so
common within the once glaciated regions that it forms perhaps the
best distinguishing mark of former glaciation. The hummocky surface
of morainal deposits is so characteristic that the lakes of this type
are never very large and are correspondingly irregular in outline.
They have often numerous islands, and their banks are formed of the
combination of rock flour and ice-worn materials known as till (Fig.
439). The smallest of the morainal lakes are mere kettles on the
marginal moraine, and these rapidly become replaced by peat bogs. In
contrast with pit lakes, morainal lakes lack the steep surrounding
slopes and the encircling plain.

[Illustration: FIG. 439.—Diagrams to illustrate the characteristics
of morainal lakes, and a sample map of such lakes from the glaciated
region of North America.]

=Pit lakes.=—The so-called pit lakes have their origin in continental
glaciation, and are found in groups within broad plains of glacial
outwash (mainly sand and gravel), which are for this reason described
as “pitted plains” (see p. 314). Those areas which lay between
neighboring lobes of the ice sheet were subject to particularly heavy
deposits of outwash material, and are, in consequence, particularly
likely to be occupied by pit lakes. As has been pointed out in an
earlier section, the water derived from surface melting within the
marginal portions of a continental glacier descends to the bottom
in the crevasses and thereafter flows in an ice tunnel under the
same conditions as water flowing in a pipe. Having in most cases a
considerable head at the outer margin of the ice, this water may rise
and issue well above the lower ice layers and so cover a portion of the
ice margin beneath sand and gravel (Fig. 440). Separated blocks, often
of massive proportions, are thus buried beneath nonconducting materials
by which they are long protected from further melting. Eventually,
however, with the approach of still milder climates they disappear,
thus causing the overlying sand and gravel to descend and form a pit of
steep walls similar to the sawdust pits over melted ice blocks within
our storehouses.

[Illustration: FIG. 440.—Diagram to show the manner of formation of
pit lakes.]

Pit lakes are thus easily recognized by their occurrence usually in
groups within a plain of glacial outwash and by their characteristic
banks inclined at the angle of repose of such materials (Fig. 441).

[Illustration: FIG. 441.—Diagrams to illustrate the characteristics of
pit lakes and a sample map from the glaciated region of North America.]

=Glint or colk lakes.=—It has been found to be true of existing
continental glaciers that where their mass has been held back by
a mountain wall, their current at the portals within this rampart
becomes greatly accelerated. Though the upper layers of the glacier
in the vicinity may move forward with a velocity of but an inch per
day, the current within the outlet may be as much as seven hundred
or a thousand times as great. In many respects these conditions are
similar to those about the raceway of a reservoir where the near-by
surface of the water is lowered by the indraught of the outlet and the
current in the raceway is so accelerated that, unless protected, the
bottom of the race is carried away and a basin excavated which extends
a short distance both above and below the position of the dam. In
Holland such basins hollowed out beneath breaks in the dykes are known
as colks. Basins which were excavated beneath the glacier outlets by
a similar process would not be open to our inspection until after the
ice had disappeared from the region; but it is most significant that
in Scandinavia, where the Pleistocene continental glacier, advancing
westward from the Baltic, was held in check by the escarpment at the
Norwegian boundary (the _glint_), lake basins have been excavated
in hard rock whose walls show the abrading and polishing which are
characteristic of glacial sculpture, and whose positions are such that
they lie beneath the former outlets partly above and in part below the
line of the escarpment. Their position in reference to the rampart and
to the former outlets is brought out in Fig. 442. The largest of the
glint lakes of this series is Torneträsk in northern Lapland (see p.
277 and Fig. 443).


FIG. 442.—Diagram to show the manner of formation of glint or outlet
lakes where the continental glacier of Scandinavia issued from the
Baltic depression through portals in its mountain rampart.]


FIG. 443.—Map showing a series of glint lakes which lie across the
international boundary of Sweden and Norway.]

=Ice-dam lakes.=—Whenever a continental glacier, either in advancing
its front or in retiring, lies across the lines of drainage upon their
downstream side, water is impounded along the ice front so as to form
ice-dam lakes. Such lakes are found to-day in Greenland and in the
southern Andes, and similar bodies of water of far greater size and
importance came into existence in Pleistocene times each time that the
continental glaciers of northern North America and Europe advanced
upon or retired from suitably directed river systems. Thus above the
Baltic depression, when the ice front lay to the eastward of the main
watershed, each easterly sloping valley was obstructed by the ice and
occupied by an ice-dam lake (Fig. 444), the beaches of which may all be
traced to-day (Fig. 445).


FIG. 444.—Ice-dam lakes (in black) between the front of the late
Pleistocene glacier of northern Europe and the divide near the
Norwegian boundary (after G. de Geer).]

One side of each ice-dam lake is formed by an ice cliff at the glacier
front, and if the region is relatively flat, the remaining shores
are likely to be formed by a marginal moraine which the glacier has
abandoned in its retreat. In their smaller stages, therefore, ice-dam
lakes on prairie country have the form of a crescent, which is the more
pronounced because the waves by their attack upon the ice front flatten
the curvature of its outline (see Fig. 360, p. 330).

The life of an ice-dam lake is begun and ended in important changes
of glacier outline, and after the draining of lakes by this process
the land shores may be traced in beaches, and the ice margin by a
water-laid moraine of low relief (Fig. 359, p. 330).

A much smaller but in many respects similar ice-dam lake is to-day to
be seen at the side of the Great Aletsch glacier, a mountain glacier of
Switzerland. The traveler who makes the easy ascent of the Eggishorn
may look directly down upon this crescent-shaped lake with its ice
cliff on the glacier side (see Fig. 446).


FIG. 445.—Wave-cut terrace at an elevation of 177.5 meters above sea
on the southern slope of the northern Dala valley north of Baggedalen
in Sweden. To the right in the foreground is a peat bog (after


FIG. 446.—View of the Márjelen Lake at the side of the Great Aletsch
glacier, seen looking directly down from the summit of the Eggishorn
(after a photograph by I. D. Scott).]

=Glacier lobe lakes.=—Upon the sites of the former lobes of the
Pleistocene glacier of North America are found the basins of the
Laurentian River system, the largest freshwater lakes in the world.
There has been much controversy concerning the manner of formation
of these lakes, but the view which has seemed to have the largest
following is that they were excavated by the eroding action of the
continental glacier over the drainage basins of former rivers. It is
but one phase of the long controversy between opposing schools, which
have advocated on the one hand the efficiency of glacier ice as an
eroding agent, and upon the other its supposed protection from the
weathering processes. The positions and the outlines of the several
lakes of the series sufficiently proclaim their connection with the
former glacial lobes, and the name which we have adopted leaves the
exact manner of their formation a still open question. The recognition
of the importance of the glacial anticyclone, in giving shape to the
glacier surface and in effecting a transfer of snow from the central to
the marginal portions, has had the effect of emphasizing the relative
importance of erosion under the marginal and lobate portions. Thus
the importance of ice lobes has been greatly accentuated, though this
applies only to the shaping of the basins and not in any important
way to the impounding of the present waters. The present Laurentian
Lakes owe their existence to the elevation by successive uplifts of
the country to the northward and eastward, since the glacier retired
from the lake region. When the ice front lay to the northward of the
Ottawa River, the discharge of the upper lakes was by a channel through
Nipissing River and Lake and thence down the Ottawa River to a gulf in
the lower St. Lawrence. The uplift of the land has had the effect of
raising a barrier where the former outlet existed, and diverting the
waters to a roundabout channel by way of Detroit and Lake Erie (see
Fig. 365, p. 335).


FIG. 447.—Diagrams to illustrate the arrangement and the characters of
rock-basin lakes, together with a map of such lakes from the Bighorn
Mountains in Wyoming.]

=Rock-basin lakes.=—The reversed grades which develop in a valley
deepened by mountain glaciers—the back-tilted treads of the cascade
stairway (see p. 376)—furnish a series of basins hollowed in rock which
are strung along the course of the valley like pearls upon a thread,
or, far better, like the larger beads in a rosary (Fig. 447). This
characteristic arrangement accounts for the name “Paternoster Lakes”
which has sometimes been applied to them in Europe. Their positions in
series within U-shaped mountain valleys, and their rock shores with
characteristically smoothed and striated surfaces, make them easy of
determination. In the higher portions of the valley, where the treads
of the cascade stairway are relatively narrow, such lakes are often
approximately circular in outline, but in the lower levels and upon
wider treads they may be ribbon-like, though lakes of this type are to
a large extent replaced in the lower levels by the valley moraine type
or a combination of the two.

[Illustration: FIG. 448.—Convict Lake, a lake behind a moraine dam
within a glaciated valley of the Sierra Nevadas, California (after a
photograph by Fairbanks).]

=Valley moraine lakes.=—The recessional moraines which mark the halting
stations of mountain glaciers, while retiring up their valleys,
form dams in the later river and so produce a type of lake which is
in contrast with the morainal lakes which result from continental
glaciation. They may, therefore, be distinguished by the name _valley
moraine lakes_. Their positions on the bed of a U-shaped mountain
valley, and the glacial materials which compose the dams, are
sufficient for their identification (Fig. 448). Moraine Lake and Lake
Louise in the Canadian Rockies are typical examples. Rock basin and
valley moraine lakes may occur in alternation or combined in mountain


FIG. 449.—Lake basins produced by successive slides from the steep
walls of a glaciated mountain valley (after Russell).]

=Landslide lakes.=—The sheer-walled valleys which are carved by
mountain glaciers are too steep to long retain their perpendicularity
when the support of the glacier has been removed. Aided by the ever
present joint planes, which admit water to the rock, they succumb to
frost action, and further give way in avalanches whenever the rock
is of sufficiently porous material to become saturated with water.
Landslides sometimes occur successively until the original valley wall
has been replaced by a terraced slope. The treads of the steps in
this terrace have generally a backward-sloping grade, so that basins
are formed to be filled by relatively long and narrow lakes or by
successions of small pools (Fig. 449 and plate 23 B).


FIG. 450.—Lake Garda, a border lake upon the site of a piedmont apron
at the margin of the Alpine highland (after Penck and Brückner).]

When the avalanched material is so disposed as to dam the valley, much
larger lakes of this type come into existence. During an earthquake
which occurred on January 25, 1348, there was a landslide within the
valley of the Gail, Carinthia, which destroyed seventeen villages and
produced a lake which even to-day is represented by a great marsh.

=Border lakes.=—Whenever mountain glaciers push out their fronts
beyond the borders of the mountain range by which they are nourished,
they spread upon the foreland in broad aprons about which morainic
accumulations are particularly heavy. This elevation of morainal
walls about the margins of the aprons yields natural basins that are
occupied by lakes so soon as the glacier retires its front within the
valley. Because such lakes are found at the borders of upland districts
they have been called _border lakes_. The beautiful Lakes Constance,
Lucerne, Maggiore, Lugano, Como, and Garda (Fig. 450), on the borders
of the Alpine highland, are all of this type.

│   PLATE 23.                                                       │
│                                                                   │
│ [Illustration: _A._ View of the American Fall at Niagara, showing │
│ the accumulation of rocks beneath (after Grabau).]                │
│                                                                   │
│ [Illustration: _B._ Crystal Lake, a landslide lake in Colorado.   │
│ (_Photograph by Howland Bancroft._)]                              │

=Ox-bow lakes.=—The cutting off of a meander within the flood plain of
a river yields a lake which is of horseshoe (ox-bow) outline and lies
generally with low banks within a plain composed of river silt. Before
separating from the parent stream the meander had begun to silt up,
especially at the ends. Ox-bow lakes are, however, relatively deep near
the convex shore and correspondingly shallow toward the concave margin
(Fig. 451).

[Illustration: FIG. 451.—Diagrams to bring out the characteristics of
ox-bow lakes, together with a map of such lakes from the flood plain of
the Arkansas River.]


FIG. 452.—Diagrammatic section to illustrate the formation of
saucer-like basins between the levees of streams flowing in a flood

=Saucer lakes.=—As we have learned, a river meandering in its flood
plain has banks which are higher than the average level of the plain,
for the reason that at flood time the main current of the stream still
persists in the channel, thus allowing the burden of sediment to be
dropped in the relatively slack water upon its margin. Because of these
natural embankments or levees, tributary streams are often compelled
to flow long distances in nearly parallel direction before effecting
a junction. Between the trunk stream and its tributaries, likewise
bounded by levees, and between streams and the valley walls, there thus
exist low basins which are more or less saucer-shaped (Fig. 452). At
flood time, when the levees are overflowed or crevassed, water enters
these depressions, and an additional supply may be derived from the
walls of the valley. Good illustrations of such lakes are furnished
by the flood plain of the former river Warren near the banks of the
present Minnesota River (Fig. 453).

[Illustration: FIG. 453.—Saucer lakes upon the bed of the former river
Warren (from the Minneapolis sheet, U. S. G. S.).]

=Crescentic levee lakes.=—As we approach the delta of a river, the
size and importance of the levee increases, and here a new type of
levee lake may develop in series (Fig. 454). At flood time the levee
is breached near the point of sharpest curvature on the convex side
(Fig. 454 _a_). When the waters are subsiding, the current is kept
away from the old channel by the rising grade of the levee as well as
by the inertia of the current, and an entrance to the old channel is
first found below the next change in curvature of the meander, since
here scour becomes effective in cutting through the levee. The new
channel is thus established in the form of a loop inclosing the old
one, and the process of levee building now erects a wall about the
territory newly acquired by the meander. This territory has the form
of a crescent, and when occupied by water produces a crescentic levee
lake often joined to its neighbors in series. The abandoned channel now
closed at both ends by levees may be occupied by water to produce a
subordinate ribbon type of curving trench (Fig. 454 _b_, _c_).

The importance of levees in obstructing drainage to form lakes is only
beginning to be appreciated. It has quite recently been shown that
when trunk streams are greatly swollen and burdened with sediment
while flowing from a receding continental glacier, they may build such
high levees as to aggrade their tributary streams above the junctions,
even producing reversed grades and so impounding the waters to form
extensive lakes. During the “ice age” lakes of this type were formed
in Illinois and Kentucky rivers just above their junctions with the
Ohio. The old lake floor with its eastern shore line and its protruding
islands is easily made out upon the new topographic maps of Kentucky.


FIG. 454.—Levee lakes developed concentrically in series within
meanders of a stream tributary to the Mississippi and flowing upon its
delta plain. _b_ and _c_ are examples of the ribbon type of levee lake
due to occupation of the abandoned river channel. The larger number of
lakes, of which Sip Lake and Texas Lake are examples, have the form of
crescents and lie between abandoned levees (from recent map of U. S. G.

=Raft lakes.=—Within humid regions the flood plains of our larger
rivers are generally forested, and as the river swings from side to
side in its perpetual meanderings, the timber which grows upon the
convex side of each meander is progressively undermined by the river
and felled upon its bank. The prostrate trees remain upon the banks
during the low water of the summer season, to be gathered up at the
time of flood in the next spring season. It is log jams thus acquired
which so generally block the main channel of a river and turn the
current across the neck of the meander when cut-offs occur with the
formation of ox-bow lakes. When the mass of timber thus gathered up
by the river is excessive, as, for example, within the flood plain of
the Red River of Arkansas and Louisiana, huge log rafts are produced
which dam up the river so effectively as to produce temporary lakes.
The impounded waters soon find an outlet over the levee at some point
higher up the river, and the waters flowing off through the timbered
bottom lands, other logs are caught by the standing timber as in a
weir. A second dam is thus formed which is separated from the initial
one by open water, and in this way the driftwood dam acquires enormous
proportions as it gradually moves up the river. After a period of
perhaps a century or more, the lower sections of the jam become decayed
and dislodged so as to float down the river.


FIG. 455.—Raft lakes along the banks of the Red River in Arkansas and
Louisiana at their fullest recorded development (after A. C. Veatch, U.
S. G. S.).]

In the lower Red River a great raft of alternating jams and open water
reached a length of about one hundred and sixty miles and moved up
the river at the average rate of something less than a mile per year.
Within the limits of the dam all tributary streams were blocked, so
that secondary lakes were formed in a double fringe about the main
river (Fig. 455). The great raft which formed here in the latter part
of the fifteenth century has now at the beginning of the twentieth
been largely removed and measures have been adopted to prevent its


FIG. 456.—The Swiss lakes Thun and Brienz, formed by deltas at the
junction of streams tributary to a steep-walled valley.]

=Side-delta lakes.=—It is characteristic of river drainage that the
tributary streams enter the main valley on steeper gradients than the
trunk stream at the point of junction. Wherever the difference in
velocity of the two streams at the junction is large, and the side
stream is charged with sediment, a delta will be formed at the mouth
of the tributary stream. Such deltas push out from the shore and
may eventually block the main channel so as to form a more or less
sausage-shaped expansion of the river—a side-delta lake. Traverse and
Big Stone Lakes in the valley of the Warren River in Minnesota have
been formed in this way (Fig. 354, p. 326). Lakes Thun and Brienz in
the Swiss Alps are of similar origin, the beautiful city of Interlaken
being built upon the delta plain over the valley of the earlier river
(Fig. 456). The Mississippi has similarly been expanded to form Lake
Pepin above the delta at the mouth of the Chippewa River.


FIG. 457.—Delta lakes formed at the mouth of the Mississippi through
the junction of the levees of radiating distributaries with the shore
of the estuary (after Berghaus).]

=Delta lakes.=—A somewhat different type of delta lake has been
formed in Louisiana, where the “father of waters” discharges into the
gulf. Here the various distributaries radiate from the main channel
to produce the “bird-foot” delta type and the toes in this foot by
their junction with the banks which outline the ancient estuary, have
separated in succession a series of basins that before were in direct
connection with the sea (Fig. 457). Lake Pontchartrain is the largest
of this series, while the so-called Lake Borgne is in process of

Where large deltas push out from the shore into the open sea, the
levees which border the individual distributaries are attacked by the
waves and their materials are transported by the shore currents and
built into barriers. These barriers cut off the re-entrants between
neighboring distributaries so as to produce lagoons or lakes (Fig. 458).


FIG. 458.—A type of delta lakes formed by levees in part destroyed
and built into barriers on the margin of the delta of the Nile (after

A type of delta lake, which more resembles the side-delta lake above
described, has formed at the mouth of the Colorado River, where it
enters the Gulf of Lower California. The Imperial Valley lying to the
north of this delta is the desiccated floor of the earlier Gulf of
Lower California which has been captured from the sea by the delta
of the Colorado. The rampart of mountains, by which this valley is
surrounded, has cut it off from any water supply derived from clouds,
and its waters being no longer renewed from the sea, the region has
passed through a period of desiccation which has left the Salton
Sink as the only existing remnant of the earlier lagoon. It will be
remembered that careless operations in diverting distributaries of the
Colorado recently reversed this process so that the waters rose in the
valley, and expensive emergency operations were necessary in order to
again turn the waters of the Colorado into their accustomed channels.

[Illustration: FIG. 459.—Diagrams to illustrate the characteristics of
barrier lakes, with an example from the southern coast of the Island of

=Barrier lakes.=—The Salton Sink illustrates a type of lake which is
formed at the border of the sea through the erection of some kind of
barrier which captures a small area of the ocean’s surface. Though such
lakes may be properly described as strand lakes, it is usually at the
mouth of a river that the process becomes effective. The common type
of _barrier lakes_ is found developed on most ragged coast lines where
the shore currents have formed first bars and later barriers at the
mouths of the estuaries (Fig. 459). Such embankments are usually gently
curving or crescent shaped and are composed of sand or shingle which
presents a steep landward and a gradual seaward slope.


FIG. 460.—Dune lakes on the coast of France (after Berghaus).]

=Dune lakes.=—Within the narrow strips of shore in which all the fine
soil that could be available for plant life has been washed away by
the waves, beach sand is exposed to the direct action of the winds. In
time of storm the sand is picked up and after drifting in the wind is
collected in long ridges parallel to the shore. Constantly traveling
along shore, these dunes block the mouths of rivers and thus produce a
series of lakes such as are indicated in Fig. 460.

=Sink lakes.=—Another class of lakes are due either directly or
indirectly to the work of underground waters. In districts which are
underlain by limestone, the surface water descending along the joints
of the limestone may widen these passageways through solution of the
rock and at lower levels flow on the floors of caverns eaten out by the
same process on bedding planes of the formation. At the intersections
of joints, more or less circular shafts known as “swallow-holes” go
down to the caves from the surface. Locally, also the cavern roofs
give way so as to choke the galleries with rubble and leave a basin at
the surface which has an irregular but generally a more or less oval
outline. If sufficiently clogged at the bottom by finer rock débris,
these basins become occupied by small lakes which are known as sinks,
and constitute one of the best surface indications of a limestone

[Illustration: FIG. 461.—Sink lakes in Florida, with a schematic
diagram to illustrate the manner of their formation (map from U. S. G.

=Karst lakes—poljen.=—In the limestone country to the north and
east of the Adriatic Sea—the so-called Karst region—there are many
interesting features which are directly traceable to the solution
of the country rock. Here all the surface water descends in certain
districts along the widened joint planes so that the drainage is
largely subterranean. The so-called _dolines_ or sinks of very regular
and symmetrical forms resembling deep bowls cover a large part of the

The entire country is, moreover, faulted in the most intricate fashion
into many rift valleys. The drainage being so largely subterranean,
these downthrown blocks of crust, the so-called _poljen_, become
flooded at certain seasons of the year when the subterranean passages
become choked or are too small to carry away all the water. A seasonal
lake of this character is the Zirknitz Lake (p. 189).

=Playa lakes.=—It is the law of the desert that the arid region be
walled in by mountains. This encircling rampart forces the clouds to
rise, and by robbing them of their moisture leaves the desert dry and
barren. Those waters which fall upon the inner margin of the ranges
drain toward the interior of this pan-like depression and are not
returned to the sea—the desert is without an outlet. Infrequent though
they be, the desert rains are of the cloudburst type and in the hills
develop torrents whose waters, emerging upon the desert floor, develop
lakes in the space of a few minutes or at most hours. In the hot and
dry atmosphere the waters of these shallow basins may be sucked up in
the space of a few hours but reappear in the same basins at the time
of the next succeeding cloudburst. Such ephemeral lakes are known as

=Salines.=—Desert lakes more favored in their supply of water may be
relatively long lived and persist for periods measured in years or
centuries. Such lakes are, however, extremely sensitive to climatic
changes (see p. 198).

For the reason that they have no outlet the waters of desert lakes
become salt through continued evaporation. They are, therefore, spoken
of as _salines_. Lake Bonneville, so long as it discharged its waters
over the sill of the Red Rock Pass, must have remained fresh; but when
the level of its waters had fallen below this outlet, its waters became
salt and the content increased as the volume diminished.

The shallow basins upon the floors of desert lakes may have come into
existence in various ways; but it would appear that the irregular
removal of the soil by the winds, modified as this is by differences
in composition of the rock materials and by vegetable growth, and the
deposition of sand by the same agent, are by far the most important.
Many of the types of tectonic and volcanic lakes which have been
described are characteristic of humid and arid regions alike.

=Alluvial-dam lakes.=—Within the mountains upon the desert borders,
the alluvial fans which form at the mouths of valleys, because of the
characteristic cloudburst, sometimes obstruct a main valley at the
junction with its tributaries. By this process the waters of the main
river are impounded in essentially the same manner as are the rivers of
humid regions by the deltas of their tributaries.

=Résumé.=—The types of lakes which we have now considered are arranged
below in tabular form so as to show their relationship to important
geological processes. While not complete, the list includes the more
important classes, as well as others which, while not of common
occurrence, are yet of interest in giving further illustration to the
processes which have been treated in earlier chapters.

By giving careful attention to criteria which have been above
suggested, it should be possible in the greater number of instances at
least to determine whether any lake which is visited has had its origin
in one or another of the processes described.


_Tectonic Lakes_                   _Volcanic Lakes_

  Newland lakes                      Crater lakes
  Basin-range lakes                  Coulée lakes
  Rift-valley lakes
  Earthquake lakes

_Continental Glaciation Lakes_     _Mountain Glaciation Lakes_

  Morainal lakes                     Rock-basin lakes
  Pit lakes                          Valley moraine lakes
  Glint or colk lakes                Landslide lakes
  Ice-dam lakes                      Border lakes
  Glacier-lobe lakes

_River Lakes_                      _Strand Lakes_

  Ox-bow lakes                       Barrier lakes
  Saucer lakes                       Dune lakes
  Crescentic levee lakes
  Raft lakes
  Side-delta lakes
  Delta lakes

_Ground Water Lakes_               _Desert Lakes_

  Sink lakes                         Playa lakes
  Karst lakes—_poljen_               Salines
                                     Alluvial dam lakes.



 I. C. RUSSELL. Lakes of North America. Boston, 1895, pp. 125, pls. 23.

 A. P. BRIGHAM. Lakes, A Study for Teachers, Jour. Sch. Geogr., vol. 1,
 1897, pp. 65-72.

 N. M. FENNEMAN. The Lakes of Southeastern Wisconsin, Bul. 8, Wis.
 Geol. and Nat. Hist. Surv., 1902 (Rev. Ed., 1910), pp. 188, pls. 37.

 A. DELEBECQUE. Les Lacs Français (with Atlas). Paris, 1898. (Work
 crowned by the Society of Geology of Paris.)

 H. R. MILL. Bathymetrical Survey of the English Lakes, Geogr. Jour.,
 vol. 6, 1895, pp. 46-73, 135-166.

 A. SUPAN. Grundzüge der Physischen Erdkunde. Leipzig, 1896, pp.

 H. BERGHAUS. Atlas der Hydrographie. Gotha, 1891, pl. 3.

 R. D. SALISBURY. Physiography. 1907, pp. 292-327.

 CHARLES RABOT. Revue de limnologie, La Géographie, Vol. 4, 1901, pp.
 110-119, 172, 189.

 I. C. RUSSELL. A Geological Reconnaissance in Southern Oregon, 4th
 Ann. Rept. U. S. Geol. Surv., 1884, pp. 442-447. (Basin range lakes.)

 ED. SUESS. The Face of the Earth, vol. 4, 1909, pp. 268-286. (Rift
 valley lakes.)

 J. S. DILLER. Crater Lake, Nat. Geogr. Mag., vol. 8, 1897, pp. 33-48,
 pl. 1; Geology of Lassen Peak Quadrangle, California, Geol. Fol. 15,
 U. S. Geol. Surv., 1895. (Coulée lakes.)

 N. M. FENNEMAN. Lakes of Southeastern Wisconsin, _l.c._, pp. 4-6. (Pit

 ED. SUESS. The Face of the Earth, vol. 2, 1906, pp. 340-346, pl. 7.
 (Glint lakes.)

 I. C. RUSSELL. A Preliminary Paper on the Geology of the Cascade
 Mountains in Northern Washington, 20th Ann. Rept. U. S. Geol. Surv.
 Pt. ii, 1900, pl. 14. (View of a rock-basin lake.)

 E. W. SHAW. Preliminary Statement concerning a New System of
 Quaternary Lakes in the Mississippi Basin, Jour. Geol., 1911, pp.
 481-491. (New type of levee lakes.)

 A. C. VEATCH. Formation and Destruction of the Lakes of the Red River
 Valley, Prof. Pap. No. 46, U. S. Geol. Surv., pp. 60-62, pls. 29-33.
 (Raft lakes.)

 M. NEUMEYER. Erdgeschichte, vol. 1, pp. 595-596. (Poljen.)



=Lakes as settling basins.=—Of all the processes which conspire to
blot out the lakes with which our northern landscapes are dotted,
the one of greatest importance is in most cases a process of filling
by the sediments brought in by tributary streams. The carrying of
sediment in suspension depends, as we know, upon the velocity of the
current, and as this is checked where it reaches the lake margin, all
coarser material is at once deposited to form a delta, while the finer
sediments are held longer in suspension and finally settle in thin
layers over the entire bottom of the lake. Clay deposits surrounded by
coarser sediments are thus characteristic of filled lake basins.

[Illustration: FIG. 462.—Map of the Arve and the upper Rhone to show
the importance of Lake Geneva as a settling basin of the larger stream.]

How waters are clarified by their passage through a lake is indicated
by a comparison of a river system such as the St. Lawrence, with
a river like the Missouri and Mississippi. Not only are the lower
stretches of the St. Lawrence in striking contrast with the muddy
floods of the Missouri and Mississippi; but the delta, which is so
remarkable a feature in the Mississippi, has no counterpart in the
northern river.

The most noteworthy examples of settling are, however, furnished by
the lakes of Switzerland, for the reason that SWISS rivers are heavily
charged with rock flour produced beneath the numerous glaciers at the
valley heads, and, further, because these rivers descend with turbulent
currents to near the borders of the larger lakes. To look out upon the
murky waters of the upper Rhone, where they enter Lake Geneva near
Villeneuve, and then to watch the flood of crystal water which issues
from the lake and passes under the bridge at Geneva, is an object
lesson which no traveling student should miss (Fig. 462). Yet even more
instructive is a visit to the _Bois de la Bâtie_ at the junction of
this clear stream with the Arve, a half hour’s walk only below Geneva.
The waters of the Arve have come on a steep descent directly from the
glaciers of the Mont Blanc district, and as they meet the cleared
waters of the Rhone, they flow beside them down the common valley
without mingling. Dull and opaque, the Arve waters can be discerned for
a long distance as a white belt against the left bank of the river,
sharply defined against the blue reflecting surface of the Rhone waters
(Fig. 463). Upon the banks of the Arve, just above its junction, a
cement manufactory has been established to utilize the clays which are
here deposited.


FIG. 463.—View looking upstream across the opaque waters of the Arve
to the clear reflecting surface of the Rhone. To the right across the
Arve is seen the cement works for recovering the Arve sediments.]

Wherever lakes are contained in long and narrow valleys, the greater
part of the tributary drainage enters at the upper end, and the delta
which there forms extends from bank to bank. As it continues to advance
into the lake, the earlier water basin is gradually transformed into a
level plain of delta deposit, a feature so common as to be deserving
of a special name. The Scottish lochs, which are lakes of this type,
are each extended in a longer or shorter delta plain described as a
_strath_, and this local term may well be given a general application
(frontispiece). The city of Ithaca, the seat of Cornell University, is
built upon a strath at the head of Lake Cayuga, and numberless Scottish
and Swiss hamlets have been located upon such fertile plains (Fig. 464).


FIG. 464.—The village of Poschiavo in eastern Switzerland, built upon
a strath at the head of Lake Poschiavo.]

=Drawing off of water by erosion of outlet.=—Next in importance to
the filling up of lake basins as a factor in their early extinction is
the cutting down of their channels of outflow. Whenever the walls of
the outlet are cut in rock, this draining process is apt to be slow,
for the reason that the outlet stream is of filtered water and so
lacks the necessary cutting tools. By far the larger number of lakes
are, however, held back by dams of loose drift deposits laid down by
the earlier continental glaciers; and so the very clarity of the water
promotes the erosion of the outlet by allowing the stream’s full burden
of sediment to be lifted and then removed from the channel.

=The pulling in of headlands and the cutting off of bays.=—The removal
of projecting headlands by wave action, though it increases the area of
the lake, yet it decreases directly the volume of lake water through
formation of the built terrace, and indirectly in far larger measure
through the transformation of bays into quiet lagoons within which the
extinguishing process of peat growth is set in operation.

=Lake extinction by peat growth.=—The first condition for the growth
of lake vegetation is quiet water. Within small lakes, such as the
kettle basins upon moraines, aquatic vegetation develops rapidly,
and bogs of peat might almost be included among the most important
distinguishing marks of a glaciated country. Within larger lakes it is
only after barrier beaches have been thrown across the mouths of the
bays to form natural breakwaters for the waves that this process of
lake extinction by peat growth can become effective.


FIG. 465.—View of the floating bog and surrounding zones of vegetation
in a small glacial lake of the Yellowstone National Park (after a
photograph by Fairbanks).]

Many erroneous notions are still held concerning the prime importance
of sphagnum in peat formation, owing to the peculiar local conditions
under which the early studies were made. Within the glaciated districts
of the United States, the formation of peat involves the successive
growths of a number of zones of vegetation and the formation of a
floating bog which advances into the lake from the shores, followed
in turn by belts of low shrubs, tamaracks, and lastly deciduous trees
(Fig. 465).

In most cases the first plants to develop in a quiet lake are the water
lilies, though these are sometimes preceded by chara and floating
bladderwort. Next behind the water lilies come the sedges, which form a
mat of floating bog by their grasslike stems sinking down in the water
and being there interwoven with the rhizomes below. This mat of sedge
is often so firm that cattle may advance upon it to the water’s edge,
but it is separated by a layer of water from the bed of growing peat
at the bottom of the lake (Fig. 466). This bed of peat appears to grow
upward toward the surface and become joined to the shore end of the
floating bog by decaying vegetation which is dropped from the bottom of
the mat above.

[Illustration: FIG. 466.—Diagram to show how small lakes are
transformed into peat bogs (after C. A. Davis).]

In order behind the floating bog come the advanced plants of the
conifer group, with sphagnum and low shrub here upon a peat base
extending to the lake bottom. Behind the belt of shrubs arise the
tamaracks and spruces, and lastly, toward the shore, come the deciduous
trees and especially poplars, maples, and marginal willows. Upon the
margin of the basin there is usually a low trench, or “fosse”, filled
with water during wet seasons, as a result, no doubt, of seasonal
inwash that does not reach the residual lake toward the center of the

=Extinction of lakes in desert regions.=—In arid regions there are
special causes of lake extinction. Thus the blowing in of sand and dust
carried for long distances in the air, a by no means negligible factor
even in humid regions, here assumes large importance. The now exposed
basins of extinct desert lakes afford the evidence, however, of an even
greater factor of extinction, in climatic change. The clouds, which
at one time found their way into the drainage basin of a lake, may
later through the rise of a mountain barrier be cut off, and so with
reduced water supply a period of lake desiccation is begun. When, in
this process of drying up, the lake level has fallen below that of the
outlet, the saline content of the waters begins to increase, and later
a stage is reached, as in Great Salt Lake, when the sodium salts are
precipitated. When the lake has become extinct, these deposits remain
as a witness to the changed climatic condition.

=The rôle of lakes in the economy of nature.=—It is natural, in
considering the extinction of lakes, to give some attention to the
rôle which they play in the economy of nature. That lakes filter the
water of rivers, and prevent the formation of important delta deposits,
has already been noticed. A curious exception to this general rule is
furnished by the great delta at the head of Lake St. Clair, just below
the outlet of Lake Huron. This anomaly is, however, explained by the
peculiar currents of Lake Huron, which are so directed as to sweep the
beach sand into the swift current of the outlet, to be deposited in the
quiet waters of Lake St. Clair (Fig. 467).


FIG. 467.—Map to show anomalous position of the delta in Lake St.
Clair, due to the peculiar currents in Lake Huron (after maps by Cole).]

As regulators of the flow of rivers, lakes perform an important
function. Such disastrous floods as are characteristic of the spring
season within the basin of the lower Mississippi could not occur in the
lower St. Lawrence, for the reason that the great basins of the lakes
serve as distributing reservoirs. The annual floods, upon which the
agriculture of Egypt depends, are explained by the flood waters from
the high mountains of Abyssinia entering the Nile _below_ the lakes of
its upper basin.

In one further respect large inland bodies of water have an important
function as regulators. It is the property of water to respond but
slowly to the variations in the quantity of heat which reaches the
earth’s surface from the sun. A larger quantity of heat must be
added to or abstracted from a body of water, in order to change its
temperature by one degree, than would be required for a like change in
the same bulk of earth or rock. Thus bodies of water by more slowly
acquiring the summer’s heat retard the coming spring, and by storing
up this energy and carrying it over into the autumn the warm season is
prolonged and early frosts prevented. The fruit belts about the lower
Great Lakes are thus dependent upon this regulating property of the
lake waters. The discomfort of the long spring of raw weather is thus
compensated by an unusually salubrious harvest season.

=Ice ramparts on lake shores.=—Small ridges known as ice ramparts are
formed upon lake shores by the action of lake ice, though subject to
so many qualifying conditions that the range of their occurrence is
somewhat limited. Within districts where a winter ice cover of some
thickness is formed, the shores of lakes are apt to present ridges of
bowlders parallel to and near the water’s edge, and such lakes have
sometimes become known as “wall lakes” (Fig. 468).


FIG. 468.—A bowlder wall upon the shore of a small lake in the
Adirondacks of New York.]

In many cases these small ridges have been formed at the time of the
spring “break up” of the ice; for the ice cover, when once loosened,
is drifted in great rafts first against one shore, and later, with a
change of wind direction, against another. Under the impact of such
heavy rafts, the half-submerged bowlders near the shore are forced up
the beach until they lie in a ridge or bowlder wall.

At other times such bowlder walls, and far more interesting ridges as
well, result from a kind of ice shove independent of the wind, but
caused by expansion within the ice itself during a sudden rise of
temperature of the surrounding air. Such ice ramparts require for their
explanation a consideration of the sequence of events from the time the
ice cover closes the lakes.

[Illustration: FIG. 469.—Diagrams to show the effect of ice shove in
producing ice ramparts upon the shores of lakes (after Buckley with a
slight modification).]

The first lake ice of early winter forms in most cases with air
temperatures a few degrees only below the freezing point of the water.
When later a severe “cold wave” arrives, the ice cover is contracted
and becomes too small for the lake surface. To this contraction it
yields and opens cracks up which the water rises, and in the prevailing
low temperature this water is quickly frozen and the lake cover again
made complete. Skaters are familiar with the opening of these cracks
and the loud “roaring” which accompanies it on cold mornings, the sharp
skate runners sometimes starting a crack in the strained ice, as does a
light scratch upon glass that is in a similar strained condition.

[Illustration: FIG. 470.—Various forms of ice ramparts (after

The original ice cover of the lake, which was formed at near-freezing
temperatures, has now received a number of inserted wedges of new ice
at a time when its contracted volume has made this possible. If now
a “warm wave” succeeds to the “cold wave” in the air, the ice cover
expands at a rate corresponding to its rate of contraction, so that a
strong pressure is exerted against the shore (Fig. 469). Sliding up
the sloping surface of the cut and built terrace, the force of this
shove may be deflected upward against the cliff, and if this is of
loose materials, the effect may be to ram bowlders into the bank, to
push up ramparts or ridges, to overturn trees, etc. (Fig. 470). In
marsh land the frozen surface layer may slide over its unfrozen base
and be forced up into broken folds (lower diagram of Figs. 469 and 470).


FIG. 471.—Map of Lake Mendota at Madison, Wisconsin, showing the
position of the ridge which forms from ice expansion, and the ice
ramparts about the shores of the bays (based on Buckley’s map).]

In order that ice ramparts may be formed, it is necessary that
the winter climate of the district be severe and characterized by
alternating cold and warm waves, involving considerable range of air
temperature below the freezing point. If the lake is small, the push of
the ice will be through so small a distance as not to yield appreciable
ramparts. If, on the other hand, the lake is too large, the ice cover
is not rigid enough to transmit the push to the distant shore, but,
like a long beam employed in the same manner to transmit a compressive
stress, it is bent out of a straight line and later broken. Thus in a
broad lake, with the coming of a “warm wave”, the ice cover opens in a
crack from shore to shore and finds relief from the stress by pushing
up a ridge above the crack. On such lakes ice ramparts are found only
about the shores of bays whose expanse does not greatly exceed a mile
(Fig. 471).

When there is heavy snowfall, ice ramparts either do not form or are
of smaller dimensions, probably in part because the ice is blanketed
by the snow and so prevented from sudden elevation of temperature
during the “warm wave”, but even more because the ice cover is sensibly
bowed down under its load and so rendered incompetent to transmit the
developed stresses to the shores.


 Lake extinction by peat growth:

 C. A. DAVIS. Peat, Essays on its Origin, Uses, and Distribution in
 Michigan, Ann. Rept. Mich. Geol. Surv. for 1906, 1907, pp. 105-182;
 Peat Deposits as Geological Records, 10th Rept. Mich. Acad. Sci.,
 1908, pp. 107-112.

 G. P. BURNS. Bog Studies. Ann Arbor, 1906, pp. 13.

Ice ramparts:

 C. H. HITCHCOCK. Shore Ramparts in Vermont, Proc. Am. Assoc. Adv.
 Sci., vol. 13, 1869, pp. 335-337.

 G. K. GILBERT. Lake Bonneville, Mon. 1, U. S. Geol. Surv., 1890, pp.

 E. R. BUCKLEY. Ice Ramparts, Trans. Wis. Acad. Sci., etc., vol. 13,
 1900, pp. 141-162, pls. 1-18.

 WILLIAM H. HOBBS. Requisite Conditions for the Formation of Ice
 Ramparts, Jour. Geol., vol. 19, 1911, pp. 157-160.



=A mountain defined.=—As ordinarily understood, mountains are
elevations upon the earth’s surface which rise above the general level
of the country. Their summits need not be at great heights above the
sea, but it is essential that they project above the average level
of the surrounding country by at least a quarter of a mile. Lower
elevations are described as hills. On the other hand, the elevation of
a plateau like the “High Plains” of the western United States may be as
much as a mile, but the vast expanse of nearly level surface precludes
the use of the term “mountain.” The word is thus applied to a feature
of the earth and not merely to an elevated tract.

In a collective sense, though more often in the plural form, the
term is properly applied to groups of similar features which have a
common origin in local uplift of the land. The origin of mountains
used in this sense of mountain complexes is thus connected with some
essentially local uplift of the earth’s surface. This may take place
by the processes of folding and superincumbent fault displacement,
by volcanic extravasations or ejections, or by a deeper seated and
essentially hydrostatic elevation of rock beds over molten rock

The existing _forms_ of mountains, as we are to see, are largely shaped
by the erosional processes which are set in operation by the uplift
itself, though often completed long subsequent to it.

=The festoons of mountain arcs.=—From our earliest studies of school
geographies, we have become familiar with the arrangement of the more
important mountains in long chains or systems. Comparatively few
persons have given any further attention to the arrangement of the
chains, though over large areas of the earth’s surface the distribution
of mountain ranges is deeply significant. The map of Asia in particular
presents a series of great sweeping arcs or crescents which are grouped
as though hung upon the map in festoons with knots or vertexes to
separate neighboring groups (Fig. 474, p. 438, and Fig. 472).


FIG. 472.—The great multiple mountain arc of Sewestan, British India
(after de Saint Martin and Schrader).]

The significance of these mountain groupings in the evolution of the
earth’s surface has been pointed out by the great Viennese geologist
Suess, to whom we are indebted for focusing upon the plan of the
earth an amount of attention which before had been largely given to
the preparation of hypothetical sections of strata which were largely
buried from sight beneath the earth’s surface. Broadly speaking,
the mountain arcs may be said to be grouped about those shields of
older rock which geological studies have shown to be the oldest land
masses upon the globe. Within the northern hemisphere these original
continents are represented by the areas of crystalline rock centered
over Hudson Bay, the Baltic Sea, and an area in northeastern Siberia
known to geologists as Angara Land. In our study of the figure of
the earth (Chapter II) it was found that these shields represent the
truncated angles of the rounded tetrahedral form toward which the
planet is tending (Fig. 3, p. 12).

=Theories of origin of the mountain arcs.=—The mountain arcs, when
studied in detail, are found to be composed of closely folded rock
strata, the flexures of which are generally so overturned that their
axial planes dip toward the center of the arc (Fig. 473). It was the
view of Suess that these arcs are to be explained by a pushing outward
of the rock strata from the center of the arc toward its periphery,
thus causing a wrinkling of the surface strata and an overriding of the
surrounding formations, which upon this hypothesis opposed a greater
resistance to the sliding movement. The folding together of the strata
due to the sliding naturally involves a very considerable diminution
of the surface area presented by the strata (Fig. 22, p. 42). In the
case of the Alpine chains it has been estimated that a flat land area,
four hundred to eight hundred miles across, has by the folding process
been reduced to a width of only about one hundred miles, or from a
fourth to an eighth of its former width.


FIG. 473.—_a_, diagram to illustrate the Suess’ theory of the origin
of mountain arcs; _b_, the author’s modification of this view.]

The weakness of Professor Suess’ theory lies in the fact that such
compression as it implies is assumed to be due to an _outward_
movement of the relatively small area of the earth’s outer shell
which is included _within_ the arc. It must be obvious that such a
movement, being from a center toward three sides at once, would for
this circumscribed area involve enormous proportionate reduction in
superficial area of the strata and could only result in a hiatus near
the center of the arc. No such gap is to be found, and one would,
moreover, be difficult to account for upon any plausible hypothesis.
On the other hand, the general contraction of the planet as a whole,
involving as it does reduction of surface over large areas, is a
well-recognized fact; and if it be true that the shields formed by the
older continents are less subject to contraction than the remaining
portions of the surface, it is easy to understand why the earth’s outer
skin should be wrinkled by _underfolding_ and thrusting about these
continental margins. The contrast of this view with that of Professor
Suess is expressed in the diagrams of Fig. 473.


FIG. 474.—Festoons of mountain arcs about the borders of the Pacific
Ocean—Pacific type of coast (based upon Suess).]

We may illustrate this conception by a stretched sheet of rubber cloth
such as is in common use by dentists, upon which a thin layer of hot
Canada balsam has been spread. This substance congeals upon cooling to
near-normal temperatures, and if a small local area of the balsam layer
be chilled and the tension upon the rubber then released, the viscous
balsam of the unchilled portion of the layer is thrown into wrinkles
about the cooled and more resistant areas. These more resistant
portions of the stratum may thus represent the ancient continental
shields of our planet.

=The Atlantic and Pacific coasts contrasted.=—In his studies of
mountain arcs in their relation to the plan of the earth, Professor
Suess has shown how the arrangements of the mountain chains about the
two larger oceans represent two strongly contrasted types. Whereas
about the Pacific margin the mountain arcs are, as it were, strung in
festoons which trend parallel to and are convex toward the coast, or
else lie in fringing garlands of islands in the same attitude (Fig.
474); the mountain chains about the Atlantic become sharply truncated
as they reach the coast, and thus indicate that the basin of this ocean
has been produced by an inthrow or depression between great marginal
displacements in some period subsequent to the formation of the


FIG. 475.—The interrupted system of the Armorican Mountains common to
western Europe and eastern North America (after Arldt).]

Thus the mountain folds of the Appalachian system are in Newfoundland
cut off abruptly at the coast line, and the same beds, similarly
truncated, are encountered again across the expanse of ocean in the
folds at the coast of western Europe (Fig. 475). In discontinuous
remnants this ancient mountain chain may be traced in an east and
west direction across western and central Europe. We have thus here
to do with a single mountain system which extends from central Europe
to northern Alabama, out of which a great link has been taken by the
subsequent sinking in of the basin of the Atlantic Ocean.


FIG. 476.—Schematic representation of a “zone of diverse displacement”
in the Great Basin of the western United States (after Powell).]

=The block type of mountain.=—The inclusion of most elevations in
mountain chains and arcs is one of the most obvious facts to any one
who has examined world atlases with this subject in mind. Such chains
are almost invariably composed of folded rocks, thus indicating that
erosion has removed great superincumbent masses of strata since the
crustal compression produced the folds at considerable depths below the
then surface.

There are, however, large elevated tracts upon the earth’s surface
which are intersected by deep valleys, but where no arrangement of the
elevated portions within chains or ranges is to be detected. In such
cases the distribution of mountain and valley may bear a resemblance to
a mosaic of disturbed parts which stand at different levels (Fig. 476).

[Illustration: FIG. 477.—Section of an East African block mountain
(after J. W. Gregory).]

Such block mountain districts are to be found in many parts of the
earth’s surface, but notably within the Great Basin of the western
United States, and in the land area which borders the Indian Ocean
upon the west and northwest. In contrast with the mountain arcs, so
strikingly exemplified by the continent of Asia as a whole, its extreme
southwestern portion is made up of an alternation of plateau and
rift valley separated from each other by great displacements. Though
modified to some extent by erosion, the elevations seem generally to
represent the displaced crust blocks which in mutual adjustments have
been left at the highest levels (Fig. 477). The valley of the Jordan,
with the mountains of Lebanon rising above it, is near the northern
extremity of this faulted mountain region (Fig. 434, p. 404), while
the Great Rift valley, crossing east Central Africa, and the many
neighboring rifts to the east and west, are graven in lines so deep
that an observer upon a neighboring planet might perhaps detect them.

It is not necessary in all cases to assume that the block mountains of
a faulted district represent the blocks which in the adjustments were
left the highest. Erosion in the course of time accomplishes marvels of
transformation, and it may result that heavy masses of more resistant
rock eventually project the highest, even though they may represent the
downthrown blocks in the fault mosaic (Fig. 43, p. 60).

[Illustration: FIG. 478.—Tilted crust blocks in the Queantoweap

Where in addition to undergoing changes of level the earth blocks
have been tilted, the features long since described from our western
interior basin as “Basin Range structure” are developed. Here the
upper surface of the disturbed earth blocks betrays the evidence of a
definite tilt in some one direction (Fig. 478, and Fig. 431, p. 402).

=Mountains of outflow or upheap.=—An important type of mountain,
generally described as volcanic, may be due either to the outflow of
lava at the earth’s surface, or to accumulations of separated fragments
of lava, first thrown into the air, and then deposited by gravity or
admixed with water as volcanic mud. Such mountains, both before and
after modification by erosion, assume the strikingly characteristic
forms which have been fully discussed in Chapters IX and X. The
dominant types are the lava dome and the puy, the cinder cone, and
the more complex composite cone. Excepting only the surface produced
by the few great fissure eruptions and the semivolcanic mesa type, the
individual mountains of volcanic origin develop features with notably
circular bases.


FIG. 479.—Pen drawing of the laccolite of the Carriso Mountain by W.
H. Holmes, which shows the jagged surface of the igneous rock core and
the sloping tables which still remain of the roof of sedimentary rocks
(after Cross).]


FIG. 480.—Map of laccolitic mountains. A portion of the Judith
Mountains, Montana. The intrusive igneous rock is shown in black (after

=Domed mountains of uplift—laccolites.=—At a considerable number
of widely separated localities upon the earth’s surface, mountainous
regions are encountered, the central areas or cores of which are
composed of intrusive igneous rock such as granite, and about this
core the sediments dip away in all directions as though they had once
formed a continuous roof above it and had been forced into this dome by
hydrostatic pressure of the once viscous material beneath (Fig. 152, p.
143, and Figs. 479 and 480). Examples of such domed mountains of uplift
were first described by Gilbert from the Henry Mountains of Utah, but
instances are furnished by many elevated tracts, especially within
the western United States. Such mountains are known as _laccolites_,
but when one margin at least of the igneous core corresponds to a
displacement, the mountain is described as a _bysmalite_ (Fig. 481).

[Illustration: FIG. 481.—Ideal sections of laccolite and bysmalite.]

When subjected to long-continued erosion, the generally fissured
granitic core of the laccolite weathers in a wholly different manner
from the bedded sediments which surround and still in part mount
over it. The former usually presents a more or less jagged surface
which contrasts sharply with the gently sloping tables of the latter
(Fig. 479). About the high granite core of the mountain, the several
strata of the uptilted formations present each a steep slope toward
this higher land, and a gentler slope in the opposite direction.
Such unsymmetrical ridges which surround the mountain area are often
referred to as “hog backs” (plate 12 B). The arrangement of the
strata in the hog backs thus presents an overlapping series like the
shingles upon a roof, except that the overlapping is here from the
bottom instead of the top, and the exposed ends thus face toward the
crest. Unlike a shingle roof the hog backs do not shed the water which
descends to them from the higher levels, but, on the contrary, they
cause it to flow in troughs parallel to the base of the slope except
where outlets are found through them.

=Mountains carved from plateaus.=—In the mountain types thus far
discussed, the local uplifting of the land has itself developed
features which in the aggregate may be referred to as mountains, even
though the characters of the original surface are soon destroyed by
erosive processes of one sort or the other. Erosive processes are,
however, quite competent to produce mountain forms from a featureless
plateau, and particularly through the incision by streams of running
water, the best studied process of mountain sculpture (see Chapters
XI-XIII). This process of throwing valleys about an elevated section
of the earth’s surface, and so carving out mountains, is sometimes
described as _circumvallation_; and if the term “mountain” be applied
in its ordinary sense to describe an individual feature, it is clear
that most mountains have been formed in this way.

To discuss the characteristic shapes of such mountains would be largely
to review the contents of this book, and especially those portions
which discuss the character profiles resulting from the action of each
sculpturing or molding agent. The work of frost and other weathering
agencies, of running water, of mountain and of continental glacier,
would all have to be considered in order to evolve the history of each

In addition to discovering the agents which were chiefly responsible
for the shaping of the mountain, we may, further, in many cases
determine at what stage the work of one agent has been succeeded by
that of another, and at least at what stage of its complete cycle of
activity the latest agent is now at work.


FIG. 482.—The gabled façade so largely developed in desert landscapes
and sharply contrasted with the recurring curves in the landscapes of
humid districts (from a painting of the Grand Cañon of the Colorado by

=The climatic conditions of the mountain sculpture.=—Since the
different geological agencies operate either in a different manner
or with differences in vigor according to the varying climatic
conditions, the mountains of arid regions may in most cases be readily
differentiated from those of the more habitable humid sections of
country. In broad lines these differences may be summed up in the
greater prevalence of the curving line within the landscapes of humid
districts. This may be largely ascribed to the influence of the
mat of vegetation, which protects the rock surface from more rapid
mechanical degeneration, and arrests the sliding movements within
the already loosened rock débris. In place of the reversed curves
of the lines of beauty, so generally observed in the landscapes of
well-watered regions, the desert lands present ever a repetition of the
vertical cliff alternating with a sort of many gabled façade which is
occasionally due to truncation of mountain spurs by the waves of former
lakes, but far more often the outlines of débris cones built up beneath
each prominent joint of the cliff walls (Fig. 482).

=The effect of the resistant stratum.=—In a striking manner mountain
landscapes may disclose the influence of the diversified rock materials
and of the rock structures as well. After prolonged erosion there
is likely to be little correspondence between the positions of the
anticlinal folds and the crests of the higher mountains. Such mountains
are, in fact, much more likely to rise over synclines than upon the
site of anticlines. The traveler who enters the Alps by any of the
several railways, or who journeys by steamer over the beautiful lake
of Lucerne, has a most favorable opportunity to study the position
of the rock folds in the mountain sections that are unrolled in
succession before him. Rarely indeed will he find a definite anticline
in correspondence with a mountain peak, for the layers which are most
resistant have developed the peaks, and it is because the outer layers
of the anticlines open by local tension (see Fig. 26, p. 45) that
they were first cut away by erosion, so that the hard layers within
the synclines are likely to constitute the peaks within the existing


FIG. 483.—The Mythen, composed of Jurassic and Cretaceous sediments,
and resting upon softer Tertiary formations. View from a balloon (after
a photograph by C. Schmidt).]

When, as sometimes happens, an older and likewise more resistant bed
has been folded back upon younger and softer formations, an isolated
remnant may be found “unrooted” to its base, upon which it appears as
though floating within a billowy sea of the softer formations (Fig.

=The mark of the rift in the eroded mountains.=—Applying the term
“mountain” in its collective sense for a circumscribed area of uplifted
crust, whether represented to-day by a folded or a faulted complex,
a lava mass, or a granite dome; the period of uplift has marked the
beginning of the activity of sculpturing agencies. By these the mass
is pared down as it is shaped into a more or less intricate design of
component and essentially repeating units. In the vernacular the word
“mountain” is applied to these units into which the larger mountain
mass is subdivided.


FIG. 484.—The battlement type of erosion mountains. Die Drei Zinnen
(Three Battlements) in the Dolomites (after Marr).]

It has been one of the main objects of this work to point out that the
peculiar shapes of these elementary mountains are each characteristic
of the erosive agents which produced them, and that each surface has
marks which may be recognized in those lines of profile which recur
within the landscape—the character profiles. In the subdivision
of the larger mass—the _genetical_ mountain—to form the numerous
smaller masses—the _erosional_ or _circumvallational_ mountains—there
is disclosed a pattern of fractures which has guided the erosional
agents in their incisional operations (see Chapter XVII). In high
altitudes, where the action of frost is so potent in prying at the
wider fractures, this subdivision of the mass may be revealed by the
sculpturing of squared towers or battlements (Fig. 484).

[Illustration: FIG. 485.—Symmetrically formed low islands repeated in
ranks upon Temagami Lake, Ontario.]

For other examples in which the sculptured surface is largely the
handiwork of a single erosional agent, as over vast areas in the
Canadian wilderness, the revelation of the fracture design is no less
apparent. Here a series of crystalline rocks underlie broad expanses
of territory and are without noteworthy variations of hardness and
almost bare of surface débris. Sculptured beneath a mantling ice sheet,
excavation has naturally been concentrated above the more widely
gaping fissures of the joint-fault system, doubtless already marked out
in the river network which the glacier overrode. The result has been a
division of the surface into a series of low, oval ridges or hummocks,
which over vast areas are repeated with monotonous regularity. Wherever
the lower levels have been flooded, symmetrical low islands of nearly
uniform elevation rise from the expanse of water and may be counted
by thousands. Though the smaller islands have notably regular shore
lines, the larger ones disclose their composition from smaller units
by the breaking of their shores into similar bays spaced with regular
intervals (Fig. 485, and Figs. 243 and 245, p. 229).

The ever repeating fracture design of the earth’s crust is not
restricted to the mountain masses which it has broken up, and the
unity of which it has done so much to conceal. It extends far outside
the margin of these masses, and is in fact common to whole continents
and perhaps even to the planet as a whole. The part played by this
design of fractures in the control of the sculpture of landscapes it
would be hard to overestimate. Through its influence the striking
features molded by one agent have been merged in the contrasted shapes
developed by another. It is the great outline blender in the creation
of nature’s masterpieces of form and color. Thus the lines of this
mysterious fracture network, though stamped in indelible characters
upon our landscapes, are generally lost in the ensemble effect and may
long remain undiscovered. Like a moss-grown inscription upon a slab of
marble, though veiled, it may yet be deciphered; and if the veil be
withdrawn, the runic characters are disclosed, and one of nature’s laws
lies open before us.


 Mountain arcs or festoons:—

 ED. SUESS. The Face of the Earth, vol. 2, 1906, pp. 201-207; vol. 4,
 1909, pp. 498-542.

Block mountains:—

 G. K. GILBERT. Surveys West of the 100th Meridian (Wheeler), vol. 3,
 Geology, Washington, 1875, Pt. 1, pp. 19 _et seq._, 48.

 J. W. POWELL. Report on the Geology of the Eastern Portion of the
 Uinta Mountains and a Region of Country Adjacent thereto, U. S. Geol.
 and Geogr. Surv. Ter., II Div. Washington, 1876, pp. 218.

 JOHN W. GREGORY. The Great Rift Valley. London, 1896, pp. 422.

Laccolites and bysmalites:—

 G. K. GILBERT. Report on the Geology of the Henry Mountains, U. S.
 Geol. and Geogr. Surv. Ter., 1877, pp. 18-98.

 WHITMAN CROSS. The Laccolitic Mountain Groups of Colorado, Utah, and
 Arizona, 14th Ann. Rept. U. S. Geol. Surv., 1895, pp. 157-241, pls.

 W. H. WEED and L. V. PIRSSON. Geology and Mineral Resources of the
 Judith Mountains of Montana, 18th Ann. Rept. U. S. Geol. Surv., Pt.
 iii, 1898, pp. 485-556, pl. 75.

 W. H. WEED. Geology of the Little Belt Mountains, Montana, etc., 20th
 Ann. Rept. U. S. Geol. Surv., Pt. iii, 1900, pp. 387-400.

 VERA DE DERWIES. Recherches géologiques et pétrographiques sur les
 loccolithes des environs de Piatigorsk (Caucase du Nord). Geneva,
 1905, pp. 84, pls. 3.

 R. A. DALY. The Mechanics of Igneous Intrusion, Am. Jour. Sci. (4),
 vol. 15, 1903, pp. 269-278; vol. 16, 1903, pp. 107-126.

 JOSEPH BARRELL. Geology of the Marysville Mining District, Montana. A
 study of Igneous Intrusion and Contact Metamorphism. Prof. Pap. 57, U.
 S. Geol. Surv., 1907, pp. 151-178.

Climatic condition in relation to land sculpture:—

 C. E. DUTTON. Tertiary History of the Grand Canyon District, Mon. 2,
 U. S. Geol. Surv., 1882, pp. 264, pls. 42.



Before one may gain a knowledge of rocks or the architecture of
their arrangement within the earth’s crust, it is quite essential
that some familiarity should be acquired with the appearance and
properties of the commonest minerals, and particularly those which
enter as essential constituents into the more abundant rocks. To be a
competent mineralogist, one must have a rather extended knowledge both
of inorganic chemistry and of the science of crystallography, which,
fascinating as it is to study, involves some technical knowledge of
mathematics and much laboratory experience. Though necessary to any one
who contemplates making a career as a geologist, this special study is
not essential to a cultural course like the present one. The attempt
will here be made to bring together a body of fact, from the study of
which the student may quickly learn to recognize the commonest minerals
in their usual varieties. The tests he is to apply are mainly physical,
and in place of an elaborate discussion of crystal symmetry, pictures
only can be supplied.

To the beginner the usual textbook of mineralogy is difficult to
read intelligently, for the reason that for each mineral species it
sets before him a catalogue of each physical property in its turn,
with little indication of those data which in the individual case
have special diagnostic value. None the less, however, the student
is advised to consider the several properties of each mineral in a
definite order, and the following may serve as well as any: crystal or
other form, cleavage, fracture, luster, color, streak, transparency,
tenacity, hardness, magnetism, and specific gravity. In endeavoring to
connect the specific values of these properties with individual mineral
species, the chemical composition and the manner of occurrence are
not to be forgotten. It is well for the student to be supplied with a
small pocket lens and with a pocket knife the blade of which has been

=Crystal form.=—Some mineral species generally occur in more or less
definite crystals—are bounded by definite plane surfaces developed
when the mineral was formed; others in groups of interfering crystals
or aggregates, in which case the mineral is said to be crystalline;
while still others are rarely found crystallized at all. Thus in
a given case crystal form may, or may not, be important for the
diagnosis of the substance. If a mineral species is usually to be
found in crystals, the student should be aware of the fact, and if
possible should have a mental picture of the common crystal shape or
shapes. Without an extended knowledge of crystallography, this must be
supplied him by drawings. Since crystals of most species are apt to be
distorted, owing to the fact that some planes within the same group
appear upon the crystal with a larger development than others, it is
convenient to remember that markings, such as lines or etchings upon
the crystal faces, are the same throughout the same group of planes,
and in the text figures such groups of planes are indicated by the use
of a common letter. For crystalline aggregates such terms as fibrous,
radiating, massive, or granular have their usual meanings.

=Cleavage.=—It is characteristic of most crystals that they break
or _cleave_ along certain directions so as to leave plane or nearly
plane surfaces, and the luster of the cleaved surface measures the
perfection of the cleavage property. It is important always to note
how many such directions of cleavage are present, and, roughly at
least, at what angles they intersect—whether they are perpendicular
to each other or inclined at some other angle. Further, it should be
noted whether a given cleavage is _perfect_, that is, easy, which will
be indicated by the thinness of the plates which can be secured. An
extremely perfect cleavage is possessed by the mineral mica, whose
plates are thinner than the thinnest paper. In the case of imperfect or
interrupted cleavage, the fracture surfaces are not plane throughout,
but interrupted, the surface “jumping” from one plane to a neighboring
parallel one. It is especially important to note whether, in the case
of several cleavages possessed by a crystal, all have the same degree
of perfection, or whether they exhibit differences.

=Fracture.=—In minerals with poorly developed cleavage, the fracture
surface is described as _fracture_. Fracture is thus perfect in
proportion as cleavage is imperfect. The fracture is described as
conchoidal when it shows waving spherical surfaces like broken glass.
For fine aggregates the fracture is described as even, uneven, earthy,
etc., names which are generally intelligible.

=Luster.=—This term is applied especially to the manner in which light
is reflected from mineral surfaces. The most important distinction is
made between those minerals which have a _metallic_ luster and those
which have not, the former being always opaque. Other characteristic
lusters are adamantine (like oiled glass), vitreous (glassy), resinous,
waxy, etc.

=Color.=—For minerals which possess metallic luster the color is
always practically the same, and hence it becomes a valuable diagnostic
property. Of minerals which have nonmetallic luster, the color may be
always the same and hence characteristic, but in the case of many
minerals it ranges between wide limits and sometimes runs almost the
entire gamut of hues, yet without appreciable changes in the chemical
composition of the mineral.

=Streak.=—This term is applied to the color of the mineral powder, and
is usually fairly constant, even when the surface color of different
specimens may vary within wide limits. In the case of fairly soft
minerals the streak is best examined by making a mark on a piece of
unglazed porcelain (streak stone).

=Transparency= (=diaphaneity=).—The terms “transparent”,
“translucent”, “subtranslucent”, and “opaque” are used to describe
decreasing grades of permeability by light rays. Through transparent
bodies print may be read, while translucent bodies allow the light to
be transmitted in considerable quantity through them, though without
rendering the image of objects.

=Tenacity.=—This comprehensive term includes such properties as
brittleness, flexibility, elasticity, malleability, etc.

=Hardness.=—Quite erroneous notions are held concerning the meaning
of this very common word, which properly implies a resistance offered
to abrasion. It is one of the most valuable properties for the quick
determination of minerals, since minerals range from diamond upon the
one hand—the hardest of substances—to talc and graphite, which are
so soft as to be deeply scratched by the thumb nail. For practical
purposes it is sufficient to make use of a rough scale of hardness made
up from common or well-known minerals. If we exclude the gem minerals,
this scale need include but seven numbers, which are: talc, 1; gypsum,
2; calcite, 3; fluor spar, 4; apatite, 5; feldspar, 6; and quartz, 7.
A given mineral is softer than a mineral in the scale when it can be
visibly scratched by a scale mineral, but will not leave a scratch when
the conditions are reversed. If each will scratch the other with equal
readiness, the two minerals have the same hardness.

Since it may often be desirable to test mineral hardness when no scale
is at hand, the following substitutes may be made use of: 1, greasy
feel and easily scratched by the thumb nail; 2, takes a scratch from
the thumb nail, but much less readily; 3, scratched by a copper coin
and very easily by a pocket knife; 4, scratched without difficulty by
a knife; 5, scratched with difficulty by a knife, but easily by window
glass; 6, scratched by window glass; 7, scratches window glass with
readiness, but a grain of sand may be substituted to represent quartz
in the scale.

=Magnetism.=—Though nearly all minerals which contain important
quantities of the elements iron, cobalt, or nickel may be attracted to
a strong electromagnet, there are but two common minerals, and these
of widely different appearance, whose powder is lifted by a common
magnet. Others are, however, lifted after strong heating in the air
(_ignition_), and this is a valuable test.

=Specific gravity.=—Rough tests of relative weight, or specific
gravity, may be made by lifting fair-sized specimens in the hand.
Better determinations require the use of a spring balance.

=Treatment with acid.=—The carbonate minerals react with warm and
dilute mineral acid so as to give a boiling effect (effervescence),
since carbonic acid gas escapes into the air in the process.


The more important common minerals fall into two classes according as
they have large economic importance as ores, or enter in an important
way into the composition of rocks.

I. The Minerals of Economic Importance

=Hematite.=—The sesquioxide of iron, Fe_{2}O_{3}, and by far the most
important ore of iron. Rarely in good crystals, but sometimes in thin
opaque scales bearing some resemblance to mica and known as micaceous
or specular iron ore. At other times in nodules built up from radial
needles (needle ore); in hard masses mixed with fine quartz grains
(hard hematite); or in soft reddish brown earth (soft hematite).
Color, black to cherry red. The powdered mineral always cherry red or
reddish brown, and easily lifted by the magnet after ignition. Hardness
5.5-6.5; specific gravity 5.

=Magnetite.=—The magnetic oxide of iron, Fe_{3}O_{4}, often in
crystals like Fig. 486, ^{1-2}. Black and opaque with a metallic
luster. Streak black. Lifted by a magnet and sometimes itself capable
of lifting filings of soft iron (lodestone). Hardness 5.5-6.5. Specific
gravity 5.

=Limonite.=—The most abundant and most valuable of the hydrated iron
ores, 2 Fe_{2}O_{3}. 3 H_{2}O. Chemical composition the same as iron
rust, with which in the earthy form it is identical. Never in crystals,
but often in mammillary or rounded pendant forms resembling icicles,
or sometimes clusters of grapes. Its yellow (rust) streak is its best
diagnostic property. Ignited it gives off water and becomes magnetic.
The streak and its notably lower specific gravity distinguish it from
certain forms of hematite which it outwardly resembles. Hardness 5-5.5.
Specific gravity 3.6-4.

=Pyrite, iron pyrites, or “fool’s gold.”=—The sulphide of iron,
FeS_{2}. The most widely distributed sulphide mineral and now a chief
source of the great chemical reagent, sulphuric acid or vitriol.
Often, but not always, in crystals (Fig. 486, ^{3-5}) which have
peculiar striæ upon their faces. At other times the mineral is found
massive or in radiated needles. Bright metallic luster with the color
of new brass, though often tarnished or altered upon the surface to
limonite. Hard and brittle, and so distinguished from gold, which is
soft and malleable and of the color of the paler old brass (which
contained a larger percentage of zinc). Gold is, further, about four
times as heavy as pyrite. Hardness 6-6.5. Specific gravity 5.

=Chalcopyrite, copper pyrites.=—A mixed sulphide of copper and iron.
If in crystals, like Fig. 486, ^6; otherwise massive or compact.
Luster metallic. Color orange-yellow, often with local blue and green
iridescence like a pigeon’s throat. Distinguished from pyrite by the
deeper color and lower hardness, and from gold, particularly, by its
brittleness and lower specific gravity. Hardness 3.5-4. Specific
gravity 4.

=Galenite, galena.=—Sulphide of lead, PbS. The chief ore of lead, and,
from admixture of a silver mineral, of silver as well. Usually found in
crystals (Fig. 486, ^7). Always cleaves into blocks bounded by six very
perfect rectangular faces which, when freshly broken, show a bright
silvery luster and quickly tarnish to a peculiarly “leaden” surface.
Very heavy. Color and streak lead-gray. Hardness 2.5. Specific gravity

=Sphalerite, zinc blende.=—Sulphide of zinc, ZnS, usually with
considerable admixture of sulphide of iron. The great ore of zinc.
Not infrequently in crystals (Fig. 486, ^{8-9}), but more often in
cleavable crystalline aggregates. The cleavage in fine aggregates is
sometimes difficult to make out, but in coarse-grained masses it is
seen to be equally and highly perfect in six different directions,
so that a symmetrical twelve-faced form may sometimes be broken out
(dodecahedron). Luster like that of rosin (rosin jack), though when
with large iron admixture the color may approach black (black jack).
The lighter colored varieties are translucent. Hardness 3.5-4. Specific
gravity 4.

=Malachite.=—Hydrated (basic) copper carbonate. The green copper ore
and the common surface alteration product of other copper minerals.
Usually has a microscopic structure made up of fine needle-like
crystals, but generally massive in various imitative shapes not unlike
those of the iron ores. Sometimes earthy. Its color is bright green,
and it is usually found in association with other characteristic copper
ores, such as chalcopyrite and azurite. When relatively pure and in
large masses, it is a beautiful ornamental stone. Effervesces with
acid. Hardness 3.5-4. Specific gravity 4.

[Illustration: FIG. 486.—Forms of Crystals: 1-2, magnetite; 3-5,
pyrite; 6, chalcopyrite; 7, galenite; 8-9, sphalerite; 10-13, calcite.]

=Azurite.=—Hydrated (basic) copper carbonate, less hydrated than
malachite, and known as the blue carbonate of copper. Generally in very
minute and quite complex crystals, but also in imitative shapes similar
to those of malachite, and at other times earthy. Slightly lighter in
weight than malachite, from which it is easily distinguished, as from
most other minerals, by its bright azure blue color and its somewhat
lighter blue streak. Effervesces with nitric acid. Hardness 3.5-4.
Specific gravity 3.7-3.8.

=Calcite.=—Calcium carbonate, CaCO_{3}. Almost always in crystals
(Fig. 486, ^{10-13}), or in confused crystal aggregates, though
rarely fibrous or dull and earthy. Some of the forms of the crystals
are described as “dog-tooth spar”, others as “nail-head spar”, while
still others are modified hexagonal prisms. There is a beautifully
perfect cleavage of the mineral along three directions which make
angles of about 105° with each other, so that under the hammer the
substance breaks into blocks which are shaped like the crystal of Fig.
486, ^{10}. Usually white or gray, but occasionally faintly tinted.
Streak white. Effervesces with cold and dilute mineral acids. An
associate of many ores and the chief mineral of limestone. A similar
mineral—dolomite—contains in addition magnesium carbonate, has
simpler crystals (like the drawing of Fig. 486, ^{10}, but often with
rounded faces), and effervesces only when the acid is warmed. Hardness
3. Specific gravity 2.7.

=Gypsum.=—Hydrated calcium sulphate, CaSO_{4}.2 H_{2}O, and the
source of plaster of Paris. Often in simple crystals (Fig. 487, ^1)
or else “swallow tail”, like Fig. 487, ^2; in which case the mineral
is generally either transparent or translucent and is described as
selenite. Such crystals show a cleavage approaching in perfection that
of the micas, but, unlike the mica laminæ, those produced by cleavage
in gypsum though flexible are not elastic. There are also fibrous forms
of gypsum (satin spar), a fine-grained form (alabaster), and the impure
earthy form (rock gypsum). Very soft, light in weight, and difficultly
fusible. Color usually white, gray, or pale yellow. Hardness 2.
Specific gravity 2.3.

=Copper glance.=—A sulphide of copper, Cu_{2}S. Not usually well
crystallized, but generally massive and associated or variously admixed
with other copper ores such as chalcopyrite, malachite, etc. Fracture
conchoidal, luster metallic, color and streak blackish lead-gray,
though often tarnished blue or green from surface alterations to the
copper carbonates. Softer and heavier than chalcopyrite. Blowpipe or
chemical tests are necessary for its identification. Hardness 2.5-3.
Specific gravity 5.5-5.8.

=Cerussite.=—The white or carbonate lead ore, PbCO_{3}, and an
important ore of silver as well. Often in crystals of considerable
complexity, though Fig. 487, ^{3-4}, shows some common shapes. Often
granular, massive, or earthy (gray carbonate ore). Very brittle and
with conchoidal fracture. The luster is adamantine or like that of
oiled glass. Color generally white or gray. Very heavy, the heaviest
of light colored and nonmetallic minerals. Dissolves in nitric acid
with effervescence. Hardness 3-3.5. Specific gravity 6.5.

=Siderite.=—The carbonate or “spathic” ore of iron, FeCO_{3}. Either
in crystals resembling in form Fig. 486, ^{10}, but with rounded faces,
or cleavable massive to finely granular and earthy. The crystalline
varieties cleave easily into smaller blocks of the same form as those
of calcite. Color usually gray or brown and streak white. On strongly
igniting, the white powder becomes black and magnetic. Lighter in both
color and weight than the other iron ores, and unlike them siderite
effervesces with acid. Distinguished from calcite by its higher
specific gravity and its change upon being ignited. Hardness 3.5-4.
Specific gravity 3.9.

=Smithsonite.=—Carbonate of zinc, ZnCO_{3}, and an important ore of
that metal. Seldom found in crystals except as a replacement of calcite
crystals, in which case it shows the forms characteristic of the latter
mineral. Usually kidney-shaped, stalactitic, or else in incrustations
upon other minerals. Sometimes granular or earthy. Brittle. Luster
vitreous, color white or greenish gray, though often stained yellow
with iron rust. Streak white except when the mineral is stained with
iron. Effervesces with warm acid. Hardness 5. Specific gravity 4.4.

=Pyrolusite.=—Black oxide of manganese, MnO_{2}, though generally
impure from admixture with other manganese oxides. Usually in intricate
aggregates which may be columnar, fibrous, mammillary, earthy, etc.
Opaque, with color and streak both black. Soft and easily soils
the fingers. With hydrochloric acid gives off the choking fumes of
chlorine. Hardness 2-2.5. Specific gravity 4.8.

II. The Minerals important as Rock Makers

These minerals are in most cases complex silicates of one or more of
a certain number of metals such as aluminium, calcium, magnesium,
iron, sodium, potassium, or hydroxyl (OH). For their identification an
examination of the physical properties is usually sufficient, whereas
of the typical ore minerals already considered, additional chemical
tests may be necessary.

=Feldspars.=—A group of similar alumino-silicates of potassium,
sodium, and calcium. The most important of all rock-making minerals.
Although with wide variation in chemical composition, the feldspars are
yet broadly divided into two classes; the one striated, and the other
an unstriated potash or orthoclase variety. The pocket lens is usually
necessary in order to make out the striations upon the crystal or
cleavage surfaces. When formed in veins, feldspar appears in crystals
(Fig. 487, ^{5-6}), but as a rock constituent the mutual interference
of crystals prevents the development of bounding faces. Two cleavage
directions, nearly or quite perpendicular to each other, are notably
different in their perfection. Hard enough to scratch glass, but easily
scratched by sand. Color pink (usually orthoclase or microline), white
(often albite) to gray. Sometimes with beautiful “pigeon’s throat”
effect of iridescence (labradorite). Low specific gravity. Hardness 6.
Specific gravity 2.5-2.8.


FIG. 487.—Forms of Crystals: 1-2, gypsum; 3-4, cerussite; 5-6,
feldspar; 7, quartz; 8, pyroxene (cross section); 9, hornblende
(cross section); 10, garnet; 11, nephelite; 12-14, staurolite; 15-16,
tourmaline (cross sections); 17, olivine.]

=Quartz.=—Oxide of silicon or silica, SiO_{2}. Both an important vein
mineral associated with the ores and a rock maker. In the former case
particularly, often in crystals of notably simple forms (Fig. 487, ^7).
Few minerals which are not gems are so hard. Remarkable freedom from
cleavage so that the mineral breaks much like window glass—conchoidal
fracture. Wide range in both transparency and color. Transparent and
colorless crystalline variety (rock crystal), brown translucent (smoky
quartz), turbid white (milky quartz), and various colored varieties
(carnelian, jasper, jet, etc.). Insoluble in acids and infusible.
Hardness 7. Specific gravity 2.6.

=Micas.=—Like the feldspars a group of complex silicates, but here
chiefly of potassium, magnesium, iron, and hydroxyl. Abundant as rock
makers, the micas are all characterized by the thinnest and toughest of
elastic cleavage plates, such as are generally known as isinglass. When
a needle is driven sharply through a thin scale of mica, a six-rayed
puncture star forms about the needle point. The darker common variety
of mica is rich in iron and magnesium and is called biotite, and the
lighter colored alkaline variety, muscovite. Hardness 2.5-3.1. Specific
gravity 2.7-3.1.

=Chlorite.=—Generally an intricate mixture of more or less similar
microscopic crystals having varying and rather complex chemical
compositions and related to the micas, but all characterized by
a peculiar leaf green color. These minerals are a common product
of hydration weathering in rocks which are rich in magnesium and
iron—especially those that contain biotite, pyroxene, or hornblende
(see below). Hardness 1-2.5. Specific gravity 2.5-3.

=Pyroxenes.=—An important group of related rock-making minerals all
of which are silicates of the bases magnesium, calcium, aluminium,
iron, and manganese. Quite generally developed either in columnar or
needle-like crystals which are uniformly shaped in cross section like
Fig. 487, ^8. Two rather imperfect cleavages are directed parallel
to the longer axis of the crystal and nearly at right angles to each
other. The colors of all but the lime varieties are dark and generally
green, dark brown, bronze, or black. The lime varieties are white,
gray, or pale green. A dark colored and common iron variety is known as
augite. Streak generally either white or lightly tinted. Hardness 5-6.
Specific gravity 3.2-3.6.

=Amphiboles.=—A group of minerals of the same chemical composition
as the pyroxenes, with which also in most physical properties they
agree. The principal distinction is found in the shape of the cross
section and in the cleavage (Fig. 487, ^9). Whereas the cross sections
of pyroxenes are generally eight sided, those of the amphiboles have
six sides, and whereas the cleavage directions of pyroxenes are nearly
at right angles to each other (87°), the similar but much more perfect
cleavage directions of the amphiboles are inclined at an obtuse angle
(124½°). Owing to the obliquity of the amphibole cleavage, fractured
surfaces of the mineral appear splintery, which is not in the same
measure true of the pyroxenes. A fibrous variety of amphibole, and
occasionally other varieties of the mineral, is a not uncommon product
of weathering of pyroxenes. Other physical properties of the amphiboles
are in the main almost identical with those of the pyroxenes.

=Garnet.=—Complex alumino-silicates or ferro-silicates of calcium,
magnesium, iron, or manganese, or several of these combined. Nearly
always in crystals, and usually found in mica schist (see below).
The crystals usually have twelve similar faces, each a lozenge
(dodecahedron), or else twenty-four similar faces, or the two forms
combined (Fig. 487, ^{10}). Brittle. From any but the gem minerals garnet
is easily distinguished by its hardness, which in different varieties
ranges from somewhat below to somewhat above that of quartz. The luster
is vitreous, and the color runs the gamut of reds, browns, and greens,
but with the common hue dark red to black. Streak white. Hardness
6.5-7.5. Specific gravity 3.1-4.3.

=Nephelite= (=nephelene=).—An alumino-silicate of sodium and
potassium. In certain special provinces this mineral is developed in
abundance as an essential constituent of igneous rocks, but elsewhere
practically unknown. The rare crystals are hexagonal prisms (Fig.
487, ^{11}), but the mineral is most easily determined by its general
resemblance to feldspar, but with the differences of cleavage, luster,
and reaction with acid. Whereas the feldspars have two cleavages,
either nearly or quite perpendicular to each other and of different
degrees of perfection, nephelite has three equal cleavages inclined
60° and 120° to each other and of less perfection than either feldspar
cleavage. The luster of nephelite is perhaps the best clew to its
identity, since this is greasy and simulated by but few minerals.
The fine powder of the mineral treated for some time with strong
hydrochloric acid forms a perfect jelly of silicic acid, whereas
the feldspars do not. Though itself gray or white and unobtrusive,
nephelite is usually associated with brightly colored minerals, which
are often the first clew to its presence in a rock. Hardness 5.5-6.
Specific gravity 2.5-2.6.

=Talc= (=soapstone=).—A silicate of magnesium and hydroxyl which is
an important alteration product through weathering of certain pyroxene
rocks especially. Usually a foliated mass, this product is occasionally
fibrous or even granular. Talc is one of the softest of minerals,
having a greasy feel and being easily scratched with the thumb nail.
The luster of the foliated varieties is apt to be pearly, and the
color apple-green to white, though sometimes stained brown from oxide
of iron. The streak of the mineral is white except when stained by
iron. Although the rocks which are composed mainly of talc (soapstone)
are exceedingly soft, they are very tough and remarkably resistant.
Hardness 1-1.5. Specific gravity 2.7-2.8.

=Serpentine.=—Like talc, serpentine is a silicate of magnesium and
hydroxyl, and an important product of the breaking down of magnesium
minerals in the process of weathering. The mineral is usually found
as a fine web of microscopic needle-like fibers, and is best roughly
diagnosed by its color and its associated minerals. Like talc it is
usually developed within those igneous rocks from which feldspar is
lacking, but where either pyroxene or olivine is found in abundance or
was previous to alteration. The characteristic color of serpentine is
leek-green. The rock largely composed of serpentine is called by the
same name, and being exceedingly tough and unchanging is, in spite of
its softness, a valuable building and ornamental stone. A red magnesium
garnet is apt to be associated with such serpentine masses. Hardness
2.5-4, because of impurities. Specific gravity 2.5-2.6.

=Staurolite.=—A silicate of aluminium, iron, and hydroxyl. Found
in metamorphic rocks usually in association with garnet. Always in
crystals bounded by simple forms generally crossed, as shown in Fig.
487, ^{12-14}. The color is dark reddish brown, and the streak is
colorless to grayish. The hardness is exceptional and higher than that
of quartz. Hardness 7-7.5. Specific gravity 3.6-3.7.

=Tourmaline.=—An exceptionally complex silicate of boron and aluminium
as well as iron, magnesium, and the alkalies. Found in metamorphic
rocks and always crystallized. The crystals are columns or needles
whose cross section is the best guide to their identity, since this
is a modified triangle unlike that of any other mineral (Fig. 487,
^{15-16}). Additional diagnostic properties are the characteristic
striations which run lengthwise of the crystals upon prism faces, and
the lack of any cleavage (difference from hornblende). The hardness is
also a valuable property, since this is greater than that of quartz.
The mineral is brittle and the fracture subconchoidal. The range in
color is as great as, or greater than, that of garnet, though the
common forms are jet black. Streak uncolored. Hardness 7-7.5. Specific
gravity 3-3.2.

=Olivine.=—A silicate of magnesium and iron and a rock-making mineral
found only in those igneous rocks which have little or no feldspar. It
easily suffers alteration by weathering and passes into serpentine,
and in fact is seldom found except when at least partially altered to
the fibrous webs of that mineral. The form of the unaltered crystals
within the rocks is shown in Fig. 487, ^{17}, and, cut in sections, the
mineral appears in more or less elongated hexagons. The hardness of
the unaltered mineral is about that of quartz. It has rather imperfect
cleavages in two rectangular directions, and is usually translucent,
with a vitreous luster and a color which is olive-green when not
stained brown by oxide of iron. Streak uncolored. Hardness 6.5-7.
Specific gravity 3.2-3.3.



In Chapter IV the classification and the structure of rocks have been
briefly discussed. Below are added brief descriptions of the more
important common rocks. For rocks as for minerals it is, however,
essential that a collection of well-chosen specimens be studied for
purposes of comparison. A small pocket lens is a valuable aid in making
out the component minerals and the textures of the finer grained rocks.

1. Intrusive Rocks

=Granite.=—Of granitic texture, though sometimes porphyritic as well.
The most abundant mineral constituent is a pink or white feldspar,
usually without visible striations, with which there is usually in
subordinate quantity a white striated feldspar. Next in importance to
the feldspar is quartz, which because of its lack of cleavage shows a
peculiar gray surface resembling wet sugar. In addition to feldspar
and quartz there is generally, though not universally, a dark colored
mineral, either mica or hornblende. The mica is usually biotite, though
often associated with muscovite.

=Syenite.=—Like granite, but without quartz, with more striated
feldspar, and generally also the rock has a darker average tint.
While biotite is the commonest dark colored constituent of granite,
hornblende is more apt to take its place in syenite. Less common than
granite, to which it is closely related in origin and in composition.

=Gabbro.=—A dark colored rock of granitic texture composed of striated
feldspar with broad cleavage surfaces and usually an abundance of
pyroxene. In contrast to the feldspars of granite, those of gabbroes
are often dull and colored grayish yellow or greenish. The pyroxene
is often in part changed to fibrous amphibole. Magnetite may be an
abundant accessory mineral.

=Diabase.=—In color dark like gabbro, and of similar constitution. In
diabase, however, the feldspar crystals, instead of being broad and of
irregularly interrupted outline, are relatively long (“lath-shaped”),
and the pyroxene acts as a filler of the residual space between them.

=Peridotite.=—A heavy and dark colored rock of granitic texture which
is nearly or quite devoid of feldspar but contains olivine. When
altered, as it generally is, it is largely a mass of serpentine, talc,
and chlorite, surrounding cores, it may be, of still unaltered pyroxene
and olivine. Magnetite is an abundant constituent, and a red garnet is
apt to be present.

2. Extrusive Rocks

=Obsidian.=—A rock glass rich in silica. It is usually black and
breaks with a perfect conchoidal fracture. It often passes over through
insensible gradations into pumice, which differs only in its vesicular
structure. As regards chemical composition, obsidian and pumice are not
notably different from rhyolite (below).

=Rhyolite.=—A light colored rock of porphyritic texture, often also
with fluxion or spherulitic textures, or both combined. The porphyritic
appearance is given the rock by large crystals of a glassy, unstriated
feldspar and crystals of quartz. Rhyolite is a very siliceous lava
containing rather more silica than granite, to which of the intrusive
rocks it is most closely related, and from which it differs in its
texture and in the manner of its occurrence in nature. Whereas
granite is found in great batholites, laccolites, and bysmalites,
and consolidated in most cases beneath the earth’s surface, rhyolite
generally occurs in sheets, flows, or dikes, and consolidated either
above or in fissures near to the surface.

=Trachyte.=—Similar to rhyolite, but usually with a peculiar gray
aspect from the greater abundance of feldspar crystals. The rock
is less siliceous than rhyolite, contains no quartz crystals, and
approaches a feldspar in its average composition.

=Andesite.=—Similar to rhyolite in appearance and in origin, but more
basic and correspondingly dark in color. The porphyritic crystals
are of lath-shaped, striated feldspar, with which are associated
crystals of either biotite or hornblende or both. A fluxion texture is
particularly characteristic of this type of extrusive rock.

=Basalt.=—A dark colored or black basic rock of porphyritic texture
which differs but little from diabase. It may show under the lens fine
lath-shaped crystals of striated feldspar associated with crystals
of augite, but more frequently the rock is dense and without visible
mineral constituents. It is particularly likely to occur divided up
into columns six inches to a foot in diameter and known as basaltic
columns. Especially fine examples are known from the Giant’s Causeway
and other localities in the western British Isles.

3. Sedimentary Rocks of Mechanical Origin

=Conglomerate= (“=pudding stone=”).—A rock made up from pebbles which
are cemented together with sand and finer materials. The pebbles are
usually worn by work of the waves upon a shore, and may vary in size
from a pea to large bowlders. They may consist of almost any hard
mineral or rock, though the sand about them is largely quartz.

=Sandstone.=—A rock composed of sand cemented together either by
calcareous, siliceous, or ferruginous materials. Sandstones are
described as friable when their surface grains are easily rubbed off,
or as compact when they are more firmly cemented. Sandstones are often
distinctly banded and are sometimes variously stained with oxide of
iron. Those sandstones which have been formed upon a seacoast are known
as marine sandstones, while those derived from accumulations collected
by the wind in deserts are distinguished as continental deposits.
Sandstones form much thicker formations than conglomerates, the latter
usually constituting a basal layer only of the sandstone formation
(basal conglomerate).

=Shale.=—A consolidated mud stone which is probably the most abundant
rock formation. In large part clay admixed in varying proportions with
extremely fine sandy grains.

4. Sedimentary Rocks of Chemical Precipitation

=Calcareous tufa= (=travertine=).—Not to be confused with tuff,
which is a fragmental extrusive or volcanic rock. Calcareous tufa is
formed when waters which contain carbonic acid gas and lime carbonate
in solution, give off the gas and with it the power to hold the lime
in solution. Such a liberation of the gas may occur when the stream
is dashed into spray above a cascade, and the lime is then deposited
about the site of the falls. Travertine is generally porous and formed
of more or less concentric layers or incrustations. A remarkable
illustration is furnished by the travertine deposits of Tivoli and
other localities near Rome, since here the material supplies a valuable
building stone.

=Oölitic limestone= (=oolite=).—This rock is made up of spherical
nodules and so has the appearance of fish roe. Broken apart, each grain
reveals in its center a core of siliceous sand about which carbonate
of lime has been deposited in concentric layers. It is thought that
waters charged with carbonate of lime, in issuing from a river near
a sea beach, coat the sand grains of the latter with successive thin
films of lime carbonate due to the rhythmic ebb and flow of the tides,
evaporation of the adhering water taking place when the sands are
exposed at low tide.

5. Sedimentary Rocks of Organic Origin

=Limestone.=—A generally white or gray rock composed of carbonate of
lime with varying proportions of clay, silica, and other impurities.
The lime carbonate is usually derived from the hard parts of marine
organisms, and the argillaceous and siliceous impurities from the finer
land-derived sediments which descend with them to the bottom.

=Dolomite= (=dolomitic or magnesium limestone=).—Differs from
limestone in containing varying proportions of the mineral dolomite
(_ante_, p. 455), which is made up of equal parts of calcium and
magnesium carbonates. Difficult to distinguish from limestone unless a
chemical test is made for magnesium, though it may be said in general
that dolomite is less soluble in cold mineral acids.

=Peat.=—An accumulation of decomposed vegetable matter within
small lakes and in lagoons separated from larger ones (_ante_, p.
429). Peat represents the first stage in the formation of coal
from vegetable matter, and differs from the coals by its larger
proportion of contained water. Because of this water its fuel value is
correspondingly small. It is usually dark brown or black and reveals
something of the structure of the plants out of which it was formed.

6. Metamorphic Rocks

=Gneiss.=—A generally more or less banded (gneissic) metamorphic rock
with a mineral constitution similar to granite, and often developed
by metamorphic processes from that rock. It may at other times, by
processes not essentially different, be derived from sedimentary
formations. It usually contains as important constituents unstriated
feldspar and quartz, but in addition it may include a striated
feldspar, biotite, muscovite, or hornblende, or several of these
combined. In proportion as mica or hornblende is abundant, it has a
marked banded texture, but it differs from mica schist (see below) not
only in the presence of its feldspar, but in the smaller proportion
of mica. Biotite gneiss, hornblende gneiss, etc., are terms used to
designate varieties in which one or the other of the dark colored
constituents predominate.

=Mica schist.=—A metamorphic rock without feldspar and mainly composed
of quartz and light colored mica (muscovite). The abundant mica lends
to the rock its characteristic schistose texture, which differs from
the usual gneissic texture. In some cases the mica is wrapped about
the grains of quartz, but at other times it forms a series of almost
continuous membranes separating layers of quartz.

=Sericite schist.=—A variety of schist which is characterized by
an abundance of a peculiar silvery mica rich in the element group
hydroxyl. The mica scales are often microscopic and wrought into an
intricate web with the quartz constituent.

=Talc schist.=—A schist made up largely of talc, but with varying
proportions of quartz, magnetite, etc. From the abundance of the talc
it is usually pale green or white.

=Chlorite schist.=—A greenish, fine-grained metamorphic rock in which
chlorite is the principal mineral, but in which magnetite is a quite
characteristic accessory constituent.

=Staurolitic garnetiferous mica schist.=—A mica schist in which garnet
and staurolite are so abundant as to be essential constituents.

=Clay slate.=—A metamorphosed mud stone or shale. In the process of
metamorphism the rock has been hardened, given a slaty cleavage, and
innumerable minute scales of mica have developed to produce a silky
luster upon the cleavage faces. The color may be gray, green, purple,
or black.

=Quartzite.=—A metamorphosed sandstone in which the sand grains have
become enlarged by accretion of silica. Whereas a sandstone fractures
about its constituent grains, a break in quartzite is continued through
the grains and the cement alike. In contrast to sandstones, the
quartzites derived from them are usually lighter in color and often
nearly white.

=Marble= (=crystalline limestone=).—The result of metamorphism upon
limestones. Usually white in color but sometimes gray, blue gray, or
yellow, and sometimes variously broken or brecciated and stained with
iron oxide. Effervesces with cold dilute acid.

=Coals.=—Under the head of peat the first stage in the formation of
coals from vegetable matter has been briefly described. Lignite, or
brown coal, represents a further stage and one in which the vegetable
structure is still recognizable. It is usually brownish black or
black in color and contains a considerable proportion of water. With
increased pressure or dynamic metamorphism, further percentages of the
volatile constituents are eliminated, and when from seventy-five to
ninety per cent of carbon remains, the material burns with a yellow
flame and is known as bituminous coal. This is the great fuel for
the production of steam. A continuation of the metamorphic processes
carries off a further proportion of the volatile matter and leaves a
dense, hard, black substance with sometimes as much as ninety-five
per cent of carbon. This is the so-called “hard coal” or anthracite
generally used for fuel in our houses, for which purpose it is so well
adapted because it burns with a production of much heat and almost
without smoke.



=Topographical maps a library of physiography.=—For the satisfactory
working out in detail of the geology of any region of complex
structure, an accurate topographical map is prerequisite. This is so
much the more true because nearly all complexly folded or faulted rock
masses are to be found in mountainous, or at least in hilly regions.
The making of the topographical map must, therefore, precede that of
the geological map, and in modern usage the latter is a topographical
and a geological map combined in one.

Within certain narrow limits, predictions concerning the geological
history of a province may often be made by an expert geologist from
examination of an accurate topographical map. Just as in forecasting
the weather upon the basis of the usual weather maps, such predictions
can sometimes be made with entire confidence in their accuracy,
while at other times a guess only may be hazarded. The great value
of the modern topographical map is becoming, however, universally
acknowledged, and every highly civilized nation has either completed
or has in preparation sectional topographical maps of its domain on
such a scale as is warranted by its financial condition and its state
of development. Thus there is now being accumulated a vast library of
geographical and to some extent geological information, of which the
student of geology must be prepared to make use.

=The nature of a contour map.=—More and more the contour map is
replacing the earlier and less scientific methods of representing
topography on the large scale sectional maps, and hence this type only
need here be considered. In the contour map, the relief of the land
is represented by a series of curving lines, each the intersection of
a particular horizontal plane with the land surface, and the several
planes separated by uniform differences of elevation. This altitude
interval is known as the contour interval. Its choice is a matter
of considerable importance, for though regions of relatively simple
topography may be adequately represented upon a map of large contour
interval, say one hundred feet, another district may require an
interval as short as five feet. A contour map with this interval may be
conceived to have been made by flooding the region which it represents
and preparing maps of the shore lines for each rise of five feet of the
water surface, and superimposing the several maps thus derived with
accurate registration one above the other. Wherever the land slopes
are steep, the shore lines of the several maps will be crowded closely
together and give the effect of a relatively dark local shade; where,
upon the other hand, the surface is relatively flat, the several shores
will be widely spaced and the effect will be to produce a white area
upon the map. Thus in contour maps dark tones indicate steep gradients
and pale tones a flatness of surface.

=The selection of scale and contour interval.=—With the use of
the small scale in the contour map, the tones of the map will be
correspondingly dark, though the relative differences in tone will
remain the same. With the use of a closer contour interval the tones
will deepen throughout. The adjustment of scale and contour interval
to any given region is a matter requiring experience in topographical
mapping, and in addition a knowledge of the geological significance of
topographic features. Unfortunately, the element of expense and the
special commercial objects held in view, conspire to select scales
and contour intervals which are often little adapted to the districts

=The method of preparing a topographical map.=—Having fixed upon the
scale and the contour interval which is to be employed, the task of
the topographical surveyor is next to fix accurately the positions and
the elevations of a sufficient number of points to _control_ the map,
and then to hang, as it were, upon these points as attachments the
design represented by the relief. Were the surface of the ground to
be represented by a flexible fabric, the map maker might raise from a
flat base a series of stout posts of the heights and in the positions
which he has determined, and upon these supports arrange the slopes of
the fabric much as drapery is adjusted. The determination of the exact
positions and the elevations of his control stations is, therefore,
a process coldly precise and formal; whereas in the shaping of the
surfaces his attention should be fixed more upon correctly reproducing
the shapes than upon fixing accurately the position of every point.
As a matter of fact, the position of the average point will be most
accurately fixed when the shapes of the features are most clearly
comprehended. To some extent, therefore, the topographer should be
familiar with the geological significance of the earth features which
he is representing.

=Laboratory exercises in the preparation of topographical maps.=—The
principles which underlie the surveyor’s method for preparing a
topographical map may be learned in the laboratory by the use of
models and the simple device shown in plate 24 A and B. To represent
the section of country to be mapped a model in plaster of Paris is
substituted, and this is placed within a rectangular tank to which
locating carriages and altitude gauges are attached that allow the
student to fix the position and the elevation of any point upon the
surface of the model.

│   PLATE 24.                                                      │
│                                                                  │
│ [Illustration: _A._ Apparatus for exercise in the preparation of │
│ topographic maps.]                                               │
│                                                                  │
│ [Illustration: _B._ The same apparatus in use for testing the    │
│ contours of a map.]                                              │
│                                                                  │
│ [Illustration: _C._ Modeling apparatus in use.]                  │

Upon each model the student “locates”, or fixes, the position of a
sufficient number of points for the control of his map, entering upon
an appropriate map base for each position the altitude which was read
from the gauges. Now _with the map always before him_ he “sketches in”
the forms of the surface by means of contour lines. For this purpose
it is often desirable to fix roughly the direction of the steepest
slope at a number of places, and noting the differences in elevation
between control stations, divide up the distance in accordance with the
curves of slope and start the contours at right angles to the slope.
Afterwards such sections are connected by sketching in with the model
always in view for control (Fig. 488).

[Illustration: FIG. 488.—A student’s map prepared from a model by the
use of the contour apparatus represented in plate 24 A.]

=The verification of the map.=—The map prepared, its accuracy may be
tested by a simple method which is denied the topographer who has to
do with the actual surface of the ground. The locating carriages and
altitude gauges are removed from the tank, which is next filled with
water and leveled by means of guide marks upon the interior. A few
drops of milk or of ordinary clothes blueing are added to the water to
render it opaque, and it is then drawn off at the faucet in successive
installments, so that the surface drops by layers corresponding in
thickness to the contour interval of the map, plate 24 B. As each layer
is withdrawn, that contour of the map to which the shore line should
correspond is carefully examined and corrected. By such corrections
the nature of the first errors made is soon appreciated, and the
method of procedure is thus more easily acquired. At the same time the
significance of the design of the map is more quickly learned than by a
mere examination of the standard government maps.

The work above outlined calls for waterproofed models of suitable form
and size, and a series, each of which sets forth some typical feature
or series of features, has been designed by Mr. Irving D. Scott.[2]

=The preparation of physiographic models.=—The apparatus used
to prepare the topographic map is adapted also for preparing a
physiographic model from a standard topographical map. For this purpose
the method is essentially reversed, though the tank is replaced to
advantage by a light metal frame elevated upon one side so as to permit
a free use of the hands in modeling the clay.

The material used in preparing the model is artists’ modeling clay[3]
which has a base of beef suet, and hence does not dry out and crack as
does ordinary clay. Its form is, therefore, retained indefinitely, and
it may be used again and again. Most maps must be enlarged in modeling,
and the simplest way is often to photographically or by pantograph
enlarge the map to the scale of the model. The map prepared, it is
covered by a thin celluloid plate which has cut upon it a series of
crossed lines spaced in inches and larger subdivisions to correspond to
those of the locating carriages (plate 24 C).

The enlargement of the map is not essential to experienced workers,
and the standard map may be covered in similar manner by a transparent
plate with “checkerboard” design, the squares of which bear some simple
relation in size to the larger divisions of the locating carriages
(Plate 24 C, rear).

The method of preparing the model is comparatively simple. Beginning at
any point upon the map, the intersection of a heavy contour line with
one of the guide lines of the celluloid “position plate” is carefully
noted. Both the position and the elevation of this point are fixed by
the point of the altitude gauge of the modeling frame, and the clay
built up beneath it to that height. With the fingers the clay is now
roughly shaped in various directions from this point, the altitude
gauge is advanced by the locating carriage so as to correspond in
position to the intersection of the next heavy contour line with the
same guide line of the position plate, and the elevation for this point
similarly adjusted upon the model. As before, the surface of the clay
is roughly shaped in advance and upon the sides so as to conform to
the indications of the map; and this process is repeated until the
work is finished. Corrections for intermediate positions may be
carried to any desired degree of refinement which the scale and the
accuracy of the map permit. Models which are larger than the area of
the modeling frame are prepared by making a square foot at a time by
the above described process, and then moving the frame forward and
adjusting in a new position by means of the sharp pins in the legs
of the apparatus.


 WILLIAM H. HOBBS, New Laboratory Methods for Instruction in Geography,
 Journal of Geography, vol. 7, 1909, pp. 97-104. Also Scot. Geogr.
 Mag., vol. 24, 1908, pp. 643-652. The Modeling of Physiographic Forms
 in the Laboratory, _ibid._, vol. 8, 1910, pp. 225-228.



[Illustration: FIG. 489.—Models to represent outcrops of rock.]

The laboratory models which have been described on page 63, and are
used to represent outcrops in the study of geological maps, are shown
in Fig. 489. The drum-shaped blocks serve to represent massive rocks
which occur in irregularly shaped masses such as batholites and flows.
The long, narrow strips are for intrusive rocks in the form of dikes,
while the larger blocks provided with a swivel joint are used for
outcrops of sedimentary rocks, and after adjustment they give the dip
and strike of the exposure. The wing bolts used in their construction
should be of bronze, because of the effect of iron upon the compass.
For the same reason tables should not be placed near iron beams or
columns. All these blocks can be made by an ordinary carpenter, and
should be available in sufficient numbers to arrange problems like
those of Figs. 47, 48, and 490. With a view to supplying suggestions
for other problems of the same general nature, the three additional
field maps of Fig. 491 have been introduced.

[Illustration: FIG. 490.—Special laboratory table set with a problem
in geological mapping which is solved in Figs. 47 and 48.]

[Illustration: FIG. 491.—Three field maps to be used as suggestions in
arranging laboratory tables for problems in the preparation of areal
geological maps.]

The list of questions given below is intended to indicate the nature of
some of the problems which the student should be asked to solve in the
preparation of each map. The numbers in parentheses refer to pages in
this book where further information is given:—


 1. Of the formations represented what ones are sedimentary and what
 igneous (Chap. IV, App. B)?

 2. Which formations, if any, are separated by unconformities (51-53)?

 3. What is the order of age of the sedimentary formations (65)?

 4. What are the _exposed_ thicknesses of each of these formations

 5. Do any of these values represent full thickness of the formation,
 and if so, which ones?

 6. What is the age in terms of the sedimentary formations of each of
 the igneous rock masses (65)?

 7. Which igneous rocks, if any, occur in batholites (143, 441)? Which,
 if any, in dikes (140)?


 8. What formations, if any, have monoclinal dip (42)?

 9. Indicate upon the map by dashed lines the crests of all anticlines
 and the trough lines of synclines.

 10. Indicate by arrows the direction of pitch of all plunging
 anticlines and synclines wherever disclosed by changes of dip and
 strike (43).

 11. Indicate the approximate position of all faults whose position
 is disclosed (58-61), and, if possible, state which limb is the one

 12. Prepare suitable geological sections.


 WILLIAM H. HOBBS. Apparatus for Instruction in Geography and
 Structural Geology. III. The Interpretation of Geologic Maps. School
 Science and Mathematics, vol. 9, 1909, pp. 644-653.



The chief value of the laboratory studies discussed in the preceding
appendices is as a preparation for observations made in the field—the
laboratory _par excellence_ of the geologist. The pilgrimages whose
itineraries are here suggested have been planned especially for
impressing by observation the lessons of this book. Such journeys are
best interrupted at a relatively small number of localities which,
because already studied in some detail, are specially adapted to serve
as centers for local excursions. These localities will in most cases
be the great scenic places to which tourists resort, or the seats of
universities near which specially detailed explorations have been often

Within the United States a few local geological guides have been
published, and the Geologic Folios published by the United States
Geological Survey are already available for a number of such centers.
For one long geological pilgrimage we are fortunate in having a
carefully prepared guide, namely, from New York to the Yellowstone
National Park and back, with a side trip to the Grand Cañon of the
Colorado. Except for the side trip this route, in large measure,
corresponds with one here chosen, and for the return journey especially
the student is referred to it for information (Geological Guide Book of
the Rocky Mountain Excursion, edited by Samuel Franklin Emmons. Comte
Rendu de la Congrés Géologique Internationale, 5me Session, Washington,
1891, 1893, pp. 253-487, map and plates 13, figs. 32).

Our journey is begun at New York City, which is built about the deeply
submerged channels of an estuary choked with glacial deposits, though
the channel may be followed as a deep cañon across the continental
shelf to its margin (252,[4] pl. 17 B). New York City is also upon the
margin of the glaciated area, the outer terminal moraine of which is
well represented on Long Island (298). Across the Hudson in New Jersey
is the great Coastal Plain which meets the oldland in a well-defined
margin (159, 246, 247). A local geological guide of the vicinity of the
metropolis has been written by Gratacap (Geology of the City of New
York, Greater New York. Brentanos, New York, 1904, pp. 119, pls. and

Traveling by the New York Central Railway, we follow up the Mohawk
outlet of the glacial lakes Iroquois and Algonquin (334), first
skirting upon the east the great sills of intrusive basalt known as the
Palisades, with their markedly columnar jointing and intersections by
numerous faults. Above Peekskill we enter the picturesque narrows of
the river (174), cut in the hard crystalline rocks of the Highlands.
Entering the Mohawk Valley, we pass Syracuse with limestone caverns
and well-oriented joints widened by solution through the agency of the
descending ground water (181, pl. 6 B). A branch line to the southwest
reaches the vicinity of Cayuga Lake and Ithaca, where are well-oriented
joints which have controlled the drainage directions, and there is also
a typical strath (55, 87, 428).

To Niagara Falls at least a day should be allotted for the “gorge ride”
by trolley car, thus making the complete circuit of the brink of the
gorge with interruptions and local studies at all important points
(352-366, pl. 23 A). From Niagara Falls over the Michigan Central
Railway we reach Detroit on the present outlet of the upper Great Lakes
as well as of the later Lake Algonquin (334). From this city as a
center a trip is made by electric railway to Ypsilanti and Ann Arbor,
across the bottoms of the early glacial lakes from the first Maumee to
Warren (330-333). The strong Whittlesey beach is encountered at the
little station of Ridge Road, and one of the Maumee beaches on Summer
Street in Ypsilanti. The city of Ypsilanti is built upon a terrace
(165) of the Huron River, and another terrace in the same series is
crossed by the electric line. In an excursion of a few miles down the
river, passing meanders (164-165) and ox-bow lakes (165, 415), is
found an interesting case of stream capture near the little village of
Rawsonville (175. See Isaiah Bowman, Jour. Geol., Vol. 12, 1904, pp.

Continuing our journey from Ypsilanti over a high moraine (312), Ann
Arbor is reached, built upon the level plain of outwash with fosses
sometimes separating it from the moraine (281, 314). Upon the campus of
the university are great bowlders of jasper conglomerate and jaspilite,
which were transported from the north by the continental glacier (305).
Across the river from the Michigan Central station and behind the
little church is a delta formed in one of the glacial lakes Maumee and
here opened in section (168). West of the city is a great valley which
was the former course of the Huron River when thus diverted by the
continental glacier lying to the eastward of Ann Arbor—border drainage
(see Ann Arbor folio by the U. S. G. S., and, further, R. C. Allen and
I. D. Scott, An Aid to Geological Field Studies in the Vicinity of Ann
Arbor, George Wahr, publisher, Ann Arbor).

Returning to Detroit (M. C. Ry.), the great Sibley quarries in
limestone near Trenton may be visited. They display perfect jointing,
numerous fossils, and especially well-glaciated surfaces interrupted
by deep troughs and showing striæ of several glaciations (304). From
Detroit the journey is continued by steamer to Mackinac Island in the
strait connecting Lakes Michigan and Huron, passing on the way through
the peculiar delta of the St. Clair River (431), and coming in view of
the notched headlands, which are a monument to the post-glacial uplift
of the glaciated area (250, 341). A day is spent at Mackinac Island
and St. Ignace in order to study with some care these uplifted strands
of the late glacial lakes (341-344). Chicago may now be reached either
by steamer or by rail, and in its vicinity we may see the elevated
beaches and the ancient outlet of Lake Chicago (331-332, 347, pl. 22
A. See Chicago Folio, U. S. G. S.). By the Chicago and Northwestern
Railway the area of recessional moraines and intermediate outwash
plains, and later that of the drumlins, are crossed in journeying to
Madison, Wisconsin. By examination of the maps on pages 308 and 317
in connection with the larger scale atlas sheets of the United States
Geological Survey (Janesville, Evansville, and Madison sheets), this
car journey can be made most instructive in gaining familiarity with
the characteristic glacial features, and this study is continued to
special advantage in excursions about Madison as a center (316-317,
407). This is the more true since at numerous localities in the
vicinity of Madison the well-striated glacier pavement is exposed for
comparison of the striæ as regards direction with the axes of the
several types of glacial features.

An especially instructive excursion may be made by carriage in a single
day to the “driftless area” some twelve miles west of the city. Before
reaching it we cross in alternation a series of recessional terminal
moraines (pl. 17 C) and outwash plains, and near Cross Plains encounter
the partially dissected upland with its arborescent drainage and
even sky line (298, 300-301, 312-313, pl. 16 A and B). Typical shore
formations (233, 241, 242) are studied to advantage about Lake Mendota
in a walking trip to and beyond Picnic Point, where are found the best
ice ramparts (431-434. See Buckley, Trans. Wis. Acad. Sci., Vol. 13,
pp. 141-162, pls. 18).

Our journey is now continued over the Chicago and Northwestern Railway
to Devils Lake near Baraboo, where we cross a salient of the driftless
area, within which lies Devils Lake, imprisoned in a former valley of
the Wisconsin River, since diverted to another course as a result of
the glacial invasion (312-313). The valley here is a former narrows in
hard quartzite (466), which towers above the lake in unstable chimneys
(300), such as the Devils Tower, but such remnants are not found on
the other side of the moraine, being there replaced by rounded rock
shoulders. Just north of the lake the marginal moraine which blocks
the valley is so characteristic as to merit special study (pl. 17
C). Only a few miles northward along the railway from Devils Lake is
Ableman, where, exposed in a high cliff, the hard purple quartzite
with beautiful ripple marks to reveal its plane of sedimentation
(pl. 11 A) dips vertically, and is overlain by horizontally bedded
yellow sandstone. The marked angular unconformity which is thus
displayed is further made evident by a basal layer of conglomerate
(463) in the sandstone (51-53). Here also are deposits of loess along
the river, which display their vertical joint surfaces (207). An
excellent geological guide to this interesting district and that of
the neighboring “Dalles” of the Wisconsin River has been written by
Salisbury and Atwood (The Geography of the Region about Devils Lake and
the Dalles of the Wisconsin, etc., Bull. 5, Wis. Geol. and Nat. Hist.
Surv., 1900, pp. 151, pls. 38, figs. 47).

If we have taken a conveyance at Devils Lake for Ableman, we may
continue in the same manner to Kilbourn, where begin the picturesque
Dalles of the Wisconsin River—here a young gorge cut in sandstone,
because the Wisconsin was diverted from its old valley to border
drainage at the edge of the driftless area (300, 321). The side cañons
of the river, through their abrupt zigzags, reveal the control of their
courses by the joint system (224). In the journey up the rapids by
steamer to inspect the Dalles, we observe many beautiful examples of
cross bedding in the sandstone (37).

From Kilbourn we continue our journey to Minneapolis over the Chicago,
Milwaukee, and St. Paul Railway, and near Camp Douglas are over a
peneplain, out of which rise prominent monadnocks (171). At La Crosse
the Mississippi River is reached, flowing beneath bluffs of sandstone
which are capped by loess (207). The meanderings and the numerous
cut-offs of the Mississippi may be observed to the left (415). Lake
Pepin is a side-delta lake blocked by the deposits of the Chippewa
River (419).

From Minneapolis an excursion is made to Fort Snelling to view the
young gorge of the Mississippi, cut by the Falls of St. Anthony for a
distance of about eight miles in manner similar to that of the seven
miles of Niagara gorge (354), and to compare this narrow gorge with
the broad valley of the Warren River which drained Lake Agassiz (327).
Somewhat farther up the Warren River are examples of saucer lakes (416).

From Minneapolis the journey may be continued by the Great Northern
Railway to Livingston, Montana, thus crossing between the stations of
Muscoda and Buffalo the bed of Lake Agassiz and its marginal beaches
(325-328. For local geology of Minnesota consult C. W. Hall, Geology of
Minnesota, Vol. 1, Minneapolis, 1903).

The Yellowstone Park is entered from Livingston (Livingston Geological
Folio, U. S. G. S.) and departure from it made at the relatively
new Union Pacific terminal at the southwest margin. The regular
trip through the Park includes visits to the several geyser basins
(191-194), Obsidian Cliff (33, 463), the Cañon of the Yellowstone, etc.
Good climbers can make a side trip from near the Mammoth Hot Springs
to the top of Quadrant Mountain, the remnant of a “biscuit cut” upland
(372), and there study the nivation process (368, Yellowstone National
Park Folio, U. S. G. S.).

The trip from the Park to Salt Lake City, over the Union Pacific
Railway, passes through the Red Rock Pass, the former outlet of Lake
Bonneville (423), into the desert of the Great Basin (Chaps. XV and
XVI). Great Salt Lake is a saline lake or sink with an interesting
record of climatic changes (198, 401). The front of the Wasatch Range,
in view and easily reached from Salt Lake City, is deeply scored by
the horizontal shore terraces of Lake Bonneville (198, 199), and
these terraces are extended at every reëntrant by barrier beaches of
great perfection. In the Pleistocene period mountain glaciers in part
occupied the valleys of this range, though they did not always extend
as far as the mountain front. Big Cottonwood Cañon, which realizes this
condition, and the neighboring Little Cottonwood Cañon, from whose
front its glacier spread into an expanded foot (264), thus show for
comparison in a single view the V and the low U sections respectively
(172, 376). Here are also alluvial fans (213) and recent faults which
intersect them.

From Salt Lake City the return to New York may be made by the Denver
and Rio Grande Railway across deserts and through the Royal Gorge,
the cañon of the Arkansas River. A full itinerary of the points of
geological interest along this route, and continued to Chicago,
Washington, and New York, is supplied in much detail in the guide of
the geological excursion to the Rocky Mountains above cited. This
the traveling geologist should not fail to study. Some references to
points along this journey will be found on preceding pages of this
book (219-220, High Plains; 170, Allegheny Plateau in West Virginia;
176, water gap of Harper’s Ferry; 176-177, 184-186, side trip up the
Shenandoah Valley to Luray Caverns and Snickers Gap; 251, Chesapeake

Instead of returning directly from Salt Lake City, the traveler,
if he has sufficient time at his disposal, may extend his journey
southwestward across the Great Basin to Los Angeles. A branch line
from this route leaves the Vegas Valley and passes within reach of the
famous Death Valley (201) to Tonopah (79) and the Owens Valley (77-78,
92), where are many surface faults dating from the earthquake of 1872
and other less recent disturbances. Returning to the junction point,
the route continues across the Colorado and Mohave deserts to Los
Angeles. From Los Angeles as a center the exceptionally interesting
terraces, caves, and stacks of an uplifted coast are to be seen
to best advantage near Pt. Harford (Chap. XIX). The islands of San
Clemente and Santa Catalina may also be reached from Los Angeles (239,
248, 249, 250, 256, 257, pls. 5 B, 7 A, 12 A). The return to the East,
if made by the Santa Fe Railway, permits of a visit to the Grand Cañon
(174, 443) from the station of Williams. From that point eastward the
geology of the route is fully covered in Emmons’ Guide to the Rocky
Mountain Excursion already cited.

       *       *       *       *       *

For the benefit of those who are privileged to travel in Europe, and
the number increases yearly, a pilgrimage is suggested which may easily
be made to correspond with plans laid out on the basis of historical,
artistic, and scenic points of interest. The only popular guide of a
general nature written for geologists traveling abroad appears to be
a brief but valuable little paper by Professor Lane (The Geological
Tourist in Europe, Popular Science Monthly, Vol. 33, 1888, pp.
216-229). The publishing house of Gebrüder Bornträger in Berlin is
now publishing a quite valuable series of geological guides dealing
with special districts and written by well-known authorities (Sammlung
Geologischer Führer). Of this series some thirteen numbers have already
been issued. Many other valuable local guides of a geological nature
are the Livrets Guides of the International Geological and Geographical
Congresses, and the similar pamphlets supplied in connection with
annual meetings of national or provincial geological societies.

Passengers on steamships sailing from the harbor of New York pass
out over a deeply submerged cañon (252) largely filled with glacial
deposits, through the Narrows (174), and in sight of Sandy Hook, a
modified spit (238, 240). To the left are seen the great morainic
accumulations at the border of the glaciated area on Long Island (298).
In the course of the trans-Atlantic voyage a much-rounded iceberg may
be encountered (291), though this is much more apt to occur upon the
northern routes from Quebec, and late in the season. Upon entering the
English Channel the land on both coasts rises in steep cliffs, where
are found all the common shore features well developed (Chap. XVIII).
The German steamships pass in sight of Heligoland, that last remnant of
wave erosion (236).

While traveling in Europe, the student should consult a map of the
glaciated area (299), and so learn to recognize its peculiarities, and
carefully mark its marginal moraine (311) and other strongly marked

If the British Isles are visited and the more rugged areas are
selected, one may study the cirques and other characteristic features
due to the presence of mountain glaciers about Snowdon (Chap. XXVI).
More mature stages of the same processes are to be found in the
Scottish Highlands and the Inner Hebrides, but especially upon the
Island of Skye (Fig. 492). A very valuable aid to excursions in
this district is Baddeley’s Scotland (part I, Dulau, London) and
Sir Archibald Geikie’s Explanatory Notes to accompany Bartholomew’s
Geological Map of Scotland (map and notes in cover, Edinburgh, 1892,
pp. 23).

[Illustration: FIG. 492.—Sketch map of Western Scotland and the
Inner Hebrides to show location of some points of special geological

It is from Oban, the “Charing Cross of the Highlands”, that one should
start out upon the summer steamers in order to reach both Skye and
Staffa, the latter with fine basaltic columns (463), and Fingal’s
Cave. In sailing to Skye one passes upon either shore of the narrow
fjords many relics left in the dissection of volcanoes (139-143 and
Sir A. Geikie, Ancient Volcanoes of Great Britain, Vol. II); also
rocky islands and skerries marking submergence (252), and the coast
terraces which register a later uplift (250). Skye is a complex of
many intrusive and volcanic rocks of such markedly different colors
as to appear as tints in the landscape. In the Cuchillin Hills of dark
green rises the massive gabbro (462) cut by cirques into the jagged
pinnacles of horns and comb ridges (373); while lower down and to the
east are rounded domes of rhyolite (463) abraded beneath the glaciers
and of a delicate salmon tint. Still lower and to the westward are
flat mesas composed of horizontal layers of black basalt under a rich
carpeting of the brightest verdure. Eastward across the channel are
seen the purplish walls of an ancient sandstone. The jagged gabbro
core of the island thus represents a fretted upland (372) and is now
the training ground of the Alpinist (Abraham, Rock Climbing in Skye,
Longmans, London, 1908), while nestled in one of the bottoms of a
U-valley is Loch Coruisk, a typical rock-basin lake (412), its shores
of hard rock planed and scored.

From Skye we may go to study the remarkable thrusts (45) on the north
shore of Loch Maree, a marked lineament, and one directed at right
angles to that on the course of the Caledonian Canal connecting
Loch Linne with Loch Ness. This northeast wall of Loch Maree is a
strikingly rectilinear fault represented by an escarpment, up which
we climb to find at the top the crushed and fluted thrust planes of
movement dipping southeastward at a flat angle. Here also are beautiful
rock-basin lakes, lying in hollows molded beneath the continental
glacier. On our way from Skye we have passed up Loch Carron, a sea loch
or fjord (252), and along the strath at its head known as Strathcarron

Returning now to Oban, it is but a short trip by steamer up Loch Linne
to Fort William along the striking lineament (226) which continues to
Loch Ness and beyond (Fig. 492), and thence by rail to Glen Roy and the
neighboring glens of Lochaber (322-325).

From Paris as a starting point, we may visit in a most picturesque
region the beautifully preserved craters of extinct volcanoes in the
Auvergne of Central France (105, 124, 145), which district is entered
from Clermont-Ferrand. Here are found the characteristic puys, steep
lava domes of viscous lava (105), which figured largely in the early
controversies of geologists concerning the origin of rocks.

[Illustration: FIG. 493.—Outline map of a geological pilgrimage across
the continent of Europe.]

The rest of our pilgrimage will be so planned as to enter the noble
river Rhine at its mouth (Fig. 493), ascend its course to its
birthplace in the snows of Switzerland, and after further exploration
of the features of this fretted upland, traverse northern and central
Italy so as to make our departure for America by the southern route.
Entering then upon this course in the Low Countries, we have first
the opportunity of observing the characteristics of a great delta
with natural levees artificially strengthened as dikes (165-168).
Here also are found dunes of beach material which has been raised by
the wind into a great rampart near the shore (209-211). Such a wall
of dune sand is well displayed at the bathing resort at Scheveningen
near the Hague (421). The flood plain of the Rhine (162-165) may be
studied in a journey up the river to the university town of Bonn, from
whence a day’s excursion should be devoted to the relics of volcanoes
known as the Seven Mountains (H. von Dechen, Geognostischer Führer in
das Siebengebirge, Bonn, 1861). As a preparation for this trip and
others in the volcanic Eifel higher up the river, a visit should be
made to the mineral and rock collections of the Poppelsdorfer Schloss
at the University. In the volcanic Eifel are found some of the most
interesting of crater lakes (405), the largest being Lake Laach
with its somewhat peculiar volcanic ejectamenta and its picturesque
abbey (see von Dechen, Geognostischer Führer zu der Vulkanreihe der
Vorder-Eifel, etc., Bonn, 1886. Consult also Lane, A Geological Tourist
in Europe, _l.c._).

Continuing our course up the river from Bonn, we soon enter the gorge
of the Rhine cut in an uplifted peneplain (169, 171, 174). From
Coblenz, where the Moselle enters the Rhine, a side trip may be made up
this tributary river past Zell with its entrenched meanders (173) to
the ancient Roman city of Treves. Above Bingen on the Rhine we leave
behind us the narrow gorge and rapid current of the river and continue
over the broad floor at the bottom of a rift valley (403), lying
between the forest of Odin and the Black Forest on the east and the
“Blue Alsatian Mountains” far away to the west. At the margins of this
plain are beds of loess with their characteristic joint structures and
inclusions (207), and in the higher hills on either hand a wealth of
intrusive igneous rocks.

At the entrance of the Neckar River to this broad plain is nestled the
picturesque castle and university town of Heidelberg, a convenient
center for excursions (Julius Ruska, Geologische Streifzüge in
Heidelbergs Umgebung, etc., Nägele, Leipzig, 1908, pp. 208, map).
At Strassburg (Schwarzwaldstrasse 12) is located the German Chief
Station for Earthquake Study, with a particularly large set of modern
seismographs. In the university cabinet is also one of the largest and
most representative mineral collections in Europe. For excursions in
the neighborhood consult Benecke, Sammlung Geognostische Führer, Vol.
5, Elsass, 1900.

From Strassburg we may go by the Black Forest Railway to the Hegau with
its volcanic plugs (140), each surmounted by a picturesque castle.
We enter next the broadly extended piedmont apron site, above which
Lake Constance still remains as a border lake (399). Outwash aprons
(314), moraines (311), and drumlins (317) are each in turn encountered.
Still continuing our course up the Rhine from Bregenz, we enter the
fretted upland (372) of the Alps, mountains composed of great folds and
thrusts about a core of intrusive rock (Rothpletz, Sammlung Geologische
Führer, Vol. 10, 1902, Thrusts in the Alps between Lake Constance and
the Engadine). Some fourteen miles above Chur we pass the terrace
produced by successive landslides (414), known far and wide as the
Flimser Bergstürz. The further assent of the cascade stairway of this
glacier-carved valley brings us to the Furka Pass, from which point
magnificent views of the fretted upland are obtained. At the Känzli, a
mile from the hotel, one may view the névé of the Rhone Glacier, which
may also be easily visited.

We have now followed a great river from its mouth in the sands of
Holland to its source in the snows of the higher Alps. Passing over
the divide and descending to Gletsch, we may observe the lower end,
or foot, of the Rhone glacier and the crevasses and séracs (391) on
the steep descent of this radiating glacier (383, 386). The response
which glaciers make to climatic changes is here well illustrated by the
recession of the glacier front from near the hotel (its position in the
’50s of the nineteenth century) to its present position about a mile
farther up the valley.

The characteristics of a glaciated mountain valley may be further
illustrated by climbing to the Grimsel Pass, which is scratched and
striated (377, 385), and then descending the valley of the Aar to
Meyringen (377). Near the Grimsel Hospice are the characteristic rock
basin lakes (412), and upon the Aar Glacier to our left were carried
out the epoch-making researches of Louis Agassiz, the founder of the
glacial theory for explaining the drift. We encounter some thirteen
rock bars (377). Just before reaching Meyringen we pass the last of
these, the Gorge of the Aar, cut by the stream through limestone.

Interlaken (419) may be made the center for additional excursions up
the Lauterbrunnen Valley, with its prominent albs (376) and its ribbon
fall of the Staubbach (378). By the Jungfrau Mountain railway we may
now ascend partly in tunnels of the rock to the Ewigeismeer, and look
down upon the névé and bergschrunds of the Great Aletsch Glacier (370,
see Baltzer, Sammlung Geologische Führer, Vol. 10, Bernese Oberland,
1906). Returning to Interlaken by way of Grindelwald, one may study the
foot of a radiating glacier, the Untergrindelwald glacier, with its
tunnel and its milky and braided stream.

Crossing now the Alpine foreland to Villeneuve at the upper end of
Lake Geneva and upon a well-developed strath (426, 428), we may look
out upon the turbid waters extending far from the shore of the lake.
Journeying to Geneva by steamer we note the gradual clearing of the
water until at the outlet of the lake it is as clear as crystal. A
walking trip from Geneva takes us to the Bois de la Bâtie, where the
Arve with turbid waters meets this clear stream (427).

The railroad to Chamonix ascends another cascade stairway (376),
affords views of complexly folded sedimentary rocks (43), and at
Chamonix itself the mer de glace supplies opportunities for the study
of moraines (386, 393) and glacial movement (390-392). To experienced
Alpinists the summit of Mount Blanc offers a remarkably extended
outlook over the fretted upland of the Alps (pl. 18 A). From the
station of LeFayet below Chamonix, one may ascend to the Désert de la
Platé, where are Schratten in limestone due to solution (188).

Crossing by one of the passes to the valley of the Rhone at Martigny
we may reach Zermatt, to-day the climbing center of the Alps. From
the subordinate cirques surrounding this village descend the Gorner,
Findelen, St. Theodul, and other components of this radiating glacier.
A black tooth of rock, the Matterhorn, towers above the other peaks and
shows to greatest advantage this feature of glacial sculpture (374),
while the Gorge of the Gorner is a severed rock bar like that of the
Aar (377). Either on foot or over the mountain railway we may ascend to
the Gorner Grat, a subordinate comb ridge (373) which affords one of
the most magnificent and instructive views of radiating glaciers.

From Brig, farther up the Rhone Valley, an excursion is made to the
Eggishorn Hotel, a center for study on and about the Great Aletsch
Glacier (329, 371, 385, 388, 395, 410). The easy ascent of the
Eggishorn is rewarded by a view almost directly downward upon the
ice-dammed Márjelen Lake (329, 411).

From Brig one may make his entry into Italy, either over the
picturesque Simplon route afoot or by diligence, or else beneath it
through the railway tunnel. By an alternation of short steamboat and
rail trips the journey is continued in a direction transverse to the
longer axes of the border lakes Maggiore, Lugano, and Como, and later
southward to Milan. In leaving the village of Como we pass over heavy
morainic deposits on the apron borders of the expanded-foot glacier
(383, 385) which once occupied the valley above. On the journey from
Milan to Venice, over the fertile plains of Lombardy, the similar
accumulations about Lake Garda (414) are first encountered at the
little station of Lonato and left behind at Somma Campagna (Tornquist,
Sammlung Geologische Führer, Vol. 9, Northern Italy, 1902).

The city of Venice is built upon pile foundations in the lagoon behind
the barrier beach known as the Lido (242, 428-429). From here we may
reach the Karst country by way of Trieste, some of the more interesting
and typical features being found near Divača (187-189, 422, pl. 6 A).
In a different direction from Venice by way of Belluno we enter the
Dolomites with their patterned relief and battlemented towers (228,

Additional centers for geological excursions on the route to our point
of departure from Italy are Rome and Naples. At the Italian capitol and
in its neighborhood we may study the volcanic Campagna with its beds
of tuff (105) and its crater lakes (405. See Sir A. Geikie, The Roman
Campagna, Landscape in History and other Essays, Macmillan, 1905, pp.
308-352; also Deecke, Sammlung Geologische Führer, Vol. 8, Campagna,
1901). From Rome it is an easy journey to the cataract of Tivoli with
its deposits of travertine (184). In the opposite direction from Rome
across the Campagna rise the Alban Hills, ruins of a composite cone
with several crater lakes on the sites of former vents. On the summit
of the encircling crater rim, like the Monte Somma of the Vesuvian
Mountain now a crescent only, is located the chief Italian station for
earthquake study.

From Naples we may reach in short excursions and study with some
care still active volcanic mountains. To the east is Mount Vesuvius
(94, 97, 122, 124, 127-137), which was in grand eruption in April,
1906. Westward from Naples are the Campi Phlegraeii, or burning
fields, with many craters. Of these Astroni offers a fine example of
a large-cratered cinder cone (105). In the same vicinity are Monte
Nuovo (96) and the Solfatara (97), the latter a type of volcano which
no longer erupts lava, but in its place emits carbon dioxide and other
gaseous emanations (Grotto del Cane). The starting point for excursions
in the Phlegræan fields is Pozzuoli with its Temple of Jupiter Serapis
(254-255), reached from Naples by an electric line which pierces the
wall of an immense crater (Posilippo) composed of fine yellow volcanic
ash known as Pozzuolan.

From Naples steamers make short excursions to Sorrento with its deep
ash deposits, and to Capri with its blue grotto (257-258). Herculaneum
(139) and Pompeii (122), buried during the eruption of 79 A.D., are on
the line of the Circum-Vesuvian Railway.

Steamships to New York from Naples call at Gibraltar, the land-tied
island _par excellence_ (241). Most steamships of the southern route
pass through or near the volcanic islands of the Azores, and certain
boats touch at Algiers, from which a line of railway gives access to
Biskra on the borders of the Desert of Sahara.

Throughout these pilgrimages the traveler should be on the alert to
note not only the agent responsible for the features which come under
his observation, but, especially where this is the common sculpturing
agent of running water, he should not fail to notice the stage of the
erosion cycle which is represented (Chapter XIII).


  Abrasion, beneath glaciers, 275.

  Abyssinia, fissure eruptions in, 101.

  Accordance, of tributary valleys, 162.

  Adiabatic refrigeration, in relation to glaciers, 262.

  Adolescence, in cycle of erosion, 169.

  Advancing hemicycle of glaciation, 263-266.

  Advective zone, of atmosphere, 270.

  Aftershocks, of earthquakes, 83.

  Agassiz, glacial lake, 325-328.

  Agassiz, Louis, cited, 339, 400.

  Age, of strata, 38, 52.

  Aggradation, 162.

  Aktian deposits, 36.

  Alaskan coast, map of, 79.

  Albs, 376.

  Alden, W. C., cited, 316, 318, 319.

  Algæ, growth of, in hot springs, 194.

  “Alkali” in deserts, 201.

  Alluvial bench, 214.

  Alluvial cone, 213.

  Alluvial-dam lakes, 423.

  Alluvial fan, 213.

  Alpine glaciers, 383, 386.

  Alterations of minerals, 27.

  Altitude, of different parts of lithosphere, 18.

  American Falls, future extinction of, 357.

  Amphiboles, 459.

  Amphitheaters, formed on drift sites, 369.

  Amundsen, R., cited, 23.

  Analysis, of folds, 54.

  Anderson, Tempest, cited, 146, 147.

  Andersson, J. G., cited, 157, 295.

  Andesite, 463.

  Angular unconformity, 53.

  Antarctica, 154, 281.

  Antarctic protuberance, 17.

  Antarctic shelf ice, 289, 290.

  Anticlinal folds, 42.

  Anticlines, 42;
    tension in, 45.

  Anticyclone, glacial, 284.

  Ants, factor in rock decomposition, 156.

  Apron, alluvial, 213.

  Aprons, outwash, 280, 281.

  Arbenz, P., cited, 195.

  Arches, of folded strata, 42;
    sea, 233, 234.

  Architecture, of fractured earth superstructure, 55.

  Arctic depression, 17.

  Areal geological map, 62.

  Arêtes, 373.

  Arldt, Theodore, cited, 11, 19, 438.

  Arnold, Ralph, cited, 157.

  Arrangement of oceans and continents, 10.

  Artesian wells, 190, 191, 196.

  Ash, volcanic, 122.

  Askja, eruption of, in 1875, 101.

  Assmann, R., cited, 294.

  Astronomical _vs._ geodetic observations, 12.

  Atlantis, North, 16.

  Atmosphere, compressibility of, 8.

  Attack, of the weather, 149.

  Atwood, W. W., cited, 7, 160, 298, 300, 313, 372.

  Axial plane, of folds, 42.

  Axis, of folds, 42.

  Azurite, 453.

  Bacteria, part taken in weathering, 156.

  “Bad Lands”, control of relief in, 223, 224.

  “Bad Land” topography, 214.

  _Bajir_, 216.

  Balance, between degradation and aggradation, 161.

  Bandai-san, dissection of, 141.

  Barchans, 211.

  Barrancoes, 139.

  Barrell, J., cited, 221, 447.

  Barrier beaches, 240;
    sections of, 242;
    uplifted, 249, 250.

  Barrier lakes, 420.

  Barriers, 240;
    mountain, in relation to glaciers, 262.

  Bars, 240.

  Basal conglomerate, 37, 53.

  Basalt, 463;
    faulted blocks of, 58;
    of Hawaii, 105.

  Base level, 159.

  Basin-range lakes, 402, 403.

  Basin Range structure, 440.

  Basins, flat bottomed, separating dunes, 216;
    of exudation, 272;
    of sedimentation, earlier, 38.

  Bastin, E. S., cited, 210.

  Batholites, 143.

  “Bath tubs”, 395.

  Beach pebbles, 239.

  Beach sand, 206, 238.

  Beaches, remaining from ice-dam lakes, 410;
    shingle, 239;
    storm, 240;
    uplifted, “feathering out” of, 344.

  Bedded structure of rocks, 31.

  Beede, J. W., cited, 195.

  “Bee-hive” mountains, 380, 381.

  _Belgica_ expedition, 289.

  Belt of sea which divides land masses, 11.

  Berghaus, H., cited, 424.

  Bergschrund, 370.

  Berson, A., cited, 294.

  Berthaut, General, cited, 7.

  “Bird-foot” delta, 167.

  “Biscuit cutting” effect of glacial sculpture, 372.

  Blackwelder, E., cited, 318.

  Block mountains, 446.

  Blocks, orographic, 58.

  _Bocchi_, 125.

  Bog, floating, 429.

  Bogs, of peat, 429, 430.

  Bonney, T. G., cited, 146.

  Borax deposits, in deserts, 201.

  Border drainage, about glaciers, 316, 320, 321.

  Border lakes, 399, 414.

  Bosses, 143.

  “Bottoms”, from entrenched meanders, 173.

  “Bowlder clay”, 310.

  “Bowlder pavement”, 237.

  Bowlders, faceted, 310;
    glacial, 298;
    “soled”, 276, 310;
    thrown up during earthquakes, 69.

  Bowlder trains, 306.

  Bowman, Isaiah, cited, 179.

  Box cañons, 214.

  Braided streams, 280.

  Branner, J. C., cited, 6, 91.

  “Bread-crust” lava projectiles, 119.

  Breakers, 232.

  Breccia, fault, 60.

  Bridges, nature of damage to, during earthquakes, 75, 76.

  Brigham, A. P., cited, 424.

  Brögger, W. C., cited, 66.

  Bruce, W. S., cited, 290, 382, 399, 414.

  Bryant, H. G., cited, 289.

  Buckley, E. R., cited, 433, 434.

  Built terraces, 235.

  Bunsen, cited, 192.

  Burns, G. P., cited, 434.

  Burton, W. K., cited, 92.

  Buttes, 216.

  Bysmalite, 442, 447.

  Calcareous ooze, 36.

  Calcareous sinter, 184.

  Calcareous tufa, 464.

  Calcite, 455.

  Caldera, 405, of composite volcanic cones, 126.

  Camiguin volcano, birth of, 96, 97.

  Campbell, M. R., cited, 178.

  Cañons, 160;
    box, 214.

  Capri, blue grotto of, 257, 258.

  Capture, river, 175, 176, 179.

  Carbonization, 151.

  Cascade Mountains, fissure eruptions of, 102.

  Cascade stairway, 376.

  Caspian Depression, 14.

  Cauliflower cloud, 130.

  Caverns, galleries directed by joints, 182;
    of limestone, 182, 195;
    refuge of predatory animals, 185.

  Caves, sea, 234.

  Cellular structure, of lava domes, 112.

  Centers of dispersion, of North American Pleistocene glaciers, 298.

  Centrosphere, 8.

  Cerussite, 455.

  Chaix, A., cited, 195.

  Chaix, E., cited, 195.

  Chalcopyrite, 453.

  Challenger expedition, 38, 96, 97, 293.

  Chamberlin, T. C., cited, 29, 156, 191, 196, 205, 221, 222, 293, 295,
                            318, 319, 337, 339.

  Character profiles, coast, due to uplift or depression, 259;
    composite, 229;
    directly due to volcanic agencies, 145, 146;
    from stream erosion in humid climates, 177;
    of arid lands, 220;
    of shore features, 243;
    referable to continental glaciers, 318;
    referable to mountain glaciers, 379.

  “Checkerboard topography”, 226.

  Chemical sediments, 34.

  Chicago outlet, 331.

  Chimneys, in “driftless area”, 300.

  Chimneys, shore feature, 234.

  China, loess of, 207.

  Chlorite, 458.

  Chlorite schist, 465.

  Cicatrice, from dissection of volcanoes, 142.

  Cinder cones, 105;
    corrugations upon, 138;
    diameter of crater in relation to violence of explosions, 123;
    grander eruptions of, 117;
    profiles of, 123;
    secondary, 111.

  Cinder eruptions, artificially simulated, 122.

  Cirques, 371;
    life history of, 371;
    subordinate, 371.

  Cities, destruction of, by drifting sand, 218.

  Clastic rocks, 30.

  Clay slate, 466.

  Cleavage, mineral, 27, 450;
    rock, 44.

  Clefts, volcanic, in Iceland, 99.

  Cliffs, notched, 233.

  Climatic conditions, in relation to mountain sculpture, 443.

  Clinometer, 48.

  Cloudbursts, in deserts, 201, 212.

  Cloud zones, 268, 269, 294.

  Coals, 466.

  Coast, Dalmatian, grottoes of, 258.

  Coast, elevation of, during earthquakes, 80;
    submergences of, during earthquakes, 80.

  Coastal plains, 246;
    belted, 247.

  Coast lines, even, 246;
    indicative of uplift or submergence, 245, 246;
    ragged, 246.

  Coast records, 245.

  Coasts, Atlantic and Pacific contrasted, 438;
    embayed, 251.

  Coast terraces, 80, 250, 241;
    uplift, effect of, on sediments, 38.

  Coats Land, shelf ice of, 290.

  Cobalt, in meteorites, 23.

  Cobb, Collier, cited, 179.

  Coigns, of earth’s tetrahedral figure, 15.

  Coleman, A. P., cited, 318.

  Colk lakes, 408, 409.

  Colks, scape, 277.

  Collet, L. W., cited, 39.

  Colorado desert, 74.

  Color, of minerals, 450.

  Cols, 374;
    origin of in cirque intersection, 372.

  Comb ridges, 373.

  Compass, geologist’s, 47, 48.

  Competent layer, 42;
    in relation to lava reservoirs, 144.

  Composite cones, _caldera_ of, 126, 127.

  Composite groups of joints, 57.

  Composite volcanic cones, 105.

  Composition of earth, 29.

  Composition of the earth’s core, 21.

  Compression of a district during earthquakes, 76.

  Cones, alluvial, 213;
    cinder, 105;
    composite volcanic, 105.

  Conformable series, 51.

  Conglomerate, 34, 463;
    basal, 37, 53.

  Constructional topography, 309.

  Construction of buildings, in earthquake regions, 89-91.

  Continental glacier, behind rampart, 281;
    in Victoria Land, 280-285;
    of Antarctica, literature of, 295;
    of Greenland, 271;
    of Greenland, melting on margin of, 278;
    of Greenland, literature, 295.

  Continental glaciers, contrasted with mountain glaciers, 266-268;
    defined, 266-267;
    of “ice age”, 297;
    of ice age, cross section of, 302;
    nourishment of, 283, 286, 295;
    profiles of, 267.

  Continental platform, 19.

  Continental shelves, 18, 19;
    origin, 232.

  Continents, arrangement of, 10;
    development of, 14;
    increase in area of, through wave action, 241;
    past history of, 14.

  Contortions of the strata, 40.

  Contours, of topographic maps, 62.

  Contraction of earth’s surface, during earthquakes, 74.

  Contrary movements upon coasts, 254, 257.

  Convective zone, of atmosphere, 270.

  Conway, W. M., cited, 294.

  Copernicus, cited, 10.

  Copper glance, 455.

  Coquina, 35.

  Cornish, Vaughan, cited, 211, 222, 244.

  Corrasion, 162.

  Corrosion, of rocks, 156.

  Coulée lakes, 406.

  Coves, 233, 234.

  Cracks, earthquake, 74.

  Crater, evolution of form of, 128.

  Crater lakes, 405, 406.

  Craterlets, 84;
    sections of, 85.

  Craters, mechanics of explosions in, 115.

  Crater, volcanic, 95.

  Credner, G. R., cited, 179.

  Crescentic levee lakes, 416, 417.

  Crestline, of an anticline, 42.

  Crevasse, marginal, on mountain glaciers, 370.

  Crevasses, in connection with river cut-offs, 164;
    on glaciers, 391.

  Cross, Whitman, cited, 216, 441, 447.

  Cross-bedded structure, 37.

  “Crystal cellars”, 27.

  Crystal form, of minerals, 449.

  Crystals, behavior under special treatment, 24, 25;
    essential nature of, 23;
    forms of, 454, 457;
    individuality of, 24;
    mutilated, later growth of, 26;
    symmetry of form of, 23.

  Crustal shortening, 42.

  Cuestas, 246, 247;
    south of Lake Ontario, 361, 362.

  Cut and built terrace, on steep shore of loose materials, 237.

  Cut-offs, of meanders, 164.

  Cut rock terraces, 235.

  Cuvier, cited, 199.

  Cvijić, J., cited, 195.

  Cycle of glaciation, 263, 294.

  Cycles, of glaciation, Pleistocene, 297;
    of stream meanders, 163.

  Dana, J. D., cited, 6, 104, 106, 109, 111, 146, 147.

  Dana, E. S., cited, 29.

  Daly, R. A., cited, 447.

  Dante, cited, 9.

  Darton, N. H., cited, 179.

  Darwin, Charles, cited, 199, 322, 323, 339.

  Daubrée, A., cited, 54.

  David, T. W. E., cited, 23.

  Davis, C. A., cited, 434.

  Davis, W. M., cited, 7, 178, 179, 221, 247, 276, 317-319, 378, 382.

  Deceptive unconformity, 53.

  Decomposition, 149, 156;
    mechanical results of, 150.

  Débris cones, 395.

  Deep sea deposits, 36, 38.

  Deflation, 204.

  Deforestation, in relation to agriculture, 156;
    of Karst region, 188;
    relation to erosion, 157.

  Degeneration, 149.

  De Geer, G., cited, 351, 366, 410.

  Degradation, 161, 162.

  Dekkan, fissure eruptions of, 101.

  Delebecque, A., cited, 424.

  De Lorenzo, cited, 125, 132.

  Delta, “Bird-foot”, 167;
    bottom-set beds, 167;
    dry, 213;
    of Mississippi River, rate of growth of, 168.

  Delta deposits, manner of growth of, 167.

  Delta lakes, 419, 420.

  Delta region, of a river, 35.

  Deltas, abnormal, below outlets of lakes, 431;
    in relation to agriculture, 166;
    in relation to population, 166;
    lake, 428;
    of rivers, 165, 166, 179;
    sections of, 168.

  Dendritic glaciers, 383, 385, 386.

  Deniston, cited, 121.

  Deposition, in zones about desert, 216, 217.

  Deposits, aktian, 36;
    chemical, 34;
    continental, 37;
    deep sea, 36, 38;
    delta, manner of growth of, 167;
    fluviatile, 35;
    fluvio-glacial, 31, 310;
    in valley vacated by glacier, 398;
    glacial, 31;
    lacustrine, 35, 217;
    littoral, 36;
    marine, 35;
    mechanical, 34;
    organic, 34;
    salt, 217;
    shoal water, 26;
    sinter, 184;
    terrigenous, 36.

  Derangement of water flow, during earthquakes, 83, 84.

  Derwies, V. de, cited, 447.

  Descent of ground water, 180.

  Desert, due to deforestation, 156;
    erosion in, 214, 222;
    law of, 197.

  Desert lakes, 423.

  Desert landscapes, features in, 209.

  Desert rains, 212.

  Desert rocks, red color of, 222.

  Desert varnish, 201, 222.

  Deserts, former shore lines in, 198;
    self-registering gauge of past climates, 198.

  Destructional topography, 309.

  Detection of plunging folds, 49, 50.

  Detonations, during Vulcanian eruptions, 131.

  Device, to simulate building of cinder cones, 122.

  Diabase, 462.

  Diagram, to illustrate formation of lava reservoirs, 143.

  Diagrams for comparison of fold types, 42;
    to show the effect of spheroidal weathering, 150.

  Diamonds, in the drift, 307.

  Diffission, 204.

  Dikes, hollow, 140;
    in China, 167;
    in Holland, 166;
    from volcanic dissection, 140.

  Diller, J. S., cited, 39, 425.

  “Diluvium”, 305.

  Dimples, on margin of continental glaciers, 272.

  Dip, 46.

  Dirt cones, 396.

  Disintegration, 156;
    of rocks in deserts, 202;
    through root expansion, 154;
    through tree growth, 154, 155.

  Dislocations, marginal, about deserts, 212.

  Dispersion of the drift, 304-309, 319.

  Displacement, total, on faults, 59.

  Dissection of volcanoes, 139.

  Distributaries, on alluvial fans, 213, 220.

  Divides, 170;
    migration of, 175.

  Dolines, of Karst region, 187, 422.

  Dolomite, 465.

  Dolomites, 203, 228, 445.

  Domed mountains of uplift, 441.

  Dome structure, of granite masses, 152, 157.

  Domes, lava, 105.

  Dovetailing, of sea and land, 11, 17.

  Drainage, changes of, due to glaciation, 336-338;
    haphazard, of glaciated area, 301;
    interference of glaciers with, 320;
    of glaciers, 397;
    reversals of, due to glaciation, 337, 338;
    trellis, 175.

  Drainage lines, control of, by fractures, 224.

  Drainage networks, controlled by fractures, 225, 226;
    repeating pattern in, 225.

  Drake, Sir Francis, circumnavigation of the globe, 10.

  _Dreikanten_, 205.

  Driblet cones, 104, 125;
    of Kilauea, 107.

  “Drift”, 305.

  Drift, assorted, 309;
    dispersion of, 304-309;
    englacial, 277, 278;
    unassorted, 309.

  “Driftless area”, 300, 313, 318.

  Driftless area, map of, 298.

  Drift sites, 368, 369.

  Drowned rivers, 251.

  Drumlins, 311, 316, 317, 399.

  Dry deltas, 213.

  Drygalski, E. von, cited, 273, 279, 295, 296.

  Dry weathering, in deserts, 201.

  Dune, war with oasis, 216.

  Dune lakes, 421.

  Dunes, 222;
    forms of, 210, 211;
    in relation to obstructions, 209, 210;
    stopped by vegetation, 211;
    wandering, 209, 211.

  Dust, carried out of desert, 206, 222;
    volcanic, 122.

  Dust wells, 395.

  Dutton, C. E., cited, 85, 92, 178, 200, 222, 447.

  Earlier figures of the earth, 14.

  Earth, a magnet, 23;
    composition of, 20;
    oblateness of, 10;
    rigidity of, 20, 21, 29;
    scale of its elevations, 10, 11;
    theories of origin of, 20, 29;
    surface shell, chemical constitution of, 23;
    surface shell, response to load, 340.

  Earth features, shaped by running water, 169.

  Earth figure, evolution of ideas concerning, 9.

  Earthquake cracks, 74.

  Earthquake fountains, 190.

  Earthquake lakes, 404.

  Earthquake, of Alaska, 1899, 72, 77, 79, 80, 81;
    of Assam, 1897, 72, 77;
    of California, 1906, 70, 72, 73, 74, 90, 91;
    of Casamicciola, 1883, 87;
    of Costa Rica, 1910, 68;
    of India, 1819, 84;
    of Jamaica, 1692, 80;
    of Jamaica, 1907, 80;
    of Japan, 1891, 72, 75;
    of lower Mississippi Valley, 1811, 83;
    of Messina, 1908, 68;
    of Owens Valley, California, 1872, 73, 77, 78, 79;
    of Servia, 1904, 84;
    of South Carolina, 1886, 85.

  Earthquake shocks, heavy over loose foundations, 88.

  Earthquakes, aftershocks of, 83;
    associated with growing mountains, 86;
    changes in earth’s surface during, 71;
    connected with lines of fracture, 86;
    descriptive reports upon, 92;
    due to adjustments between blocks of shell, 78, 79;
    faults and fissures, 71;
    focused at fault intersections, 87;
    fountains during, 83, 86;
    localized at corners of earth blocks, 87;
    manifestations of changes in level, 68;
    nature of shocks, 67;
    of Ischia, localization of, 87;
    shown by coast terraces, 250;
    special lines of heavy shock, 86;
      in unstable areas of earth’s crust, 86;
    wave motions of, 68;
    zones in distribution of, 86.

  Earth relief, repeating patterns in, 223.

  Eckert, cited, 188.

  Effect of contraction upon a spherical body, 13.

  Egg-spinning demonstration of earth rigidity, 20.

  “Elevation-crater” theory of volcanoes, 95, 139.

  Embankments, shore, 240.

  Embayed coasts, 251.

  Emerson, B. K., cited, 19.

  End moraines, 394.

  Engell, M. C., cited, 296.

  Englacial débris, 393.

  Englacial drift, 277, 278.

  _Entonnoirs_, 182.

  Entrenchment of meanders, 172, 173, 179.

  Eolian sand, 206.

  Eolian sediments, 30.

  Erosional unconformity, 53.

  Erosion cycle, 159.

  Erosion, effect of, in adding curves to landscape, 65;
    glacial, in contrast with normal weathering, 377;
    in desert, 214;
    shadow, 206;
    stream, as modified by resistant rocks, 174.

  “Erratic blocks”, 304.

  Eruptions, Strombolian, 117;
    Vulcanian, 117, 125.

  Escarpments, from faults, 59.

  Eskers, 311, 315, 316, 363.

  Estes, L. A., cited, 93.

  Estuaries, 251.

  Etna, eruption of 1669, 122.

  Evolution, doctrine of, in connection with fossils, 38.

  Evolution of ideas concerning the earth’s figure, 9.

  Exfoliation, 151, 203.

  Expanded foot glaciers, 383, 385.

  Experiment, to illustrate relation of earthquake shocks to
           foundations, 88.

  Experiments, on fracture and flow, 40, 41;
    for demonstration of earthquakes, 81, 82.

  Exposures, rock, 46.

  Extrusive rocks, 463.

  Fairbanks, H. W., cited, 155, 170, 174, 201, 205, 214, 224, 248, 249,
                           250, 260, 302, 375, 406, 413, 429.

  Fairchild, H. L., cited, 339.

  Falls, “Bridal veil”, 378.

  Falls, ribbon, 378.

  Fan, alluvial, 213.

  Farrington, O. C., cited, 29.

  Fault, drag upon, 60.

  Fault breccia, 60.

  Fault topography, 65.

  Faults, 58, 440;
    during earthquakes, 71;
    earthquake, change in throw upon, 76, 77, 78;
    earthquake, disappear in loose materials, 73;
    earthquake, of small displacements, 74;
    earthquake, plan of, 76, 78;
    illusory nature of, 59;
    methods of detecting, 59;
    post-glacial, 74;
    relation of escarpments to, 60;
    shown by changes in strike and dip, 61;
    shown by offsets, 61.

  Feldspars, 456.

  Fenneman, N. M., cited, 424, 425.

  Festoons of mountain arcs, 435, 436.

  Field ice, 286.

  Field map, geological, 62, 63.

  Figure of the earth, the, 8.

  Figures, earlier, of the earth, 14;
    earth, evolution of, 15.

  Figure toward which the earth is tending, 12.

  “Fire girdle” of the Pacific, 98.

  Firn, 369.

  Fissure eruptions, of volcanoes, 101.

  Fissures, during earthquakes, 71;
    earthquake, 74;
    in connection with volcanoes, 99-101.

  Fissure springs, 61, 190, 195.

  Fjords, 290, 340.

  “Float copper”, 305.

  Flooded portions of continents, 18.

  Flood plain, 178;
    manner of grading of, 162.

  Floors of hydrosphere and atmosphere, 18.

  Flow, experiments on, 41;
    zone of, 40.

  Flow texture, of extrusive rocks, 33.

  Fluviatile deposits, 35.

  Fluvio-glacial deposits, 31.

  Fluxion texture, of extrusive rocks, 33.

  Folds, analysis of, 54;
    comparison of shapes of, 44;
    mutilated, restoration of, 45;
    pitching, 43;
    secondary, 44;
    shapes of, 43.

  Fold topography, 65.

  Forbes, J. D., cited, 294.

  Fore-set beds, 167.

  Forest, destruction of, in relation to agriculture, 156.

  Formation of lava reservoirs, 143.

  Formations, measurement of thickness of, 48, 49.

  Fort Snelling, on Warren River, 327, 331.

  Fosses, glacial, 281, 314;
    in connection with peat bogs, 430.

  Fracture control, of drainage lines, 224.

  Fracture, experiments on, 41;
    of minerals, 450;
    zone of, 40, 46.

  Fractures, in rocks, shown by rectilinear lines on map, 65;
    system of, 55.

  Free, E. E., cited, 222.

  Free waves, 232.

  Fretted upland, 372, 373.

  Frost, prying work of, 152.

  Frost action, 223.

  Frost snow, 285.

  Fuller, M. L., cited, 157, 195.

  Fumeroles, 97.

  Gabbro, 462.

  Gabled façade, in desert landscapes, 221, 443.

  Galenite, 453.

  Gannett, Henry, cited, 178, 386.

  Gaps, water, 176;
    wind, 176.

  Garnet, 459.

  Gautier, E. F., cited, 221.

  Geikie, A., cited, 6, 7, 148, 178, 244, 318.

  Geikie, James, cited, 6, 318.

  Geoid, departure from spherical surface of, 10.

  Geological map, 46, 54;
    areal, 62, 63;
    base of, 61;
    field, 62, 63.

  Geological section, 46, 47.

  Geology, defined, 1.

  Geyserite, 194.

  Geysers, 191-194;
    effect of plugging with sod, 193;
    in relation to drainage lines, 191;
    soaping of, 194.

  _Geysir_, 192.

  Gilbert, G. K., cited, 93, 148, 157, 178, 179, 198, 221, 224, 240,
                         244, 294, 344, 345, 347, 350, 355, 356, 357,
                         358, 359, 362, 366, 370, 381, 434, 446, 447.

  _Gjás_, volcano fissures in Iceland, 99.

  Glacial anticyclone, 284.

  Glacial deposits, 30, 31.

  Glacial fringe, of Grant Land, 285.

  Glacial Lake Agassiz, 325-328, 339.

  Glacial lakes, at close of ice age, 320;
    of St. Lawrence Valley, 329.

  Glaciated regions, aspects of, 302;
    characteristics of, 301;
    contrasted with nonglaciated, 299, 309.

  Glaciation, conditions essential to, 261;
    cycle of, 263;
    Permo-Carboniferous, 298.

  Glaciations, following changes in earth’s figure, 15;
    previous to “ice age”, literature of, 318.

  Glacier broom, over continental ice, 285.

  Glacier cornices, 397.

  Glacier deposits, upon its bed, 390.

  Glacier drainage, 397.

  Glacier flow, 390, 400;
    data from accidents to Alpinists, 392.

  Glacier gravings, 301, 319;
    multiple records, 304.

  Glacier lobe lakes, 411.

  Glacier milk, 398.

  Glacier mills, 278.

  Glacier pavement, 276.

  Glaciers, birth of, 369;
    crevasses on, 391;
    dendritic, 383, 385, 386;
    grinding tools of, 276;
    horseshoe, 383, 386, 387;
    inherited basin, 387-389;
    initiation of, 262;
    in relation to wind direction, 262;
    main types of, 266;
    mountain, cross sections of, 394;
    mountain, expanded-foot type, 264;
    mountain, land sculpture by, 367;
    mountain, successive stages, 383;
    nivation, 387;
    nourishment of, 268-270;
    piedmont, 383, 384;
    radiating, 383, 386;
    sensitiveness to temperature changes, 263;
    séracs, 391;
    surface features of, 390;
    tide water, 290, 386.

  Glacier stars, 395.

  Glacier tables, 395.

  Glacier types, successive, during waning glaciation, 383.

  Glacier wells, 278.

  Glassy texture, of extrusive rocks, 32.

  Glen Roy, 322, 339.

  Glint, 409.

  Glint lakes, 408, 409.

  Gneiss, 465.

  Gneiss banding, 31.

  Goethe, cited on volcano structure, 139.

  Gold, E., cited, 294.

  Goldthwait, J. W., cited, 259, 320, 341, 345, 351.

  Gondwana Land, 16.

  Gorges, through rock bars, 378.

  Grabau, A. W., cited, 361, 366.

  Grading of flood plain, 162.

  Grand Cañon of the Colorado, 146, 169, 174, 215, 443.

  Grand River outlet, 333.

  Granite, 462;
    dome structure in, 152, 157.

  Granite domes, 221.

  Granitic texture, of igneous rocks, 33.

  _Grats_, 373.

  Gravel, kame, 310.

  “Gravel piedmont”, 214.

  Great Basin, 190, 198, 439.

  Great Lakes, probable future of, 347, 348;
    submergence of certain shores of, 349, 350.

  Great Ross Barrier, 282.

  Great Salt Lake, 199;
    fluctuations of level of, 198.

  Green, W. Lowthian, cited, 19.

  Gregory, J. W., cited, 11, 19, 439, 446.

  Grooved upland, 372, 373.

  Gross, H., cited, 294.

  Grossman, cited, 268.

  Grottoes, sea, colors of, 258.

  Ground water, 180;
    descent of, in relation to joints, 181.

  Ground water lakes, 424.

  Grund, A., cited, 195.

  Gullies, early stages of, 160.

  Gulliver, F. P., cited, 244, 319.

  Gullying process, started by deforestation, 156.

  Gypsum, 455.

  Hade, on faults, 59.

  Hague, Arnold, cited, 196.

  Halemaumau, Kilauea, 107, 108.

  Hamilton, Sir William, cited, 128.

  Hanging valleys, 378.

  Hardness, of minerals, 451.

  Harwood, W. A., cited, 294.

  Haug, E., cited, 7, 133, 211.

  Haughton, Samuel, cited, 56.

  Hawaii, lava domes of, 105;
    lava surfaces of, 113;
    map of, 106;
    section through, 106.

  Hayes, C. W., cited, 156.

  Headlands, notched, 341.

  Heave, of faults, 59.

  Hebrews, conception of the universe, 9.

  Hedin, Sven, cited, 221.

  Heilprin, A., cited, 148.

  Heim, A., cited, 54.

  Heligoland, 236.

  Helland, A., cited, 99.

  Hematite, 452.

  Hemicycles, of glaciation, 263, 264.

  Herculaneum, buried beneath mud flows, 139.

  Hess, H., cited, 267, 272, 294, 393, 400.

  High plains, 435;
    origin of, 219.

  Hilgard, E., cited, 222.

  Hinge lines, of uptilt, 344-347.

  Hitchcock, C. H., cited, 106, 147, 434.

  Hobson, B., cited, 120.

  Hogarth, William, cited, 170.

  Hogarthian line of beauty, in landscapes, 170-171.

  “Hog backs”, 442.

  Holmes, W. H., cited, 441.

  Horns, 374.

  Horseshoe glaciers, 383, 386, 387.

  Hot springs, 191;
    colors in, due to algæ, 194.

  Hovey, E. O., cited, 136, 137, 148.

  Hovey, H. C., cited, 183, 195.

  Howchin, W., cited, 298.

  Howe, E., cited, 140.

  Howell, cited, 325.

  Hudson River, narrows of, 174.

  Hudsonian channel, 252.

  Hummocks, on pack ice, 286.

  Humphrey, R. L., cited, 90, 93.

  Humphreys, cited, 404.

  Humus, in relation to weathering, 156.

  Huntington, Ellsworth, cited, 216, 217, 221, 222.

  Hus, H. T. A. de L., cited, 183.

  Hydration, 151.

  Hydrosphere, 8.

  Hypothesis, the value of, 6;
    Laplacian, of the universe, 20.

  Icebergs, 296;
    Antarctic, 292, 293;
    Antarctic, formation of, 292;
    blue, 292;
    manner of formation of, 291, 292;
    northern, 291.

  Ice caps, profiles of, 267, 268;
    sculpture, 380.

  Ice-dammed lakes, 321, 323, 410, 411;
    in St. Lawrence Valley, 339;
    of Scottish glens, 322.

  Ice floes, 287.

  Iceland, fissure eruptions of, 102.

  Ice pyramids, 395.

  Ice ramparts, 431-434;
    manner of formation of, 433.

  Igneous rocks, 30;
    textures of, 32.

  Imlay outlet, 332.

  Inbreak, of lava surface, 107.

  Incised topography, 301.

  Inherited basin glacier, 387-389.

  Interlobate moraines, 314.

  Inter-pluvial periods, 198.

  Intricate pattern of river etchings, 158.

  Intrusive rocks, 32, 462.

  Islands, land-tied, 241;
    steep rocky, due to submergence, 252.

  Isobases, 347.

  Isoclinal folds, 42.

  Isothermal zone of atmosphere, 270.

  Jagger, T. A., Jr., cited, 148.

  Jamieson, T. F., cited, 221, 322, 339.

  Jeannette exploring expedition, 287, 295.

  Jensen, H. I., cited, 110, 113, 147.

  Johnson, D. W., cited, 7, 148.

  Johnson, W. D., cited, 77, 213, 219, 220, 222, 370, 381.

  Johnston-Lavis, H. J., cited, 87, 131, 132, 134, 138, 147, 148.

  Joint blocks, in Niagara limestone, 353.

  Joint plane, seat of frost action, 370.

  Joints, 56;
    effect on surface features, 57;
    closed during earthquakes, 76;
    composite nature of, 58;
    composite groups of, 57;
    disorderly, 57;
    displacements upon, 58;
    master, 56;
    space intervals of, 58;
    sets of, 55;
    system of, 55.

  Joint series, combinations of, 56.

  Joint systems, 66.

  Jorullo, birth of, 96.

  Judd, John W., cited, 116, 118, 139, 148.

  Julien, A. A., 156.

  Jura Mountains, 46.

  Kame gravel, 310.

  Kames, 311, 314.

  Kammerbühl, 139.

  _Karrenfelder_, 188.

  Karst, characters of, 186-187;
    once forested, 188.

  Karst conditions, 195.

  Karst lakes, 422.

  _Katavothren_, 188.

  Katzer, F., cited, 195.

  Kearney, Th. H., cited, 222.

  Kelvin, Lord, cited, 20, 29.

  “Kettle moraines”, 311-314.

  “Kettles” on moraines, 312.

  Kikuchi, Y., cited, 148.

  Kilauea, 101, 106;
    draining of lava in crater of, 108;
    eruption of 1840, 109, 111, 112;
    lava movements in, 106, 107;
    moving platform in crater, 107;
    range in height of lava in, 107.

  King, F. H., cited, 157, 195.

  Knebel, W. von, cited, 185, 195, 258, 260.

  “Knob and basin” topography, 314.

  Knott, C. G., cited, 92.

  Kopisch, August, cited, 258.

  Kotô, B., cited, 92.

  Krakatoa, dissected by eruption, 142.

  Krakatoa, eruption of 1883, 141, 142.

  _Kuppen_, 105.

  Kurische Nehrung, wandering dunes of, 210.

  Laboratory apparatus, for simulation of cinder eruptions, 122.

  Laboratory models, for study of geological maps, 63.

  Laccolites, 143, 441, 442, 447.

  Lacroix, A., cited, 148.

  Lacustrine deposits, 35.

  Lake Agassiz, glacial, 325-328.

  Lake Algonquin, 334, 342.

  Lake Arkona, 332, 333.

  Lake basins, study of, 401.

  Lake Bonneville, 199.

  Lake Chicago, 330, 332, 333.

  Lake Eulalie, draining of, during earthquake, 83.

  Lake Iroquois, 334, 335.

  Lake Maumee, 330, 331, 332, 345.

  Lake Ojibway, glacial, 338.

  Lake stages, in St. Lawrence Valley, 336.

  Lake Warren, 333, 334.

  Lake Whittlesey, 332, 333.

  Lakes, alluvial dam, 423;
    as regulators of air temperature, 431;
    as regulators of river flow, 431;
    as settling basins, 426-428;
    barrier, 420;
    basin range, 402, 403;
    become extinct through wave action, 428;
    border, 399, 414;
    classification of, 424;
    colk, 408, 409;
    continental glaciation, 424;
    coulée, 406;
    crater, 405, 406;
    crescentic, 329, 330;
    crescentic levee, 416, 417;
    currents in, 431;
    delta, 419, 420;
    desert, 424;
    drained by cutting down of outlet, 428;
    dune, 421;
    drained during earthquakes, explanation of, 83;
    earthquake, 404;
    ephemeral existence of, 426;
    extinction by peat growth, 429-430;
    extinction of, in desert regions, 430;
    fresh water, 401;
    glacier lobe, 411;
    glint, 408, 409;
    ground water, 424;
    ice dam, 410, 411;
    intramorainal, about continental glaciers, 279, 280;
    karst, 422;
    landslide, 414;
    morainal, 315, 406, 407;
    mountain glaciation, 424;
    newland, 401, 402;
    ox-bow, 165, 415;
    pit, 315, 407, 408;
    playa, 422;
    raft, 417, 418;
    rift-valley, 403, 404;
    river, 424;
    rock basin, 376, 377, 400, 412;
    rock basin about continental glaciers, 279;
    rôle of, in economy of nature, 430;
    saline, 401;
    salines, 423;
    saucer, 415, 416;
    seasonal, 189, 422;
    side delta, 326, 327, 418, 419;
    sink, 421;
    strand, 424;
    tectonic, 424;
    valley moraine, 400, 413;
    volcanic, 424;
    “wall”, 432.

  Laki, eruption in 1783, 99.

  Laminated structure, of rocks, 31.

  Lamplugh, G. W., cited, 225.

  Land, growth of, from volcanic outflow, 113, 114;
    sliced during earthquake, 80;
    uptilt of, at close of ice age, 340.

  Land areas, concentration of, in northern hemisphere, 11.

  Land sculpture, by mountain glaciers, 367;
    in relation to climatic conditions, 443;
    referable to ice caps, 380.

  Land shields, 15.

  Landslide lakes, 414.

  Land-tied islands, 241.

  Lane, A. C., cited, 148.

  Lankester, E. Ray, cited, 260.

  La Noe, G. de, cited, 7.

  _Lapilli_, 119, 122.

  Laplacian hypothesis of the universe, 20.

  Lateral moraines, 393.

  Lateral movements, deep seated, during earthquakes, 81.

  Lava, 32;
    block, 113;
    composition and properties of, 103;
    discharging from tunnel, 111;
    fluidity of basic, 103;
    movements, in caldron of Kilauea, 107;
    probable origin from shale, 144;
    ropy, 113;
    viscosity of siliceous, 103.

  Lava domes, probable structure of walls of, 112;
    slopes of, 103, 104, 105.

  Lava projectiles, pear-shaped type, 121.

  Lava reservoirs, formation of, 143.

  Lava streams, appearance of, 133, 134.

  Lava surface, 113, 124.

  Law of the desert, 197.

  Lawson, A. C., cited, 92, 260, 351.

  Leads, in pack ice, 286.

  Le Conte, Joseph, cited, 6.

  Leffingwell crater, California, 104.

  Levees, 166.

  Leverett, Frank, cited, 6, 104, 166, 312, 318, 321, 330, 332, 333,
                          334, 337, 339, 344, 345.

  Lewiston escarpment, at Niagara, shaping of, 360-362.

  Libbey, W., cited, 274.

  Life histories, of rivers, 158.

  Light figure, from surface of crystal, 25.

  Lightning, in connection with volcanic eruptions, 130.

  Limbs of faults, 59;
    of folds, 43.

  Limestone, 464;
    origin of, 36;
    sinks, 182.

  Limestone, caverns of, 182.

  Limonite, 452.

  Linck, G., cited, 122.

  Lindenkohl, A., cited, 260.

  Lineaments, 87, 226, 227.

  Line of beauty, Hogarthian, in landscapes, 170, 171.

  _Lithodomus_, borings of, in records of oscillation, 254.

  Lithosphere, a complex of interlocking crystals, 25;
    and its envelopes, 8.

  Littoral deposits, 36.

  Loess, 35, 207;
    erosion of, 208.

  Loessmännchen, 208.

  Lubbock, Sir John, cited, 7.

  Luray caverns, Virginia, 186.

  Luster, of minerals, 450.

  Lyell, Sir Charles, cited, 7, 96, 146, 199, 259, 260, 304.

  _Maare_, 405.

  McGee, W. J., cited, 157, 259.

  Mackinac Island, records of uplift of, 341-344.

  Madison, Wisconsin, 233, 237, 241, 317, 434.

  Magellan, circumnavigation of globe, 9.

  Magma, defined, 30.

  Magnetism, of minerals, 451.

  Magnetite, 452.

  Malachite, 453.

  _Mamelons_, 105.

  Mammoth Cave, 182, 183.

  Mantle, rock, 155.

  Map, contour, nature of, 467;
    of Armorican mountains, 438;
    of barrier beaches, 242-243;
    of bowlder train from Iron Hill, 306;
    of cirques and niches, in Bighorn Mountains, 371;
    of coast lines, 246;
      geological, 54, 61;
      geological, method of preparing, 46, 63;
    of continental divide in Colorado, 377;
    of continental glacier in Victoria Land, 282;
    of Dalager’s nunataks, 277;
    of expanded foot glaciers, 264;
    of front of Green Bay lobe, 317;
    of glacial features, Southern Finland, 315;
    of glacial Lake Agassiz, 325, 326, 328;
    of glaciated area, Europe, 299;
    of glaciated area, North America, 298;
    of ice ramparts on Lake Mendota, 434;
    of inner Sandusky Bay, 350;
    of Kilauea and neighboring slopes, 109;
    of Lake Chicago and later Lake Maumee, 332;
    of Lake Maumee, 330;
    of Lakes Whittlesey and Saginaw, 333;
    of lava outflows on Vesuvius, 1906, 131;
    of lava streams on Mauna Loa, 126;
    of marginal moraines, 312;
    of mountain arcs of Eastern Asia, 438;
    of mountain arc of Sewestan, 436;
    of North Polar regions, 288;
    of part of “fire girdle” of the Pacific, 98;
    of Scottish glens, 322-324;
    of Volcano, 118;
    of volcano belts, 98;
    of Warren River, 326, 327;
    topographical, 61;
    topographical, preparation of, 467, 468;
    topographical, verification of, 469;
    to show dispersion of diamonds in Lake region, 308;
    to show dispersion of peculiar rocks, 305;
    to show distribution of existing glaciers, 263;
    to show formation of shore features, 238;
    to show glaciated areas of Pleistocene period, 297;
    to show reciprocal relation of land and sea, 11.

  Marble, 466.

  Margerie, Emm. de, cited, 7, 54.

  Marginal moraines, 278-280, 311-314.

  Marine clays, as marks of uplift, 253.

  Marine deposits, 35.

  Märjelen Lake, 329, 411.

  Marks, of origin of rocks, 30;
    of uplift, on coasts, 245.

  Marr, John E., cited, 7, 445.

  Martel, E. A., cited, 181, 187, 195.

  Martin, Lawrence, cited, 77, 92, 260, 280, 351.

  Martonne, E. de, cited, 7, 195, 222, 382.

  Massive structure, of rocks, 31.

  Master joints, 56.

  Matavanu, eruption in 1906, 110, 113, 147.

  Mat of vegetation, shield to lithosphere, 155.

  Matthes, F. E., cited, 7, 371, 381.

  Maturity, of upland, 170.

  Mauna Loa, 106;
    eruptions of, 109.

  Meander scars, 165.

  Meanders, entrenchment of, 172, 173, 179;
    stream, 163;
    stream, undermining by, 164.

  Measurement of thickness, of formations, 48, 49.

  Mechanical sediments, 34.

  Medial moraines, 393;
    from nunataks, 274.

  Mediterranean seas, 14.

  Melting, selective, on glacier surface, 394.

  Melville, G. W., cited, 289.

  Mercalli, G., c