Principles of Physical Geology

GOVERNMENT OF INDIA
DEPARTMENT OF ARCHAEOLOGY
CENTRAL ARCHAEOLOGICAL
LIBRARY
Class
J
Call No._ 551 JlQrfc

t
D.G.A 79.
PRINCIPLES OF PHYSICAL GEOLOGY
O p r r ,>
P -
1
; .
1 .
j J
PRINCIPLES OF
PHYSICAL GEOLOGY
ARTHUR HOLMES D.Sc. F.R.S.

Emeritus Professor of Geology and Mineralogy
University of Edinburgh
THOMAS NELSON AND SONS LTD

LONDON EDINBURGH PARIS MELBOURNE
TORONTO AND NEW YORK
THOMAS NELSON AND SONS LTD
Parkside Works Edinburgh 9
36 Park Street London Wi
312 Flinders Street Melbourne Ci
302-304 Barclays Bank Building
Commissioner and Kruis Streets
Johannesburg
Thomas Nelson and Sons (Canada) Ltd

91-93 Wellington Street West Toronto 1
Thomas Nelson and Sons

19 East 47th Street New York 17
Societe Fran^aise d’Editions Nelson

97 rue Monge Paris 5
First published September 1944

Reprinted 1943 (twice), 1946, 1947 , 1948 , 1949
1951 y *95 ? » i953> 1954 ’ 1 955 1956 y [ 957
,
1959
CENTR AL ARCHAEOLOC.IGAI
LIBRARY, NEW UhLHt.
Aoc. No 1 A.43.7..
Daie
Cali No JTZLUJ-hZ ,«•
PREFACE
Like many other teachers of geology and geography in this
country, I have long felt the need for a thoroughly up-to-date
book on Physical Geology. I therefore readily responded when

the publishers invited me book that seemed
to write the sort of
to be required. The method of treatment adopted is one that
twenty years’ experience has shown to be successful in training
students and in holding and developing their interest, even
when, as has often been the case, some of them have come to
the subject without any preliminary acquaintance w'ith scien-
tific principles and methods. For this reason it is hoped that
the book will appeal not only to students and teachers and to
the senior classes in schools, but also to the general reader
who washes to see something of the “ wild miracle ” of the
world we live in through the eyes of those who have tried to
resolve its ancient mysteries.
The earth’s activities may be compared to an intricate
interplay of combined operations, and the results, whether they
be landscapes, natural catastrophes, or materials such as
building stones and fuels, are correspondingly varied. It is
inherent in the character of the subject that the full significance
of any one aspect can be properly appreciated only in relation
to the whole. A
broad preliminary survey has therefore been
presented in Part I, to serve as an introduction to the more
detailed treatment that follows. Part II deals with the outer
earth and only turns aside from tracks already familiar to take
in the more interesting results achieved by recent progress.
Part III is mainly concerned with the activities of the inner
earth and their surface expressions. Thanks to the variety of
detective methods that have been developed during the present
century, the depths are less inscrutable than formerly and the
time is an elementary book of
ripe for the incorporation into
the many
spectacular discoveries that have already been made.
But even so, the frontiers of established knowledge are soon
reached, and wdiat we think we see by peering beyond is merely
v
PREFACE
interpretation,and must on no account be confused with the

solid facts of observation.While I have not hesitated to intro-
duce current views, since these reveal the active growth of the
subject, it should be clearly realized that topics such as the
cause of mountain building, the source of volcanic activity and

the possibility of continental drift remain controversial just
because the guiding facts are still too few to provide a foundation
for more than tentative hypotheses. It is my hope that recogni-
tion of some of the outstanding problems may stimulate at least
a few of my readers them.
to co-operate in the attempt to solve
Professor Alfred Brammall, Dr. Leonard Hawkes and my
wife (Dr. Doris L. Reynolds) have given generous assistance by
reading the first draft of the book, and I gratefully acknowledge
that the subsequent revision owes much to their constructive
criticism and helpful suggestions. For any defects that still
—
remain whether of fact, treatment, judgment or style I am, —
of course, entirely responsible.
As befits the subject, special care has been taken to illustrate
the book as fully and effectively as possible. Of the 262 text-
figures over two hundred have been specially drawn. Some of
the figures are original, others are based on diagrams already
published, and a few have been directly reproduced with due
acknowledgment. In addition to the photographs procured
from professional photographers and press agencies, many
more have been contributed by various friends and official
organizations. It is a pleasure to record my cordial thanks to
the following :
The Director of H.M. Geological Survey, for fifty subjects

selected from the Survey’s unrivalled collection of British
geological photographs. Copyright of these is reserved by the
Crown, and this has been specifically stated beneath eleven of
the reproductions, to indicate that they are included by per-
mission of the Controller of H.M. Stationery Office.
The Director of the United States Geological Survey, and
also the National Park Service and the Department of the
Interior of the United States, for providing batches of photo-
graphs from which many striking and instructive subjects have
been selected.
vi
PREFACE
The Director-General of the South African Air Force, for
supplying the originals of Figs. 69 and 81 and Plate 87a, which
are “ air photographs published under the Union of South
Africa Government Printer’s Copyright Authority No. 489 of
12.11.42,” and of Plates 30, 45a and 88, which were taken by
the Aircraft Operating Company of South Africa Limited.
The Curator of the Belfast Municipal Museum and Art
Gallery for permission to use three photographs (Plates 8a, 17a,
and 35b) from the well-known series taken by the late Mr.
R. Welch.
The Geological Photographs Committee of the British
Association (Fig. 18 and Plate 74b) ;
the Egyptian Government
(Plate 62a and b) ;
the Burma Forest Service (Plate 75b) ;
and the Burmah Oil Company Limited (Plate 77b).

A word of personal appreciation is due to Mr. F. N.
Ashcroft for the privilege of using ten of his superb geological
photographs. Other friends and correspondents to whom I
am indebted for illustrative material are Capt. J. Brown,
Mr. F. A. Bannister, Dr. H. S. Bell, Dr. A. J. Bull, Dr. A. M.
Cockburn, Dr. R. M. Craig, Mr. A. D. Combe, Prof. R. A.
Daly, Dr. D. Griggs, Dr. L. Hawkes, Prof. H. G. A. Hickling,
Dr. W. F. Hume, Prof. A. Lacroix, Dr. C. E. Marshall, Mr.
G. O’Neill, Prof. S. H. Reynolds, the late Dr. R. W. Sayles,
Mr. G. S. Sweeting, Dr. C. T. Trechmann, Dr. G. W. Tyrrell,
Prof. W. W. Watts, Mr. E. J. Wayland and Dr. C. E. Wegmann.
It will be noticed that the illustrations are not listed in the
preliminary pages. When such lists become unduly long, as
would here have been the case, they tend to defeat their pur-
pose. As an alternative which it is hoped will facilitate easy
reference, the plates and text-figures are included in the index.
ARTHUR HOLMES
Durham Edinburgh
July ig42 May igjg.
vii
CONTENTS
PART I A PRELIMINARY SURVEY
I INTRODUCTION
Interpretations of Nature : Ancient and Modern
The Major Fields of Scientific Study
The Scope and Subdivisions of Geoiogy
II THE SHAPE AND SURFACE RELIEF OF THE EARTH 8
The Outer Zones of the Earth 8
Continents and Ocean Floors
The Shape of the Earth .... ... io
!3
Isostasy
The Distribution of Land and Sea
5
17
III THE CHANGING FACE OF THE EARTH

Weathering, Erosion, and Denudation .
Deposition of Sediment
The Importance ot Time
Earth Movements ... . .
»oco
Volcanic and Igneous Acu\

Metamorphism of Rocks
ity
Sunimarv of the Geological Processes

.... .
00
O
O
Isostasy and Geological Processes . Cl
IV MATERIALS OF THE EARTH'S CRUST: MINERALS

Elements and Crystals
Rock-forming Minerals
V MATERIALS OF THE EARTH'S CRUST: COMMON

ROCKS 45
Igneous Rocks 45
Granite 45
Basalt
Classification of the
Sedimentary Rocks
Common Igneous Rocks .... 47
49
53
Sandstone and Shale 53
Limestones 55
Metamorphic Rocks 57
Marble and Crystalline Limestones 57
Slate do
Kinds of Metamorphism 6i
Crystalline Schists 63
Gneisses and Migmatites <4
The Cycle of Rock Change . .
65
VI ARCHITECTURAL FEATURES OF THE EARTH’S CRUST 69

Fracture, Flowage, and Flow 69
Folds
Joints ..............78 7
7^
Faults
Modes of Occurrence of Igneous
The Emplacement of Batholiths
Intrusions .... 82
90
ix
CONTENTS
VII ROCKS AS THE PAGES OF EARTH HISTORY ... 93

The Key to the Past 93
The Succession of Strata 97
The Significance of Fossils 99
The Geological Time Scale ioi
Igneous Rocks and the Geological Time Scale . . . .102
Earth Movements and the Geological Time Scale . . .106
PART II EXTERNAL PROCESSES AND THEIR EFFECTS

\ III ROCK WEATHERING AND SOILS 112
Weathering and Climate 112
Disintegration by Temperature Changes 113
The Role of Animals and Plants 1
15
Chemical Weathering 116
Weathering Residues 119
The Mantle of Rock-waste 121
The Growth and Nature of Soils 122
IX UNDERGROUND WATERS 126

Sources of Ground-water 126
I he Storage and Circulation of Ground-water . . . .127
Springs and Wells 129
Swallow Holes and Limestone Caverns 134
Hot Springs and Geysers 137
Deposition from Ground-wateis 140
X RIVER ACTION AND VALLEY DEV ELOPMENT 143

Some General Considerations 143
Rain Erosion 146
Soil-creep and Landslides '47
Erosion and Transport by Rivers .
150
Grading of Rivers ....

Lengthening and Deepening of Valleys '
5*
'53
Waterfalls
Widening of Valievs ... 156
160
Meandeis
Flood Plains
Deltas
.... 164
167
169
XI DEVELOPMENT OF RIVER SYSTEMS AN D ASSOCIATED

LANDFORMS >73
Tributaries and Drainage Patterns
'73
Shifting of Disides and River Capture .
'75
Superimposed Drainage ....

Escarpments and Related Features
I he Normal C> cle of Erosion

'79
182
185
The Isostatic Response to Denudation .
189
Uplifted Peneplains >
9'
Interruptions in the Cycle of Erosion .
'94
Riser lerraces
>95
Incised Meanders and Gorges
'97
Antecedent Drainage of the Himalayas 200
Alluvial Fansand Cones . . . . 201
x
1
CONTENTS
XII GLACIERS AND GLACIATION 204
Snow Fields and the Maintenance of Glaciers .... 204
Types of Glaciers
The Movement of Glaciers
Surface Features of Glaciers
.... ... 206
209
21
Glacial Erosion 214
Corries and Associated Features
Modifications of Valleys by Glacial Erosion
Glacial Deposits
.... 218
220
226
Graciofluvial Deposits 232
Ice-dammed Marginal Lakes 235
Lakes : A General Summary 240
The Pleistocene Ice Age 245
Suggested Causes of Ice Ages 249
XIII WIND ACTION AND DESERT LANDSCAPES . . .253

The Geological Work of Wind 253
Wind Erosion 235
Coastal Dunes and Sandhills 260
Desert Dunes and Sand Sheets 262
Loess
Weathering and Stream Work in the Desert
The Cycle of Erosion in Arid Regions
.... 267
269
273
XIV COASTAL SCENERY AND THE WORK OF THE SEA 277

Shore Lines 277
Tides and Currents 279
Waves 282
Marine Erosion
The ....
Profile of Equilibrium . .
286
291
....
Transport and Deposition Transverse to the Shore
Transport and Deposition Along the Shore
293
295
Shore Lines of Submergence
Shores Lines of Emergence
Submarine Canyons
...
.... .
.
.
.
301
304
306
XV LIFE AS A ROCK BUILDER ”

Life as a Geological
Marine Deposits
.... Agent
3
3H
313
Pelagic Deposits 316
Geological Interest of the Pelagic Deposits .
3‘9
Coral Reefs and Atolls 321
The Origin of Barrier Reefs and Atolls 325
XVI LIFE AS A FUEL MAKER COAL AND OIL 330

....
:
The Sources of Natural Fuels 33 °

Peat
Coal and its Varieties . . ... 332
334
The Constitution of Coal 33 6
Coal Seams and Coalfields 339
Petroleum 343
The Origin of Petroleum 345
Migration and Concentration of Petroleum 34 8
The Discovery of Oilfields 35i
XI
CONTENTS
PART III INTERNAL PROCESSES AND THEIR EFFECTS

XVII EARTHQUAKES 358
The Nature of Earthquakes
The Effects of Earthquakes
....
.... 358
361
Isoseismal Lines and Depth of Origin .
363
Distribution of Epicentres Earthquake Belts
:
365
Seismographs and Seismic Waves . 368
The Structure of the Earth’s Crust 37i
The Structure of the Deep Interior 373
XVIII EARTH MOVEMENTS MOUNTAIN BUILDING
.... 377
:
The Nature of Orogenic Belts 377

Geosvnclines 379
Structures of Orogenic Belts and their Implications 381
Orogenic Belts of Europe 384
The Appalachians 387
The Western Alps 390
Orogenic Belts of the Alpine Revolution 397
The Orogenic Cycle 401
Present-day Orogenesis 403
The Cause of Mountain Building 406
XIX EARTH MOVEMENTS : PLATEAUS AND RIFT V ALLEYS 414
Surface Expressions of Epeirogenic
Fluctuations of Sea Level
Dislocations of Orogenic Belts
.... Movements 414
417
4i9
The Cordilleran Plateaus of North America 420
The Plateaus and Basins of Central Asia 425
The Plateaus and Basins of Africa
The African Rift Valleys
The Origin of the Rift Valleys
.... 428
432
438
XX VOLCANIC ACTIVITY 443
General Aspects
443
Volcanic Gases 446
Las as
448
Pyroclasts
450
Cones and Other Volcanic Structures .
452
Types of Central Eruptions 460
Kilauea 462
Vesuvius
465
Mont Pelee
468
Krakatao
470
The Distribution of Volcanoes
473
Speculations on the Causes of Vulcanism
478
XXI CONTINENTAL DRIFT 487
Continental and Oceanic Relationships
487
4 aylor s Hypothesis of Continental Drift
Wegener's Hypothesis of Continental Drift 49 °
The Opposing Lands of the Atlantic
.
492
.
498
The Climatic Zones of the Late Carboniferous
INDEX
The Search for a Mechanism .... 499
5°5
5 ”
xii
PART I A PRELIMINARY SURVEY
Chapter I
INTRODUCTION
Interpretations of Nature : Ancient and Modern
The world we live in presents an endless variety of fascinating

problems which excite our wonder and curiosity. The scien-
tific worker, like a detective, attempts to formulate these
problems in accurate terms and, so far as is humanly possible,
to solve them in the light of all the relevant facts that can
be collected by observation and experiment. Such questions
as What ? How ? Where ? and When ? challenge him to
find the clues that may suggest possible replies. Confronted
by the many problems presented by, let us say,an active
volcano (Plate 2), we may ask What are the lavas made
:
of? How does the volcano work and how is the heat gene-
rated ? Where do the lavas and gases come from ? When
did the volcano first begin to erupt and when is it likely to
erupt again ?
Here and in all such queries the question What ? refers
to the stuff things are made
and an answer can be
of,
given in terms of chemical compounds and elements. Not
the elements of ancient philosophers, who considered the
ultimate ingredients of things to be earth, air, fire, and water,
but chemical elements such as oxygen, silicon, iron, and
aluminium.
—
The question How ? refers to processes the way things are
made or happen or change. The ancients regarded natural pro-
cesses as manifestations of power by capricious and irresponsible
gods. In the Mediterranean region, for example, Poseidon
was regarded as the ruler of the seas and underground waters.
As the waters confined below the surface struggled to escape,
r
INTRODUCTION
Poseidon assisted them by shaking the earth and Assuring the

ground. He thus became the god of earthquakes. Typhon,
the source of destructive winds, was a “ many-headed monster
of malignant ferocity ” imprisoned in the earth. From him
the dreaded typhoon takes its name. Pluto was the deity pre-
siding over the fiery regions of the underworld. The eruptions
of lavas and volcanic bombs from Stromboli and Vesuvius
were expressions of his wrath. In Ireland, on the other hand,
legendary giants made things happen. They were stone
throwers and builders. One of them flung the Isle of Man
into the Irish Sea and Lough Neagh represents the place it
was taken from. The Giant’s Causeway (Plate 3b), which is
a terrace carved by the weather and the sea from an ancient
lava flow of columnar basalt, was “ explained ” as the work
of the giant Fionn MacComhal.
To-day we think of natural processes as manifestations of
energy acting on or through matter. We no longer blindly
accept events as results of the unpredictable whims of mytho-
logical beings. Typhoons and hurricanes are no longer inter-
preted as the destructive breath of a wind god they arise
:
from the heating of the air over sun-scorched lands. The

source of the energy is heat from the sun. Volcanic eruptions
and earthquakes no longer reflect the erratic behaviour of the
gods of the underworld they arise from the action of the
:
earth’s internal heat on and through the surrounding crust.

The source of the energy lies in the material of the inner
earth. In many directions, of course, and particularly where
great catastrophes are concerned, our knowledge is still woe-
fully incomplete. Only the first of the questions we have
asked about volcanoes can as yet be satisfactorily answered.
The point is not that we now pretend to understand every-
— —
thing if we did, the task of science would be over but that
we have faith in the orderliness of natural processes. As a
result of two or three centuries of scientific investigation we
have come to believe that Nature is understandable in the
sense that if we ask her questions by way of appropriate observa-
tion and experiment, she will answer truly and reward us with
discoveries that endure.
2
THE MAJOR SCIENCES
The Major Fields of Scientific Study
The questions we ask when faced with a volcano in eruption

are typical of the kinds suggested by all natural phenomena.
—
They indicate that in general terms — scientific investigation
is concerned with the manifestations and transformations of
matter and energy in space and time.
Of all the sciences Physics is the most fundamental, for it
deals with all the manifestations of energy and with the nature
and properties of matter in their most general aspects. It
overlaps to some extent with Chemistry, which is particularly
concerned with the composition and interactions of substances
of every kind in terms of atoms and molecules, elements and
compounds. Biology is the science of living matter. The
nature of life still remains an elusive mystery, completely
inexplicable in terms of matter and energy alone, but living
organisms and their evolution can nevertheless be investigated
scientifically. All other sciences are restricted to more or less
specific fields of interest. Astronomy claims the unfathomed

universe of stars and nebulae as its held of study. It is con-
cerned with the distribution and movements of matter in space
on a celestial scale. Its interest in the earth is limited to the
purely planetary aspects of our globe. Regarded as a daughter
of the sun, the earth may be only an insignificant speck in
the immensity of space. But the earth is also the mother
of life and the home of mankind, and as such we naturally
regard her as the most important of all the celestial bodies.
Thus the earth remains as a special held for investigation,
and to this is devoted the science appropriately known as
Geology (from the Greek Ge, the earth logos, logical speech or
;
science ).
From the earliest days of exploration Geography has been

recognized as the study of the “ home of mankind.” Modern
geography focuses attention on man’s physical, biological, and
cultural environment and on the relationships between man
and his environment. The study of the physical environment
by itself is Physical Geography, which includes consideration
3
BRANCHES OF GEOLOGY
of the surface relief of the globe (Geomorphology), of the seas

and oceans (Oceanography), and of the air (Meteorology and
Climatology). Partly as an offshoot from early geography and
partly from the observations of miners, but mainly from the
work of amateur collectors of minerals, rocks, and fossils, there
developed the more general science of the earth which is
distinguished as Geology.
The Scope and Subdivisions of Geology
Modern geology has for its aim the deciphering of the whole
evolution of the earth and its inhabitants from the time of
the earliest records that can be recognized in the rocks right
down to the present day. So ambitious a programme requires
much subdivision of effort, and in practice
it is convenient
to divide the subject into a number

of branches, as shown in
Fig. 1, which also indicates the chief relationships between
geology and the other major sciences. The key words of the
three main branches are the materials of the earth’s rocky
framework (Mineralogy and Petrology) the geological pro-
;
cesses or machinery of the earth, by means of which changes of

all kinds are brought about (Physical Geology) and finally
;
the succession of these changes, or the history of the earth

(Historical Geology).
The earth is made up of a great variety of materials, such
as air, water, ice, and living organisms, as well as minerals and
rocks and the useful deposits of metallic ores and fuels which
are associated with them. The relative movements of these
materials (wind, rain, rivers, waves, currents, and glaciers ;
the growth and movements of plants and animals and the ;
movements of hot materials inside the earth, as witnessed by

volcanic activity) all bring about changes in the earth’s crust
and on its surface. The changes involve the development of
new rocks from old ;
new structures in the crust ;
and new
distributions of land and sea, mountains and plain, and even
of climate and weather. The scenery of to-day is only the
latest stage of an ever-changing kaleidoscopic series of widely
<390) r o
INTRODUCTION
—
varied landscapes and seascapes. Physical Geology is con-
cerned with all the terrestrial agents and processes of change
and with the effects brought about by them. This branch of
geology is by no means restricted to geomorphology, the study
of the surface relief of the present day, which it shares with
physical geography. Its main interest, as we have seen, is
in the machinery of the earth, past and present, and in the
various by-products, of which the existing surface relief and
the rocks now in process of formation are important examples.
Changes of ail kinds have been going on continuously for
something like 2,000 million years. To a geologist a rock is
more than an aggregate of minerals it is a page of the earth’s
;
autobiography with a story to unfold, if only he can read the

language in which the record is written. Placed in their
proper order from first to last Stratigraphy these pages embody
( ) ,
the history of the earth. Moreover, it is familiar knowledge

that many beds of rock contain the remains or impressions of
shells or bones or leaves. These objects, called fossils, are the
relics of animals or plants that inhabited the earth in ancient
times. Paleontology is the study of the remains of these ancestral
forms of life. Thus we see that Historical Geology deals not only
with the sequence of events brought about by the operation
of the physical processes, but also with the history of the long
procession of life through the ages.
Geology is by no means without practical importance in
relation to the needs and industries of mankind. Thousands
of geologists are actively engaged in locating and exploring the
mineral resources of the earth. The whole world is being
searched for coal and oil, and for the ores of the useful metals.
Geologists are also directly concerned with the vital subject
of water supply. Many engineering projects, such as tunnels,
canals, docks, and reservoirs, call for geological advice in the
and materials. In these and many other ways,
selection of sites
geology is applied to the service of mankind.
Although geology has its own laboratory methods for
studying minerals, rocks, and fossils, it is essentially an open-
air science. It attracts its followers to crags and waterfalls,
glaciers and volcanoes, beaches and coral reefs, ever farther
6
GEOLOGY IN THE FIELD
and farther afield in the search for information about the
earth and her often puzzling behaviour. Wherever rocks are
to be seen in cliffs and quarries, their arrangement and
sequence can be observed and their story deciphered. With
his hammer and maps the geologist in the field leads a healthy
and exhilarating life. His powers of observation become
quickened, his love of Nature in all her moods is deepened, and
the thrill of discovery is ever at hand.
SUGGESTIONS FOR FURTHER READING

A. N. Whitehead
Science and the Modern World. Cambridge University Press, 1927.
W. H. George
The Scientist in Action. Williams and Norgate, London, 1 936.
A. Geikie
The Founders of Geology. Macmillan, London, 1905.
F.D. Adams
The Birth and Development of the Geological Sciences. Bailliere, Tindall
and Cox, London, 1938.
7
Chapter II
THE SHAPE AND SURFACE RELIEF OF THE

EARTH
The Outer Zones of the Earth
As it presents itself to direct experience, the earth can be
physically described as a ball of rock (the lithosphere), partly
covered by water (the hydrosphere) and wrapped in an en-
velope of air (the atmosphere). To these three physical
zones it is convenient to add a biological zone (the biosphere)
The Atmosphere is the layer of gases and vapour which
envelopes the earth. It is essentially a mixture of nitrogen
and oxygen with smaller quantities of water vapour, carbon
dioxide, and inert gases such as argon. Geologically it is
important as the medium of climate and weather, of wind,
cloud, rain, and snow.
The Hydrosphere includes all the natural waters of the outer
earth. Oceans, seas, lakes, and rivers cover about three-
quarters of the surface. But this is not all. Underground, for
hundreds and even thousands of feet in some places, the pore
spaces and fissures of the rocks are also filled with water. This
ground-water, as it is called, is tapped in springs and wells,
and is sometimes encountered in disastrous quantities in mines.
Thus there is a somewhat irregular but nearly continuous
mantle of water around the earth, saturating the rocks, and
over the enormous depressions of the ocean floors completely
submerging them. If it were uniformly distributed over the
earth’s surface it would form an ocean about 9,000 feet in
depth.
The Biosphere, the
sphere of life, is probably a less familiar
conception. But think of the great forests and prairies with
their countless swarms of animals and insects. Think of the
tangles of seaweed, of the widespread banks of molluscs, of
reefs of coral and shoals of fishes. Add to these the incon-
ceivable numbers of bacteria and other microscopic plants and
8
THE EARTH'S CRUST
animals. Myriads of these minute organisms are present in

every cubic inch of air and water and soil. Taken altogether,
the diverse forms of life constitute an intricate and ever-
changing network, clothing the surface with a tapestry that
is nearly continuous. Even high snows and desert sands fail
to interrupt it completely, and lava fields fresh from the craters
of volcanoes are quickly invaded by the pressure of life outside.
Such is the sphere of life, and both geologically and geo-
graphically it is of no less importance than the physical zones.
The Lithosphere is the outer solid shell or crust of the earth.
It is made of rocks in great variety, and on the lands it is
commonly covered by a blanket of soil or other loose deposits,
such as desert sands. The depth to which the lithosphere
extends downwards is a matter of definition it depends on :
our conception of the crust and of what lies beneath. It is

usual to regard the crust as a heterogeneous shell, possibly
about 30 miles thick, in which the rocks at any given level are
not everywhere the same. Beneath the crust, in what may be
called the substratum, the material at any given level appears
to be practically uniform, at least in those physical properties
”
that can be tested. Some authors use the term “ lithosphere
to include both crust and substratum.
The dominant rocks occurring in the crust fall into two
contrasted groups :
(a) Light rocks, including granite and related types, having

an average specific gravity or density * of about 2*7. Chemi-
cally these rocks are very rich in rilica, while a/umina is the
most abundant of the remaining constituents. Since it is
often desirable to refer to them as a whole, such rocks are
collectivelyknown by the mnemonic term sial.
b
( )
Dark and heavy rocks, including basalt and related
types (density about 2*9-3*0) and still heavier rocks (ranging
in density up to about 3*4). In these rilica (40-50 per cent.)
__ ... . . the mass of any volume of the substance

* The specific gravity of a substance
the lnair<if an equal ^lume' of water'
The density of a substance is the mass of unit volume of the substance, generally
expressed as the mass in grams of one cubic centimetre. Since 1 c.c. of water has
a mass of 1 gm., the density of water is 1. In c.g.s. units specific gravity and density
are numerically the same.
9
THE SHAPE AND SURFACE RELIEF OF THE EARTH
the leading constituent, though it is much less abundant
is still
than in granite (70 per cent.). In the heavier rocks of this

group magnesia takes second place, and the whole group is
conveniently known as sima. When it is necessary to make
the distinction, the basaltic rocks are sometimes referred to as
salsima. The sial is the dominant material of the continental
crust down depth of several miles, while the sima forms
to a
the foundations of the ocean floor and extends beneath the
continents. Samples of the sima, represented by basaltic
lavas, are brought to the surface by many continental and
oceanic volcanoes.
Continents and Ocean Floors
The surface of the crust reaches very different levels in

different places. The areas of land and sea floor between
successive levels have been estimated, and the results can be
Fig. 2
Hypsographic curve, showing the areas of the earth’s solid surface between succes-
sive levels from the highest mountain peaks to the deepest oceanic
deeps
.graphically represented as shown in Fig. 2. From this diagram

it is clear that there are two dominant the continental
levels :
platform and the oceanic or deep-sea platform. The slope

connecting them, which is actually quite gentle, is called the
continental slope.
io
EARTH STATISTICS
The continental platform includes a submerged outer

border, known which extends beyond
as the continental shelf,
the shore zone to an average depth of about 100 fathoms or
200 metres. Structurally, the real ocean basins must be
SOME NUMERICAL FACTS ABOUT THE EARTH

Size
Km. Miles
Equatorial diameter of the earth . 12,757 7,926-7
Polar diameter of the earth . 12,714 7,900-0
Equatorial circumference 40,077 24,902
Polar circumference 40.000 24,860
Area
Millions of
Sq. km. Sq. miles
Area of the sea floor (70-78 per
cent.) 361 139-4
Area of the lands (29-22 per cent.) 149 57-5
Total area of the earth . 510 196-9
Volume, Density, and Mass

Millions of
Cu. km. Cu. miles
Volume of the earth . . 1,082,000
. 259,600
Density of the earth . 5'527
Mass of the earth . 5,876 million million million tons
Relief
Metres Feet
Greatest known height :
Mt. Everest . 8,840 29,140 above sea level
Mean
land ....
Average height of the
level of the sur-

825 2,707 V
face (land and sea) 250 820 ) >
Mean level of the litho-

sphere 2,450 7,040 below sea level
Average depth of the sea 3,800 12,460 J? JJ
Greatest known depth :
Swire Deep 10,800 34,430 J9

regarded as commencing, not at the visible shoreline, but at
the edge of the shelf. The basins, however, are more than
full, and the overflow of sea water inundates about 1 1 million
square miles of the continental platform. The North Sea,
the Baltic, and Hudson Bay are examples of shallow seas
(epicontinental or shelf seas) which lie on the continental
shelf. It is of interest to notice that during the Ice Age, when
enormous quantities of water were abstracted from the oceans
to form the great ice sheets that then lay over Europe and
North America, much of the continental shelf must have been
land. Conversely, if the ice now covering Antarctica and
Greenland were to melt away, the sea level would rise and the
continents would be still further submerged.
The continents themselves have a varied relief of plains,

plateaus, and mountain ranges, the last rising to a maximum
height of 29,140 feet (Mt. Everest). The ocean floors, except
locally, are less vigorously diversified than the continents, but
islands and submarine ridges and plateaus rise from their
normally monotonous surfaces, and basins and deeps sink to
more than average depths. The deepest sounding so far made
is 34,430 feet (Swire Deep, off the Philippines). Fig. 2 might
suggest that the greatest deeps are farthest away from the
lands, but such is not the case. The deeps lie close in to the
continental edge, and along the Asiatic side of the Pacific they
are particularly strongly developed.
From the figures given above it is clear that the total
vertical range of the surface of the lithosphere is just over
12 miles. To grasp the true relation between the surface
relief and the earth itself, draw a circle with a radius of 2 inches.
A moderately thin pencil line has a thickness of about 1 /100
inch. If 2 inches represents 4,000 miles, then the thickness of
the outline of the circle represents 20 miles. On this scale the
relief is all contained well within the thickness of a pencil
line.
Nevertheless, the relief very great by human standards,
is
and the question arises how it

is that there are such differences
of level. The earth might very well have been a smooth globe
with a uniform oceanic cover. Just how it comes about that
12
THE SHAPE OF THE EARTH
continental land areas exist at all is still an unsolved problem.
But there is no mystery in the fact that the continents stand

up like platforms above the ocean floor. Like ships riding
light, the continents protrude just because their rocks (sial)
are light compared with the heavier rocks (sima) which

underlie the ocean floor. In the same way, mountain ranges
stand high above the continental platforms because the sialic
rocks beneath them go down to correspondingly greater
depths. High mountains have deep “ roots ” (see Fig. 4).
To understand how these curious facts came to be ascertained,
convenient to begin by considering the effects of gravitation
it is
and rotation on the shape of the earth.
The Shape of the Earth

The first voyage around the world, begun at Seville by
Magellan in 1519 and completed at Seville by del Cano in
1522, established beyond dispute that the earth is a globe.
To-day, aviators could fly round the earth in any direction in
a few days. But long ago the nearly spherical form of the
earth had been inferred from a variety of observations, e.g.
the circular boundary of the earth’s shadow on the moon
during an eclipse, and the circularity of the horizon, wherever
observed, combined with the fact that its distance * increases
with the altitude of the observer.
The reason for the spherical shape of the earth became
clear when Newton discovered the law of gravitation. All the
particles of the earth are pulled towards the centre of gravity
and the spherical shape is the natural response to the maximum
possible concentration. Even if a body the size of the earth
were stronger than steel, it could not maintain a shape such
as, let us say, that of a cube. The pressure exerted by
the weight of the edges and corners would squeeze out
material in depth. Equilibrium would be reached only when
* distance of the horizon in miles is given very closely by the simple

The
expression V3h 2, where h is the altitude of the eye in feet. From an aeroplane
at a height of 20,000 feet, for example, the aviator can see places 174 miles away.
13
the shape and surface relief of the earth
the faces had bulged out, and the edges and corners had sunk
in, until every part of the surface was equidistant from the
centre.
The earth isnot exactly spherical, however. Again it was
Newton who firstshowed that, because of the earth’s daily
rotation, its matter is affected not only by inward gravitation,
but also by an outward centrifugal force, which reaches its
maximum at the equator. He deduced that there should be
an equatorial bulge, where the apparent value of gravity was
reduced, and a complementary polar flattening, where the
centrifugal force becomes vanishingly small. Clearly, if this
were so, the length of a degree of latitude across the equator
would be shorter than in the far north. Expeditions were
despatched to Peru in 1735 and to Lapland in 1736 to test
this idea, and Newton’s deduction was confirmed. If the
surface of the earth were everywhere at sea level its shape
would closely approximate to that of an ellipsoid of rotation
(or spheroid) with a polar diameter of 7,900 miles, nearly
27 miles shorter than the equatorial diameter.
How is it, then, that the earth is not exactly a spheroid ?
The reason is that the crustal rocks are not everywhere of the
same density. Since the equatorial bulge is a consequence of
the relatively low value of gravity around the equatorial zone,
it follows that there should be bulges in other places where
gravity is relatively low ; that is to say, wherever the outer

part of the crust is composed of light sialic rocks. Such places
are the continents. On the other hand, wherever the outer
is composed of heavy rocks (sima) the surface
part of the crust
should be correspondingly depressed. Such regions are the
ocean basins.
The earth is in gravitational equilibrium. If there were
no rotation and no lateral differences in the density of the
rocks, the earth would be a sphere. As a result of rotation
it becomes a spheroid. As a further result of density differences
in the crustal rocks, continents, mountain ranges, and oceanic
basins occur as irregularities superimposed upon the surface
of the spheroid.
14
ISOSTATIC BALANCE
ISOSTASY
For the ideal condition of gravitational equilibrium that

controls the heights of continents and ocean floors, in accord-
ance with the densities of their underlying rocks, the term
isostasy(Gr. isostasios, “ in equipoise ”) was proposed by
Dutton, an American geologist, in 1889. The idea may be
grasped by thinking of a series of woodenblocks of different
heights floating in water (Fig. 3). Theblocks emerge by
amounts which are proportional to their respective heights ;
Wooden blocks of different heights floating in water (shown in

front as a section through the tank), to illustrate the conception of
isostatic balance between adjacent columns of the earth’s crust
they are said to be in a state of hydrostatic balance. Isostasy

is the corresponding state of balance which exists between
extensive blocks of the earth’s crust, which rise to different
levels and appear at the surface as mountain ranges, wide-
spread plateaus, or plains. The idea implies that there is a
certainminimum level below the surface, where the pressure
due to the weight of the material in each unit column of the
crust everywhere the same. This isopiestic (uniform pressure)
is
level may
be regarded as the base of the crust or lithosphere.
The earth’s major relief is said to be compensated by the differ-
ences of density within the crust, and the level where the
compensation is complete, i.e. the isopiestic level, is often
referred to as the level of compensation. Naturally, individual
peaks and valleys are not separately balanced in this way ;
the minor relief features of the surface are easily maintained

by the strength of the crustal rocks. As pointed out on p. 33
15
perfect isostasy is rarely attained, though there is generally a
remarkably close approach.
If a mountain range were simply a protuberance of rock
resting on the continental platform and wholly supported by
the strength of the foundation, then a plumb line such as is —
—
used for levelling surveying instruments would be deflected
from the true vertical by an amount proportional to the
gravitational attraction of the mass of the mountain range.
The first hint that mountains are not merely masses stuck on
Mountain
Range Continental
Plateau
'"-'vF---.-:-- 'Bee an Floor -v-
SIAL :
Density about 2-1
SIMA Density about

: :
'
'
3;0 ~
' ' " - *1
Fig. 4
Diagrammatic section through the earth’s crust to illustrate the relationship
between surface features and the probable distribution of sial and sima in depth.
Based on gravity determinations and exploration of the crustal layers by earth-
quake waves
the crust was provided by the Peru expedition of 1735. Bouguer

found that the deflection of the plumb line by the Andes was
surprisingly small, and he expressed his suspicion that the
gravitational attraction of the Andes “ is much smaller than
that to be expected from the mass represented by these moun-
tains.” Similar discrepancies were met with during the survey
of the Indian lowlands south of the Himalayas. The attraction
of the enormous mass of Tibet and the Himalayas was estimated
to be sufficient to deflect the plumb line by at least 15 seconds
of arc, but the real deflection found by Everest was only
5 seconds of arc. Even more remarkable was the observation
that along the south coast of the Bay of Biscay the plumb line
was actually deflected towards the Bay instead of towards the
Pyrenees.
16
DISTRIBUTION OF LAND AND SEA
Only one physical explanation of these discrepancies is
available. There must be a deficiency of mass in the crustal

columns underlying the visible mountain ranges, i.e. the density
of the rocks must be relatively low down to considerable depths.
The possible density distributions are, of course, infinite.
Fortunately, we know something of the rocks within the crust
and can say what the probable densities are. Moreover, earth-
quake waves can be used to explore the depths (see page
371), and evidence from this source indicates that mountain
ranges have sialic roots going down to depths of 40 km. or
more ;
that under plains near sea level the thickness of the
sial is only 10 or 12 km. and that beneath the ocean floor
;
the sial is either absent or quite thin. Fig. 4 illustrates an

approximation to the structure of the crust in relation to the
surface relief.
The Distribution of Land and Sea
Certain peculiarities in the distribution of land and sea

have aroused discussion ever since the main features of the
earth’s surface were discovered.
1. The marked concentration of land in the northern hemi-
sphere and of water in the southern, combined with a reversal

of this contrast in the polar regions.
2. The occurrence of 81 per cent, of all the land in the
“ land hemisphere,” which has its pole in Brittany and in-
cludes North America, Europe, Asia, and Africa and more than
half of South America and the predominance of water in
;
the corresponding “ water hemisphere,” which has its pole

near New Zealand. The following figures bring out these
differences :
Percentage
of land
“Land hemisphere” . . 49
Northern hemisphere . . 39
Whole earth . 29
Southern hemisphere . . 19
“Water hemisphere” . . 9
i7
3. The southerly extension of the three continental blocks
of South America, Africa, and Australia.
4. The antipodal relation between land and sea. 44*6 per
cent, of the surface has sea opposite sea, but only T4 per cent,
has land opposite land. 95 per cent, of the land is antipodal
to sea.
The so-called Tetrahedral hypothesis —now abandoned
was an ingenious attempt to “ explain ” points 1, 3, and 4.
A tetrahedron is a three-dimensional figure bounded by four
Fig. 5
To illustrate the “ tetrahedral ” distribution of continents and oceans
equilateral triangular faces, like a three-sided pyramid with

a triangular base. Since, for a given surface area, a sphere is
the regular figure with the greatest volume, while the tetra-
hedron is the regular figure with the smallest volume, it was
thought that a contracting globe would tend to shrink towards
a tetrahedral form. Thus, regarding the earth as a contracting
globe, the ocean depressions would— on this view be pro- —
duced by an irregular collapse of the crust on a shrinking
interior the Pacific, Indian, Atlantic, and Arctic oceans
;
corresponding to the four tetrahedral faces, while the conti-

nents would be left as the elevated regions containing the
—
PROBLEM OF THE CONTINENTS

edges and corners (Fig. 5). It is true that the present-day
pattern of land and sea is crudely simulated by a “ tetrahedron”
mounted on an apex representing the South Pole, though it
should be noticed that the tetrahedron has to have one face
corresponding to the Pacific —very much larger than the others.
A fatal objection to the tetrahedral hypothesis is that an
earth with a homogeneous crust could never have contracted
towards a tetrahedral form, the reason being that any approach
to a tetrahedron in a body with the mass of the earth would be
gravitationally unstable and inconsistent with isostasy. The
weight of the upstanding edges and corners (if they ever de-
veloped) would be so great that they would sink in again
until the stable form of a globe was restored. A
very small
globe could contract to a form resembling a tetrahedron ; an
enormous one, like the earth, could not.
Now we have already seen that the continents are essentially
rafts of light sial, surrounded by ocean floors of heavy sima.
It is for this reason that the continents stand high above the
ocean floors. But mere collapse of the crust on a shrinking
interior could not possibly have changed the chemical and
mineral composition of the crust so that the rocks of the
tetrahedral edges and corners became light enough to continue
to stand up, or those of the faces so heavy that they would stay
depressed. Thus we see that contraction could not account
for the distribution of the continents without also accounting
for the distribution of the sial, and this it cannot do.
Our problem is thus carried a stage further back, to that
of the rafts of sial that constitute the continents. The material
of the sial may be regarded as a kind of light slag which
accumulated at the surface during the earth’s consolidation
from a molten state. We should expect it to have accumu-
lated uniformly, so that it would everywhere form the upper
layer of the crust. Where, then, is the sialic material that
is missing from the ocean floors ? Obviously, it must be
either in the earth or outside the earth, and speculative answers
have been based on each of these alternatives.
If part of the sial was removed from the earth, the moon
is the most probable place to look for it. The moon probably
J
9
separated from the earth at a very early stage in the history
of our planet. If a sial shell had already formed before the
moon was born, much of it must have been carried away
when the great rupture occurred. Long ago it was suggested
that the vast Pacific basin might be a relic of the scar that was
left behind, and it has since been discovered that this is by
far the greatest of the regions from which a cover of sial is
lacking. This hypothesis is an attractive one, but unfortun-
ately it has to meet the great difficulty that the moon could
only have separated while the parent planet was still in a
molten state, i.e. before there could have been a sial shell.
By the time the sial had formed the interior would almost
certainly have become too stiff for separation of the moon to
be mechanically possible.
If the missing sial is still in the crust, it must be concentrated
Fig. 6
Convection currents in a
layer of liquid uniformly
heated from below
in the continental rafts. A clue as to how such a concentration

might be brought about is provided by observing the behaviour
of the light scum on the surface of jam which is gently boiling.
The heat from below keeps the jam slowly circulating (Fig. 6).
A hot current ascends near the middle, and turning along at
the surface, it sweeps the scum to the edges, where the current
descends. The scum is too light to be carried down, and so
it accumulates until it is skimmed off. When the earth was
molten it would cool by means of similar circulations. Con-
vection currents would rise in certain places, spread out
horizontally, and then turn down again.There are reasons,
as we shall see in later chapters, for suspecting that a sub-
crustal circulation may still be going on within the earth, though
now at an excessively slow rate. However, when the circula-
tion was vigorous, the horizontal currents spreading out from
each ascending current may have swept certain regions clear
of sial. These would become ocean basins. Where the
20
AN UNSOLVED PROBLEM
horizontal currents of one convecting system met those of
a neighbouring system they would be obliged to turn down,
and the light sial would thus become concentrated in the
regions overlying the descending currents. These regions
would become continents. The “ convection current ” hypo-
thesis is plausible, but as yet it is no more than an intelligent
guess. The origin of the continents remains an unsolved
geological problem.

C. E. Dutton
On Some of the Greater Problems of Physical Geology (1889). Reprinted
in Journal of the Washington Academy of Sciences, Vol. XV., pp.
259-369, 1925 ; and in The Figure of the Earth, National Research
Council Bulletin 78, Washington, 1931.
R. A. Daly
Strength and Structure of the Earth. Prentice-Hall, New York, 1940.
(396 ) 21 3
Chapter III
THE CHANGING FACE OF THE EARTH

Weathering, Erosion, and Denudation
The circulations of matter that are continually goingon in

the zones of air and water, and even of life, constitute a very
complicated mechanism which is maintained, essentially, by
the heat from the sun. A familiar example of such a circula-
tion is that of the winds. Another, more complex, is the
circulation of water. Heat from the sun lifts water vapour
from the surface of oceans, seas, lakes, and rivers, and wind
distributes the vapour far and wade through the lower levels
of the atmosphere. Clouds are formed, rain and snow are
precipitated, and on the land these gather into rivers and
glaciers. Finally, most of the water is returned to the
oceanic and other reservoirs from which it came (Fig. 7).
These circulations are responsible for an important group
of geological processes, for the agents involved —wind, rain,
rivers, and glaciers —act on the land by breaking up the rocks
and so producing rock-waste which is gradually carried away.
Part of every shower of rain sinks into the soil and promotes
the work of decay by solution and by loosening the particles.
Every frost shatters the rocks with its expanding wedges of
freezing water (Plate 4a). Life also co-operates in the work of
destruction. The roots of trees grow down into cracks, and
assist in splitting up the rocks (Plate 4b). Worms and other
burrowing animals bring up the finer particles of soil to the
surface, where' they fall a ready prey to wind and rain. The
soil is a phase through which much of the rock-waste of the
lands must pass before it is ultimately removed. The pro-

duction of rock-waste by these various agents, partly by
mechanical breaking and partly by solution and chemical
decay, is described as weathering.
Sooner or later the products of weathering are removed
from their place of formation. Blowing over the lands the
22
CIRCULATION OF NATURAL WATERS
Fig. 7
The circulation of meteoric water. Part of the water which ascends from the
depths bv way of volcanoes reaches the surface for the first time
;
such water
is called juvenile water to distinguish it from the meteoric water already present
in the hydrosphere and atmosphere
winds pick up dust and sand and carry them far and wide.
Glaciers grind down the rocks over which they pass during
their slow descent from ice-fields and high mountain valleys.
Rainwash and landslips feed the rivers with fragments, large
and small, and these are not only carried away, but are used
by the rivers as tools to excavate their floors and sides. And
in addition to their visible burden of mud and sand, the river
waters carry an invisible load of dissolved material, extracted
from rocks and soils by the solvent action of rain and soil
water, and by that of the river water itself. Winds, rivers, and
glaciers, the agents that carry away the products of rock-
waste, are known as transporting agents. All the destructive
23
processes due to the effects of the transporting agents are
described as erosion (L. erodere, to gnaw away).
It is convenient to regard weathering as rock decay by
agents involving little or no transport of the resulting products,
and erosion as land destruction by agents which simultaneously
remove the debris. Both sets of processes co-operate in wearing
away the land surface, and their combined effects are de-
scribed by the term denudation (L. denudo, I make bare).
Deposition of Sediment
The sediment carried away by the transporting agents is

sooner or later deposited again. Sand blown by the wind
collects into sand dunes along the seashore or in the desert
(Plate 5a). Where glaciers melt away, the debris gathered up
during their journey is dumped down anyhow (Plate 5b),
to be dealt with later by rivers or the sea. When a stream
enters a lake the current is checked and the load of sand and
mud gradually settles to the bottom. Downstream in the open
valley sand and mud are spread over the alluvial flats during
floods, while the main stream continues, by way of estuary
or delta, to sweep the bulk of the material into the sea. Storm
waves thundering against rocky coasts provide still more rock-
waste, and the whole supply is sorted out and widely distributed
by waves and currents. Smooth and rounded water-worn
boulders collect beneath the cliffs. Sandy beaches accumulate
in quiet bays. Out on the sea floor the finer particles are
deposited as broad fringes of sediment, the finest material of
all being swept far across the continental shelves, and even
over the edge towards the deeper ocean floor, before it finally
comes to rest. All these deposits are examples of sedimentary
rocks in the making.
We have still to trace what happens to the invisible load
of dissolved mineral matter that is removed from the land by
rivers. Some rivers flow into lakes that have no outlet save by
evaporation into the air above them. The waters of such
lakes rapidly become salt because, as the famous astronomer
24
PLATE -t
the
Survey
bv
(u'ol.
M.
eiduigcd Anglesey
[II.
uatk
Treiarthen,
with
loots,
boulder,
Received]
nee
of
Icc-tiansported
Copynght growth
Crown
(B)
North
Keswick
rocks,
Ltd.,
igneous
Abraham
Wales
P.
N\
[G.
Ordowcian
Trvfan,
of
of
P'*ak
Frost-shattering
(A)
PLATE J
,n\
(B) tTerminal
, ,
and i ,
lateral
.
[C.S. Gcol. Survey
moraines deposited by the Chisana Glacier,
Alaska
DENUDATION AND DEPOSITION
Halley realized more than two centuries ago, “ the saline
particles brought in by the rivers remain behind, while the
fresh evaporate.” Gradually the lake waters become saturated
and rock salt and other saline deposits, like those on the shores
and floor of the Dead Sea, are precipitated. Most rivers,
however, reach the sea and pour into it the greater part of
the material dissolved from the land. So, as Halley pointed
out, “ the ocean itself is become salt from the same cause.”
But while, on balance, the salinity of the sea is slowly increasing,
much of the mineral matter contributed to the sea is taken
out again by living organisms. Cockles and mussels, sea-
urchins and corals, and many other sea creatures, make shells
for themselves out of calcium carbonate abstracted from the
water in which they live.When the creatures die, most of
their soft parts are eaten and the rest decays. But their hard
parts remain, and these accumulate as the shell banks of
shallow seas, the coral reefs of tropical coasts and islands, and
the grey globigerina ooze of the deep-sea floor. All of these
are limestones in the making. Life, as a builder of organic
sediments, is a geological agent of first importance.
The Importance of Time
It will now be realized that while the higher regions of the

earth’s crust are constantly wasting away, the lower levels are
just as steadily being built up. Evidently denudation and
>
deposition are great levelling processes. In the course of a

single lifetime their effects may not be everywhere perceptible.
Nevertheless they are not too slow to be measured. About
half an inch has already been worn from the outer surface of
the Portland stone with which St. Paul’s Cathedral was built
two centuries ago. Britain as a whole is wasting away rather
faster —
at an average rate of about one foot in three or four
thousand years. At this rate a million years would suffice to
reduce the varied landscapes of our country to a monotonous
plain. Evidently slowly acting causes are competent to produce
enormous changes if only they continue to operate through
sufficiently long periods.
25
Now, geologically speaking,a million years is a com-

paratively short time, just as a million milesis a short distance
from an astronomical point of view. One of the modern

triumphs of geology and physics is the demonstration that the
age of the earth cannot be much less than 2,000 million years
(p. 105). Geological processes act very slowly, but geological
time is inconceivably long. The effects of slow processes
acting for long periods have been fully adequate to account
for all the successive transformations of landscape that the
earth has witnessed.
Earth Movements
has been ample time, since land and
It follows that there
sea came and indeed for the highest
into existence, for Britain,
land areas, to have been worn down to sea level over and over
again. How then does it happen that every continent still
has its highlands and mountain peaks ? The special creation
theory, immortalized in the words :
" When Britain first at Heaven’s command

Arose from out the azure main,”
is not very helpful, yet it does suggest a possible answer. The

lands, together with adjoining parts of the sea floor, may have
been uplifted from time to time. Alternatively, the level of the
sea may have fallen, leaving the land relatively upraised. In
either case there would again be land well above the sea, and
on its surface the agents of denudation would begin afresh
their work of sculpturing the land into hills and valleys. An
additional factor is the building of new land— like the volcanic
islands that rise from oceanic depths —
by the accumulated
products of volcanic eruptions. Each of these processes of
land renewal has repeatedly operated in the course of the
earth’s long history, but it is the first —
the movement of the
crust itself — which
has most commonly and most effectively
rejuvenated the lands and compensated for their recurrent
wastage.
Relative movements between land and sea are convincingly
EARTH MOVEMENTS
proved by the presence in Islay and Jura, and in many other
places, of typical sea beaches, now raised far above the reach
of the waves (Plate 6). Behind these raised beaches the corres-
ponding cliffs, often tunnelled with sea caves, are still preserved. ,
In Scandinavia and Peru old strand lines can be traced which

rise from near sea level in the south to heights of hundreds
of feet in the north. Such tilting of the shores shows that the
movements involve actual upheaval of the crust and not
merely withdrawal of the sea. The former uplift of old sea
floors can be recognized in the Pennines, where the grey
limestones contain fossil shells and corals that bear silent
witness to the fact that the rocks forming the hills of northern
England once lay under the sea. The most spectacular example
of uplift is provided by Mount Everest, the summit of which
is carved out of sediments that were originally deposited on
the sea floor of a former age.

When earth movements take place suddenly they are
recognized by the passage of earthquake waves. In certain
restless belts of the crust, for example in Japan, there may be
several shocks every day, occasionally with terribly disastrous
consequences. Exceptionally, as in Alaska, upward jerks of
more than forty feet have been observed, but usually the
movements are on a smaller scale.
If crustal movements were all vertical, then uplifted beds
of sediment from the sea floor would generally be found lying
in nearly horizontal positions. So, indeed, they often are,
but in many places they have been corrugated and buckled
into folds (Plate 7) in much the same way as a tablecloth
wrinkles up when it is pushed along the table. The layers
of rock seen in the cliffs and on the rocky foreshores of parts
of Devon and Berwickshire have been folded tightly together,
like the pleats of a closed concertina. If the tablecloth is
pushed along stillfurther after a fold has appeared, the fold
will gradually turn over and overlap the flat part that is being
pushed along. So also in the rocks. In many an Alpine pre-
cipice great sheets of rock are visibly “ overfolded ” in just
the same way, so that parts of them now lie upside down (see
Plate 81).
27
Such amazing structures as these show that certain parts
of the earth’s crust have yielded to horizontal compressive
forces of unimaginable intensity. All the great mountain
ranges of the world are carved out of rocks that have been
folded and crumpled and overthrust. Long belts of the crust
have been so squeezed and thickened that they had no alter-
native but to rise to mountainous heights.
The crustal movements have not always and everywhere
been upwards. That parts of the land surface have recently
subsided is proved by the local occurrence around our shores
of submerged forests uncovered only at the lowest tides. But
there seems never to have been a time when all the lands were
submerged at once. Earth movements and volcanic additions
to the surface have evidently been fully competent to restore
the balance of land and sea whenever that balance has been
threatened by the levelling processes of denudation. Most of
the sediments originally deposited on the shallow sea floor,
sometimes hardened and cemented into firm and durable
rocks, sometimes bent and twisted into intricate folds, some-
times accompanied by lavas and volcanic ashes, have sooner
or later been upheaved to form new lands.
Volcanic and Igneous Activity
Earth movements are not the only manifestations of the

earth’s internal activities.Volcanic eruptions provide a most
spectacular proof that the earth’s interior is so hot that locally
even the crustal rocks pass into a molten state. A volcano is
essentially a rift or vent through which magma (molten rock
material highly charged with gases) from the depths is erupted
at the surface as flows of lava or as explosive clouds of gases
and volcanic ashes. Magma may reach the surface through
long fissures, from which lavas spread over the surrounding
country, filling up the valleys and forming widespread volcanic
plains or plateaus.
In the more familiar volcanoes the magma
ascends through a central pipe, around which the lavas and
ashes accumulate to form a more or less conical volcanic
28
IGNEOUS ACTIVITY
Dartmoor : a large granite intrusion now exposed bv denudation and carved

into a landscape of hills and valleys
mountain. Thus volcanic activity is a constructive process

whereby new materials are brought to the surface, and new
topographic forms built up.
Volcanic activity is only the surface manifestation of the
movement through the earth’s crust of magma generated in
the substratum or in exceptionally heated regions of the crust
itself. The new rocks formed in the crust from magma that
failed to reach the surface are called intrusive rocks, to distinguish
them from lavas, which are called volcanic or extrusive rocks.
In many places intrusive rocks are exposed to observation, as
on Dartmoor, as a result of the removal of the original cover
by denudation (Fig. 8). All rocks which owe their origin to
the solidification of magmas in depth (intrusive rocks) or of
lavas at the surface (extrusive rocks) are described as igneous
rocks (L. ignis, fire). This definition of igneous rocks is, how-
ever, not quite complete (see page 67).
29
Metamorphism of Rocks
When crustal rocks come under the influence of (a) the

intense pressure or stress accompanying severe earth move-
ments ;
(b) the increased temperature associated with igneous
activity ;
chemically active gases and liquids from
or (c)
magmatic by changes in structure and

sources, they respond
mineral composition, and so become transformed into new
types of rocks. All such changes are described as metamorphism ,
and examples of them will be considered in the following

chapter. Here the term is introduced to draw attention to the
fact that rocks respond to the earth’s internal activities not
only by crumpling or by fusion, but also by recrystallization.
It must be carefully noticed that metamorphism is the very
antithesis of weathering. Both processes bring about great
changes in pre-existing rocks, but weathering is destructive
w'hile metamorphism is constructive. Instead of reducing a
pre-existing rock to a decaying mass of rock-waste and soil,
metamorphism brings about its transformation, often from a
dull and uninteresting-looking stone, into a crystalline rock of
bright and shining minerals and attractive appearance.
Summary of the Geological Processes

It will now be from our rapid survey of the leading
clear
geological processes that they fall into two contrasted groups.
The first group— denudation and deposition—includes the
processes which act on the crust at or very near its surface,
as a result of the movements and chemical activities of air,
water, ice, and living organisms. Such processes are essentially
of external origin. —
The second group earth movements,
—
igneous activity and metamorphism includes the processes
which act within or through the crust, as a result of the physical
and chemical activities of the material of the substratum and
of magmas formed in or passing through the crust. Such
processes are essentially of internal origin.
30
.
CLASSIFICATION OF PROCESSES
Both groups of processes operate under the control of

gravitation (including attractions due to the sun and moon),
co-operating with the earth’s bodily movements —rotation
about its axis and revolution around the sun. But if these
were all, the earth’s surface would soon reach a state of approxi-
mate equilibrium from which no further changes of geological
CLASSIFICATION OF GEOLOGICAL PROCESSES
I. Processes of External Origin
1. Denudation (Weathering, Erosion, and Transport)

Sculpturing of the land surface and removal of the products
of rock decay mechanically and in solution
2. Deposition
(a) of the debris transported mechanically

(
e.g sand and mud)
b
( )
of the materials transported in solution :
(i) by evaporation and chemical precipitation

(e.g. rock salt)
(ii) by the intervention of living organisms

(e.g. coral limestone)
(c) of organic matter, largely the remains of vegetation

(e.g. peat)
II. Processes of Internal Origin
1. Earth Movements (including Earthquakes)

Uplift and depression of land areas and sea floors ; and
mountain building by lateral compression (folding and
overthrusting) of rocks
2. Igneous Activity
The intrusion of magmas and the extrusion of lavas and
other volcanic products
3. Metamorphism
The transformation of pre-existing rocks into new types
stress, and chemically active
by the action of heat, pressure,
migrating fluids.
1
3
significance could develop. Each group of processes, to be
kept going, requires some additional source of energy. The
processes of external origin are specifically maintained by the
radiation of heat from the sun. Those of internal origin are
similarly maintained by the liberation of heat from the stores
of energy locked within the earth.
Throughout the ages the face of the earth has been changing
its At times its features have been flat and mono-
expression.
tonous. — —
At others as to-day they have been bold and
vigorous. But in the long struggle for supremacy between
the sun-born forces of land destruction and the earth-born
forces of land renewal, neither has permanently gained the
mastery.
ISOSTASY AND GEOLOGICAL PROCESSES
It will now be realized that geological processes bring

about changes that must inevitably upset the ideal state of
isostatic balance which gravitation tends to establish. When
a mountain range carved into peaks and valleys and gradually
is
worn down by the agents of denudation, the load on the under-

lying column of the crust is reduced by the weight of the rock-
waste that has been carried away. At the same time a neigh-
bouring column, underlying a region of delta and sea floor
where the rock-waste is being deposited, receives a corres-
ponding increase of load. Unless a complementary transfer
of material occurs in depth, the two columns cannot remain in
isostatic equilibrium. At the base of the crust the pressure
exerted by the loaded column is increased, while that exerted
by the unloaded column is decreased. In response to this
pressure difference in the substratum a slow flowage of material
is set up, as illustrated in Fig. 9. The loaded column sinks
and the unloaded one rises. This process, whereby isostasy
is restored, is called isostatic readjustment.
The upper part of the substratum consists of hot rock
material which probably differs from the crystalline rocks
seen at the surface by being much richer in gases. Acting
like molecular ball-bearings, the gases facilitate flowage, but
32
ISOSTATIC READJUSTMENT
nevertheless the movement
is extremely sluggish. Moreover,
in some regionsappears that the substratum material is not
it
altogether devoid of strength. It then behaves as a plastic

substance (with a little strength) rather than as a viscous
substance (with no strength). In this case no flowage is possible
Oenudst.cn of
Fig. 9
Section illustrating isostatic readjustment in response to denudation and deposition
until the departure from isostasy is up a pressure

sufficient to set
difference that can overcome the strength. The region con-
cerned will therefore remain slightly out of isostatic balance.
In practice, perfect isostasy is rarely attained, though there is
generally a remarkably close approach.
It may happen that certain processes disturb the pre-
existing isostatic balance much more it can be
rapidly than
restored by deep-seated flowage in the substratum. For
example, when the thick European and North American ice-
sheets began to melt away towards the end of the Ice Age,
about 25,000 years ago, these regions were quickly relieved of
an immense load of ice. The resulting uplifts which then
began are still actively in progress. Far above the shores
of Finland and Scandinavia there are raised beaches which
show that a maximum uplift of nearly 900 feet has already
occurred, and every twenty-eight years another foot is added
to the total all around the northern end of the Gulf of Bothnia.
The region is still out of isostatic balance, and it can be
estimated that it has still to rise another 700 feet or so, before
equilibrium can be reached.
33
A common misunderstanding about isostasy is that it is
responsible for earth movements of all kinds.must, there-

It
fore, be clearly realized that isostasy is only a state of balance ;
it is not a force or a geological agent. It is the disturbance of

isostasy by denudation and deposition, earth movements and
igneous activity, that brings into play the gravitational forces
that restore isostasy. The restoration involves vertical move-
ments of the crust which are additional to the earth move-
ments brought about by the independent activities of the
earth’s interior.

G. A. J. Cole
The Changeful Earth. Macmillan, London, 1925.
E. T. Brewster
This Puzzling Planet. Bobbs- Merrill Co., Indianapolis, 1928.
R. L. Sherlock
Man as a Geological Agent. Witherby, London, 1922.
A. Holmes
The Age of the Earth. Nelson and Sons, Edinburgh, 1937.
34
1
Chapter IV
MATERIALS OF THE EARTH'S CRUST : MINERALS

Elements and Crystals
The vast majority of rocks are aggregates of minerals. Of the

remainder, some, like pumice, are made of volcanic glass,
while others, like coal, are composed of the products of organic
decay. All these ingredients in turn are made of atoms of the
chemical elements. Although ninety-four different elements
are now known, nine of these are so abundant that they
make up more than 99 per cent, of all the many thousands of
rocks that have been analysed. Many of the others, such as
gold, tin, and copper, though extremely rare in oi'dinary
rocks, are locally concentrated in ore deposits that can be
profitably worked.
AVERAGE COMPOSITION OF CRUSTAL ROCKS

(After Clarke and Washington)
In Terms of Elements In Terms of Oxides
Name Symbol Per cent. Name Formula Per cent.
Ox> gen '

O 40-7 Silica SiCX 59 07
Silicon
Aluminium
Si
A
27-09
8-07
Alumina AI
Fe
AG 15-22
Iron oxides * 2 3
1 t
fi
-
81
Iron Fe 5-0.3 \ Fe"0 3T?J
Calcium Ca 3 0.7 Lime CaO .7-10
Sodium Xa 2 7.7 Soda Xa..O 3-71
Potassium K 2 .78 Potash K,0 3-11
Magnesium Mg 2 -OS Magnesia MgO 3-45
Titanium Ti 002 Titania TiO a 1 -03
Hydrogen H OIL Water H,0 1-30
99-34 98-80
Some of the elements, e.g. gold, copper, sulphur, and

carbon (as diamond and graphite), make minerals by them-
* It should be noticed that iron makes two kinds of oxides, distinguished as

ferrous oxide, FeO. and ferric oxide, Fe 2 3 0
If it is desired to discriminate between
.
the two corresponding states in which iron can exist, ferrous iron is represented
by Fe" and ferric iron by Fe'".
35
MATERIALS OF THE EARTH’S CRUST : MINERALS
selves, but most minerals are compounds of two or more

elements. Oxygen is by far the most abundant element in
rocks. In combination with other elements it forms com-
pounds called oxides, some of which occur as minerals. As
silicon is the most abundant element after oxygen, it is not
surprising that silica, the oxide of silicon, Si0 2 should be the
,
most abundant of all oxides. Silica is familiar as quartz, a

common mineral which is specially characteristic of granites,
sandstones, and quartz veins. The formula, Si0 2 is a simple ,
way of expressing the fact that for every atom

of silicon in quartz there are two atoms of
oxygen. Quartz has, therefore, a perfectly
definite composition. The formulae for other
oxides and compounds of other types may be
similarly interpreted.
In the cavities of mineral veins quartz can
be found as clear transparent prisms, each
with six sides and each terminated by a
pyramid with six faces (Fig. 10). The old
Greeks gave the name crystal ( krustallos clear ,
Doubly terminated
tQ t b ese beautiful forms, and to this day
water-clear quartz is still known as rock
crystal (Fig. 11). Most other minerals and a great variety of
chemically prepared substances can also develop into sym-
metrical forms bounded by flat faces, and all of these are now
called crystals. It is an old joke in the world of crystallography
that the beauty of a crystal depends on the planeness of its
faces. In recent years the study of crystals by means of X-rays
has revealed the fact that the symmetrical shape is simply the
outward expression of a perfectly organized internal structure.
The atoms of which a crystal is composed are arranged in an
orderly fashion, the different kinds of atoms being built into
a definite pattern which is repeated over and over again, as
in the design of a wallpaper, except that in crystals the design
is in three dimensions.
Every crystalline chemical compound has a characteristic

unit-pattern, and as the different kinds of atoms are present
in definite proportions, it follows that the crystals of any given
36
)
ATOMIC PATTERNS
[British Museum ( Natural History

Fig. 11
Group of quartz crystals from Minas Geraes, Brazil
compound have a definite and constant composition. It

sometimes happens, however, that two (or more) related
compounds may have the same atomic pattern they are ;
then said to be isomorphous. In isomorphous compounds the

atoms of certain related elements are interchangeable in the
crystal edifice, just as bricks of different colours can be built
into a wall without altering either the structure of the wall
or its outward shape. In “ mixed ” crystals of this kind the
composition is therefore not constant, but ranges between
certain definite limits.
In non-crystalline, or amorphous substances the atoms are
,
arranged haphazardly, like the bricks in a tumbled heap.

Examples of such substances are glass, opal, and limonite.
Only a few non-crystalline substances are regarded, by
common and these are generally distin-
usage, as minerals,
guished as mineraloids. Apart from these, all minerals are
naturally occurring inorganic crystalline substances, each of
which has its own specific variety of crystal structure the ;
chemical composition may be constant (as in quartz) or it

( 396
)
yj ^
MATERIALS OF THE EARTH'S CRUST : MINERALS
may vary (as in the felspars) within limits that depend on the
degree to which the atoms of certain elements can substitute
for those of other elements without changing the specific
pattern of the atomic framework.
Rock-forming Minerals
Although about 2,000 minerals are known, most common

rocks can be adequately described in terms of a dozen or so,
as the following table indicates. It is, therefore, well worth
while to become familiar with these essential rock-forming
—
minerals, and especially to learn something of their chemical
compositions. An attempt is made here to present this mini-
mum equipment of chemical knowledge as briefly as possible.
AVERAGE MINERAL COMPOSITIONS OF SOME

COMMON ROCKS
Igneous Rocks Sedimentary Rocks

Minerals
Granite Basalt Sandstone Shale Limestone
Quartz 31-3 69-8 31-9 3-7

;
Felspars
Micas
....
....
52-3
11-5
46-2
—
8-4
1-2
17-6
18-4
2-2
—
— —
1
!
[
Clav minerals 6-9 10-0 1-0

Chlorite.... J
— 1-1 6-4 —
1
Hornblende 2-4 — — — —
Augite
.
1
rare 36-9 — — —
;
Olivine .... 1
— 7-6 — — —
; Calciteand dolomite — — 10-6 7-9 92-8
Iron ores 2-0 6-5 1-7 5-4 0-1
Other minerals .
0-5 2-8 0-3 2-4 0-3
The student should refer to textbooks for additional informa-

tion, and above all he should handle typical specimens of
minerals and rocks and examine actual rock exposures out of
doors whenever opportunity affords.
38
FELSPARS
The following are the chief examples of oxide minerals :
Qtiartz, s;c) 2 . already referred to above.

Hematite, Fe'”2 3 0 "i All important ores of iron. Magnetite is often
Magnetite, Fe"0.Fe'"2 03 or Fe Oj associated with Ilmenite, FeO.Ti0 especially
3 j
2,
Limonite, Fe'" 0 .H 0 2 3 2 J in basalts and rocks of similar composition
Ice, crystalline water, H O, the mineral of v/hich glaciers and ice sheets are com-
a
posed
Silica combines with one or more of the other common

oxides to form a group of extremely important rock-forming
minerals called silicates. The most abundant of the silicate
minerals are the felspars, nearly all of which are built up of
the three compounds :
Alkali (Orthoclase, Or, KAlSi 3 0 8

felspars \Albite, Ab, NaAlSi 3 O g
Plagioclase
Anorthite, An, CaAl 2 Si 2 0 8
In the Alkali felspars the compounds Or and Ab are combined

in limited proportions, Na and K
being interchangeable to a
limited degree in the crystal structure. Thus the mineral
Orthoclase, the most familiar of the alkali felspars, generally
contains a small amount of albite. Albite and anorthite,
however, are perfectly isomorphous (that is, they can combine
in all proportions), thus forming a continuous series of minerals,
known collectively as plagioclase. In this case the atomic
group interchangeable in the crystal structure
(NaSi)
is
with atomic group (CaAl).

the The formulae for albite
and anorthite can be written in a way that makes clear
the possibility of this substitution :
Plagioclase
Albite Al(NaSi)Si 2 g O
Anorthite Al(CaAl)Si 2 O s.
When albite is more abundant than anorthite the felspar

is called sodic or soda plagioclase. Varieties with anorthite
in excess are distinguished as calcic or lime plagioclase.
Orthoclase can be easily recognized as the cream, pink, or
grey mineral in granite. In some granites, like those of Corn-
wall or Shap (Fig. 12), large slab-like crystals of orthoclase, an
39
inch or more in length, are sprinkled through the rock. When

the crystals are broken across, the surfaces are smooth and glisten-
”
ing. Orthoclase does not break just anyhow it “ cleaves;
along parallel cleavage planes in the crystal structure across

which cohesion is comparatively weak. Just as in many wall-
papers the repetition of the unit pattern gives rise to a parallel
seriesof “ open ” lines, so in the atomic pattern of a crystal
there may be similar “ open ” planes, and it is along these
[G. S. Sweetin°
Fic. 12
A polished surface of Shap granite, showing large crystals of orthoclase em-
bedded in a ground-mass of finer grain. The resulting pattern is described as
porphyritic texture, p. 45
that the crystal splits most readily. Orthoclase has two such
setsof cleavage planes, and the mineral takes its name from the
fact that they are exactly at right angles (Gr. orthos, rectan-
gular clastos, breaking).
; Plagioclase has also two cleavages,
but in this case, though nearly at right angles, they are not
exactly so. Hence the name (
plagios , oblique).
When any of the felspars are decomposed (e.g. by weather-
ing or other processes involving addition of water), the usual
residual products are either (a) a very fine-grained variety of
40
CLAY MINERALS AND MICAS
white mica (see below) called sericite, or (b) a clay mineral, of
which there are several varieties. Most of the clay minerals
are hydrous silicates of aluminium, with formulae such as
Al 4 Si 4 O 10 (OH) 8 or Al 4 Si 4 0 6 (0H) 16 but some varieties (as in
,
fuller’s earth) contain a little magnesium in addition. Under

certain conditions in tropical climates all the silica may be
removed from felspars by weathering. The residue then
left is bauxite, a mixture of two aluminous minerals with the
compositions A1 2 0 3 .H 2 0 and A1 2 0 3 .3H 2 0. Bauxite is of
great value as the only workable ore of aluminium.
Quartz and felspars (orthoclase and sodic plagioclase) are
the characteristic minerals of the sialic rocks, together with
micas, of which there are two leading varieties :
White mica or Muscovite . KAl 2 (Si 3Al)O 10 (OH ) 2

Dark mica or Biotite . . K(Mg,Fe ,, ) 3 (Si 3Al)O 10 (OH) 2 .
The means that

expression (Mg, Fe”) in the formula for biotite
Mg and Fe” are interchangeable in the atomic structure of
the crystals. Similarly, A1 and Fe'” are interchangeable, and
fluorine (F) may take the place of some of the (OH). Biotite
has therefore a considerable range of composition, and the
formula given is merely illustrative of the possibilities. Micas
all have a perfect cleavage, because their atoms are arranged
in parallel layers, and splitting between the layers is very

much easier than tearing across them. Cleavage flakes of
mica are both flexible and elastic. From certain exceptionally
coarse varieties of granite, sheets of mica can be obtained
which are large enough to be used for lamp chimneys and
furnace windows.
Certain rocks of the kinds grouped together as sima (e.g.
basalt) contain calcic plagioclase, but others are free from
felspar. All of the sima rocks, however, are characterized
by the abundance or predominance of heavy, greenish
silicate minerals rich in magnesia and iron oxides, and
therefore commonly known as ferromagnesian minerals.
The leading ferromagnesian minerals are biotite, a group
known as the pyroxenes, a group known as the amphiboles,
and olivine.
41
The common pyroxenes may be regarded as built up ol
compounds such as
CaSi0 3 /"with or without

MgSi0 3 A1 2 0 3 and
j
FeSi0 3 ( Fe 2 0 3
The chief non-aluminous pyroxene is
Diopside . . . Ca(Mg,Fe) (Si 2 0 6 )
The chief aluminous pyroxene, and by far the commonest

member of the group, is
Augite . . . Ca(Mg,Fe,Al)(Si,Al) 2 0 6
The amphiboles are somewhat similar in their range of com-

position, the most essential difference being the presence of
(OH) which, as in mica, may be partly replaced by fluorine
(F). The commonest member of the group is
Hornblende . . . (Ca,Mg,Fe) 4 (Si,Al) 4 0 11 (0H)
Augite and hornblende are readily distinguished by their

crystal forms and cleavages. Both minerals have two well
developed cleavages but in augite the angle of intersection
;
between the cleavage planes is nearly 90 °, whereas in horn-

blende it is nearer 60 ° or 120 °.
Olivine is of simpler composition, its formula being
(Mg,Fe) 2 Si0 4 .The mineral is familiar as the transparent
bottle-green crystals cut as gem under the name peridot.
stones
Rocks in which olivine is the most abundant mineral (generally
in association with other ferromagnesian minerals) are called
peridotite. Nearly all the heavier rocks of the sima group are
peridotites;
the term “ sima ” obviously reflects the chemical
composition of olivine and its ferromagnesian associates.
By a process of alteration involving addition of water,
olivine is changed to serpentine, (Mg,Fe) 6 Si 4 O 10 (OH) 8 ; the
same name is given to rocks formed from peridotites by
similar alteration. The corresponding alteration products of
the ferromagnesian minerals which contain alumina and iron
42
CARBON COMPOUNDS
oxides are green, fine-grained, flaky minerals known collectively
as chlorite.
Carbon though not listed in the table on page 35, is
(C),
only a abundant than hydrogen. It forms an enor-
little less
mous number of compounds, mainly with hydrogen and oxygen.

Some of these are the chief constituents of living organisms,
fuels (wood, peat, coal, and oil), and the organic matter of
soils. dioxide, C0 2 is an important atmospheric gas.

Carbon ,
Many forms of vegetation abstract it from the air or from

natural waters which hold it in solution, and the supply is
restored partly by C0 2 liberated during breathing, organic
decay, and the burning of fuels, and partly by C0 2 given off
during volcanic eruptions. CO
a combines with many other
oxides, forming carbonates, some of which occur as minerals.

The most important of these are :
Carbonate f
Dalcite CaC03 the predominant mineral of limestones
,
rock-forming J Dolomite CaC0 3 .\IgC03 occurring in a carbonate rock,

,
minerals which is itself called dolomite

1
l Chalybite or SideriU FeC03 ,
an important iron ore

A. E. H. Tutton
The Natural History of Crystals. Kegan Paul, London, 1924.
A. B. Dale
The Form and Properties of Crystals. Cambridge University Press, 1 932.
W. H. and W. L. Bragg
The Crystalline State. Bell and Sons, London, 1933.
A. G. Ward
The Nature of Crystals. Blackie and Son, London and Glasgow, 1938.
H. A. Miers ( revised by H. L. Bowman)
Mineralogy. Macmillan, London, 1929.

H. H. Read
Rutley's Mineralogy. Allen and Unwin (Murby), London, 1936.
W. E. Ford
Dana’s Manual of Mineralogy. Wiley and Sons, New York, 1929.
43
L. J. Spencer
A Key to Precious Stones. Blackie and Son, London and Glasgow,
I93 6-
F. H. Hatch
An Introduction to the Study of Ore Deposits. Allen and Unwin, London,
I
9 2 9-
A. M. Bateman
Economic Mineral Deposits. Wiley and Sons, New York, 1942.
T. S. Lovering
Minerals in World Affairs. Prentice-Hall, New York, 1943.
44
Chapter V
MATERIALS OF THE EARTH’S CRUST :
COMMON ROCKS
The few references to rocks that have already been made
suffice to show that rocks may be divided into three major
—
groups igneous, sedimentary, and metamorphic according —
to the processes that were concerned in their origin.
IGNEOUS ROCKS
Granite
In its natural home, granite may be examined in the

tors and valleys of many a rough moor (Figs. 8 and 62),
or in the quarries where it is wrought for its durability
and handsome appearance as a building stone. In most
towns it can be seen as hewn blocks or decorative slabs
and columns.
Granite is a coarse-grained rock composed essentially of
quartz, felspar, and mica. In some examples ( e.g from Aber- .
deen) the interlocking minerals are uniformly distributed, and

all are about the same size. Felspar, mostly orthoclase, is the
most abundant mineral. Gleaming plates of mica (black or
bronze-like biotite, accompanied in some varieties by silvery
white muscovite) can easily be recognised. Between the
felspars and micas the remaining spaces are occupied by
translucent, glassy-looking quartz.
Instead of being uniformly granular, certain granites (e.g.
from Cornwall and Shap Fells, Fig. 12) have a distinctive
pattern or texture, clearly seen on polished slabs, due to the
development of orthoclase as conspicuous, isolated crystals
which are much larger than those of the granular ground-mass
in which they are embedded. This texture is technically
described as porphyritic, a term derived from an old Greek
45
MATERIALS OF THE EARTH’S CRUST : COMMON ROCKS
Basalt pierced by granophyre (a fine-grained variety of granite),

shore N. of mouth of Hvaldal, N. of Lon Fjord, south-east Iceland
word meaning “ purple.” The Romans, prospecting for

decorative stones in Egypt two thousand years ago, came
—
upon a deep purple rock which they called porphyrites lapis —
of such attractive appearance that they actively quarried it
for columns, vases, and slabs. In the course of time the same
name came to be applied to other rocks which contain large
crystals embedded in a finer ground-mass, even though they
lack the purple hue of the original porphyritic rock.
The problem whether granite crystallized from solution
in water, or from a hot molten state, aroused fierce controversy
in the early days of geology. The most influential advocate of
the first view was Werner, of Saxony, who taught his students
that granite was the oldest of rocks, and that it was pre-
cipitated from a primaeval ocean that he supposed to have
covered the whole globe. The second view was first advanced
by Hutton, of Edinburgh, who discovered that granite veins
from some of the Scottish granites penetrated the adjoining
rocks. Hutton thus realised that granite had been intruded
46
ORIGIN OF GRANITE AND BASALT
into still older rocks, and he naturally inferred that it must

have been in a liquid state when it did so (Fig. 13). Aloreover,
he found that the rocks in contact with granite had been baked
and thereby metamorphosed, very much as clay is altered
when it is fired and “ metamorphosed ” into brick. Evi-
dently, then, the granites he investigated had crystallized
not only from a liquid, but from an intensely hot liquid ;
that is to say, from a molten or magmatic state.
Basalt
Basalt is a dark-coloured, very fine-grained rock, which is

of widespread occurrence as lava flows of all geological ages.
Its igneous origin is therefore not in doubt. Nevertheless, like
granite, basalt was also the subject of an early controversy.
Some varieties of basalt contract on cooling in such a way that
the rock cracks into long polygonal columns, most of which
are six-sided, and set at right angles to the base of the flow.
The well-known columns of the Giant’s Causeway in Antrim
and of Fingal’s Cave in Stafl'a (Plate 3) are beautifully devel-
oped examples of this natural masonry. In Saxony similar
columnar basalt occurs in isolated patches, capping the hills.
Now Werner, adopting the general opinion of his day, believed
that the basaltic columns were gigantic crystals. He also
believed that crystals could only grow from aqueous solutions.
From these two erroneous ideas he drew the conclusion that
basalt must be a chemical precipitate from his imaginary
world-wide ocean. In this case, Werner’s mistake was cor-
rected in France. The Puys of Auvergne are a group of extinct
volcanoes, with cones, craters, and lava flows still perfectly
preserved. Here Desmarest found a sheet of black columnar
basalt, underlain by cindery lava and burnt soil. He traced
the basalt across country until it finally led him up to the
crater of one of the Puys. No further proof was needed that
the basalt had been erupted from the volcano as a lava
flow.
The early geologists found it extremely difficult to in-
47
vestigate fine-grained rocks like basalt, for with the limited
means at their disposal they were rarely able to identify the
tiny crystals in such compact materials. In 1851 this difficulty
was overcome by Sorby, a Sheffield metallurgist, who showed
how a slice of rock could be ground down to a film so thin
(about 1 /1000th inch) that it became transparent, thus making
it possible, after mounting the film on a glass slide, to view
Fig. 14
Photomicrograph of a thin section of basalt, showing
plagioclase [white), augite [grey), and ilmenite [black)
X 50
the rock through a microscope, and so to examine the magnified

minerals with ease.
A thin section of basalt prepared in this way has the
appearance illustrated in Fig. 14. Lath-like crystals of a clear
colourless mineral, which is calcic plagioclase, form an ir-
regular open network which extends throughout the rock. The
grey mineral, which is greenish or brownish as seen through
the microscope, is augite. Many basalts also contain olivine,
sometimes abundantly. The black opaque mineral is magnetite
or ilmenite. It is the high proportion of iron in basalt which
is
48
VOLCANIC AND PLUTONIC ROCKS
responsible for the dark colour of the rock, and for the rusty-
looking material (limonite) that encrusts its surface when it
has been exposed to the weather.
Basalt is not always wholly made up of crystals. Varieties
that solidified very rapidly, as a result of sudden chilling,
had no time to crystallize completely. In consequence, the
part that remained uncrystallized had no alternative but to
solidify into black volcanic glass. Crystals may already have
grown in the magma before eruption as lava. In this case
its
the resulting basalt is a porphyritic variety, with relatively

large crystals in a very fine-grained or glassy ground-mass. At
the top of a basalt flow the lava may be blown into a cinder-
like froth by the expansion of escaping gases. Even in the
more compact basalts, gas-blown cavities of various sizes may
occur. These may be empty ;
or lined with crystals, often
beautifully developed ; or even completely filled with minerals.
The filled “ bubbles ” sometimes look like almonds, and so the
name amygdale is given to them (from the Greek amygdalos,
an almond). Basalts which are studded with numerous
amygdales are called amygdaloidal basalts. One of the
commonest minerals found in amygdales is agate, banded in
concentric layers of different tints. Agate is a variety of
silica, and sometimes, inside a lining of agate, crystals of
quartz or of amethyst (purple quartz) may be found projecting

into the hollow space within.
Classification of the Common Igneous Rocks

Granite and basalt, and types closely related to them, are
by far the most abundant of the igneous rocks. Granite is an
example of the coarse-grained rocks that crystallized slowly in
large masses within the crust. Such rocks are described as
plutonic, after Pluto, the god of the underworld. Basalt, on the
other hand, is an example of the very fine-grained or glassy
rocks that cooled quickly from lavas that flowed over the
surface. These are volcanic rocks, after Vulcan, the god of fire.
Between the two extremes there are rocks of intermediate grain
49
that cooled and crystallized at intermediate rates, generally in
small intrusions like the dykes and sills described on page 83.
Such rocks are distinguished as hypabyssal (intermediate depths).
Many of the hypabyssal rocks of sialic composition are re-
markable for a very conspicuous development of porphyritic
texture, in consequence of which they are known by the
familiar rock name porphyry.
The simple relationship between mode of occurrence and
grain size indicated above does not completely cover all the
possibilities, because (a) the outer skin of a thin dyke or sill
may be chilled against cold wall-rocks almost as quickly as
the upper layers of a lava flow against the air and ( b ) the ;
rate of cooling in the interior of a thick lava flow may be as

slow as it is within the smaller dykes and sills. However, pro-
vided that such inevitable exceptions to the general rule are
not forgotten, the following correlations serve as a useful basis
for the classification of igneous rocks :
MODE OF OCCURRENCE TEXTURE

Extrusive Lava flows Volcanic
Glassy v and
Very fine-grained |
porphyritic
Intrusive )
Minor intrusions Hypabyssal Fine-grained I varieties
1 Major intrusions Plutonic Coarse-grained J of each
According to scheme, the rocks of any given composition

this
may have any one of eight different textures. In the case of
rocks of granitic composition, the eight textural types are dis-
tinguished by the following names :
Obsidian and Pumice (frothy) Porphyritic Obsidian

Volcanic
l Stony Rhyolite Porphyritic Rhyolite
Hypabyssal Felsite Quartz-porphyry
Plutonic Granite Porphyritic Granite
In the case of rocks of basaltic composition, the corresponding

types are tachylyte (the glassy form of basalt), basalt, dolerite or
diabase (the common rock of many dykes and sills, much used
for road metal), and Gabbro, together with porphyritic
varieties of each.
Igneous rocks are also classified according to the kinds and
50
MINERALS OF IGNEOUS ROCKS
proportions of their constituent minerals. Many of the rocks

commonly referred to as “ granite ” have more plagioclase
than orthoclase and are therefore not true granites it is ;
convenient to distinguish them by the name granodiorite,

because they are intermediate in mineral composition between
granite and quartz-diorite. Diorite, with little or no quartz,
is made up essentially of plagioclase and hornblende. Many
large
“ granite ” masses, consisting mainly of granite and
granodiorite, pass marginally into quartz-diorite and diorite.
'ation in mineral composition and silica content of the rock s

granite-diorite. Accessor)’ minerals in black (at the base)
The continuous mineral variation, and the conventional

dividing lines between the types, are shown graphically in
Fig. 15.
As this example indicates, igneous rocks can be classified
mineralogically, and therefore identified, by means of criteria
such as :
(a) the presence or absence of quartz
(b) the kinds of felspar and the proportions between

them (and in some cases by the absence of felspar),
and
(,c
)
the kinds of ferromagnesian (F.M.) minerals.
5i
u U' *
MINERALOGICAL CLASSIFICATION OF THE CHIEI

IGNEOUS ROCKS
Volcanic or very fine-grained types in italics
Hypabyssal or fine-grained types in ordinary type
PLUTONIC or COARSE-GRAINED types in CAPITALS
Orthoclase SODIC SoDIC

> THAN Plagioclase Pi agioclase
SODIC > THAN predominant
Plagioclase Orthoclase
Rhyolite Rhyodacite Dacite

Quartz 1
essential
Quartz- Granodiorite- Quartz-
F.M.
porphyry porphyry porphyrite
Biotite or
Hornblende
OR BOTH
GRANITE GRANODIO- QUARTZ-
RITE DIORITE
Orthoclase SODIC
Orthoclase and Plagioclase No
PREDOMINANT Plagioclase predominant Felspar
roughly equal
1
Little or no Trachyte Trachyandesite Andesite

Quartz
F.M. Porphyry Monzonite- Porphyrite
Hornblende porphyry
and/or Biotite
AND/OR SYENITE MONZONITE DIORITE HORN-
Augite BLENDITE
Calcic
Plagioclase No
Felspar
predominant
Little or no 1
Basalt
Quartz Dolerite or
F.M. Augite '
Diabase PYROXEN-
and Iron Ores '
GABBRO ITE
i
No Quartz I
1
Olivine-basalt
F.M. Augite, Ol. -dolerite
Olivine, and or Ol. -diabase
Iron Ores OLIVINE- PERIDOT-
GABBRO ITE
52
SANDSTONE AND ITS STRUCTURE
SEDIMENTARY ROCKS
Sandstone and Shale
Sandstone is perhaps the most familiar of all rocks, for

it is easily quarried, and it is used more than any other kind of
stone for building purposes. Examined closely, using a lens
if necessary, a piece of sandstone is seen to consist of grains
Fig. 10
Photomicrograph of sandstone, Torridon Sandstone,

North-West Highlands of Scotland X 30
of sand identical in appearance with those that are churned

up by the waves breaking on a beach. Most of the grains
consist of more or rounded grains of quartz, but there
less
are others of cloudy,weathered-looking felspar, and gen-

erally a few shining spangles of mica can be seen (Fig. 16 ).
Clearly, sandstone is made of second-hand materials, of
worn fragments derived from the disintegration of some older
rock, such as granite, which contained the same minerals. It
(396>
53 5
differsfrom deposits of modern sands only in being coherent
instead of loose. Calcite is a common cementing material.
Brown sandstones are cemented by limonite and red varieties
by haematite. In pure white, extremely hard sandstones, the
cement is quartz. These materials were deposited between
the grains by ground-waters which percolated through the
sand when it was buried under later sheets of sand or other
formations.
In the steep face of a quarry or cliff, successive beds or
layers can be seen, differing from one another by variations

in colour or coarseness of grain (Plate 8a). At intervals there
may be strongly marked bedding planes, along which the
sandstone is easily split, due perhaps to the presence of a thin
layer packed with flat-lying flakes of mica, or to the intervention
of a thin band of clay or shale. Evidently the beds or strata
have been formed by the deposition of layer after layer of
sediment.
Along the beach, and especially near the cliffs, boulders
and pebbles are heaped up by storm waves. Then come the
sands, and beyond them on the sea floor lie still finer deposits
of mud, made up of the tiniest grains of quartz and altered
felspar, shreds of mica, and minute flaky particles of clay
minerals. Each of these different types can be recognized
among the stratified rocks. Sooner or later, a sheet of sand-
stone thins out and passes laterally into clay or shale. Traced
in the other direction, it may become coarser in grain and pass
into a massive boulder bed or conglomerate (see Plate 21b).
The term conglomerate is applied to cemented fragmental rocks
containing rounded fragments such as pebbles if the frag-
;
ments are angular or sub-angular, the rock is called breccia.

The very fine-grained sedimentary rocks, corresponding
to mud, are described as clay, mudstone, or shale. Mudstone
iscompact, but shale can easily be divided into thin laminae.
This fissility is due to a structure resembling stratification on
a fine scale, which is distinguished by the term lamination. The
micas and clay minerals all occur as minute films which lie
with their flat surfaces parallel to the bedding, and in conse-
quence the shale readily divides along the lamination planes.
54
USES AND STRUCTURES OF LIMESTONES
Limestones
Limestones of suitable quality are widely used as building

stones, because of the ease withwhich they can be worked, and
some varieties, the aristocrats of a very mixed group, have
become famous through their lavish use in great public
buildings. Portland stone, for example, has been a favourite
choice for many of London’s greatest buildings, ever since
Wren selected it for the rebuilding of St. Paul’s Cathedral
after the Great Fire of 1666. The towers and steeples of the
London churches, the government offices in Whitehall, the
Buckingham Palace, and most recently of all, the new
front of
home of London University, all display its use in varied styles
of architecture.
The natural architecture of limestones can be studied in
the quarries of Portland and the Cotswolds, in the grey scars
of the Pennines (Plate 8b), and in the white gashes cut by
lime-makers in the green slopes of the Chalk Downs. Some of
the limestones of the Pennines are packed with the remains of
corals and marine shells, and the bead-like relics of the stems
of sea-lilies (animals like starfishes on long stalks Plate 9a). —
The yellow Cotswold limestones, more open and porous, are
often crowded with fossil shells, while belemnites, looking like
thick blunt pencils, and the coiled forms of ammonites add
further interest and variety (see Fig. 50). The fine-grained
and friable limestone known as chalk contains smooth white
shells and sea urchins. But besides these visible fossils there
are innumerable little coiled or globular shells which can be
seen only by microscopic examination ;
these are exacdy
likethe foraminifera dredged up from the deep-sea floor,
where globigerina ooze is collecting at the present day (see
Plate 72).
Evidently many limestones are accumulations of organic
remains, vast cemeteries in which the teeming life of the sea
has been entombed. It is easy to prove with dilute acid that
shells and corals and limestones are all composed of calcium
carbonate. They effervesce briskly when the acid is applied
55
and give off carbon dioxide, while a drop of the resulting
solution held on a clean platinum wire in the flame of a bunsen
burner colours the flame with the brick-red tint due to calcium.
Soft-bodied organisms abstracted the calcium carbonate from
sea water and deposited it around or within themselves, to
form a protective covering or a supporting framework.
The limestones of Portland and Bath also contain shells,
but they are mainly composed of rounded grains that look like
Fig. 17
Photomicrograph of oolitic limestone, Farley, Bath
X 30
insects’ eggs. For reason they are called oolites or oolitic

this
limestones (Greek oon, an egg). Under the microscope each
granule is found to be made of concentric layers of calcium
carbonate, often with a bit of shell at the centre (Fig. 17).
Along coral strands, where the conditions are favourable,
oolitic grains are forming to-day around such nuclei by the
deposition of calcium carbonate from sea w'ater. Rolled about
by the surf, the grains tend to grow equally on all sides, and so
to become round.
Limestones are thus seen to be deposits formed from dis-
56
[ff.M. Geol. Survey
(B) The Great Scar Limestone Carboniferous) above
( Malham Cove,
Yorkshire
ir. n n ntt>
.
( \) Weathered surface of crinoidal limestone Carboniferous . 1 hntshiie
s. //. h\\K"!Us
(B; I mvMorv in the making: a deposit of dir 11 -gravel mainh cockles-
LIMESTONE AND DOLOMITE
solved material :
generally, but not always, from sea water.
They usually accumulate outside the stretches of sand and
mud that in most places border the lands, but where the sea is
uncontaminated by muddy sediment, and especially where the
cliffs themselves happen to be made
of older limestones, they
may form close up to the land.Beaches may locally be com-
posed of sand made up, not of quartz grains, but of shell
debris (Plate 9b). Similarly, the sands associated with coral
reefs and atolls consist largely of coral debris, ground down by
the waves.
Many limestones are distinctly fragmental, whether formed
from organic remains or from oolitic grains. Others are
extremely fine-grained and compact, and most of these repre-
sent chemical precipitates from waters rich in calcium car-
bonate. The double carbonate of calcium and magnesium
(dolomite) may also be precipitated either directly, or by
;
the action of warm sea water on limestone already in process

of accumulation on the floor of the sea. In these ways magnesian
or dolomitic limestones, grading into pure dolomite rocks, have been
formed. Dolomite was selected as the building stone for the
Houses of Parliament. Unfortunately, the sulphurous fumes
from the neighbouring potteries of Lambeth corroded the
delicate carvings of the structure, turning the magnesia of the
stone into easily soluble epsom salts. Dolomite, however, may
be an excellent building stone where it is not exposed to such
abnormal weathering conditions.
METAMORPHIC ROCKS
Marble and Crystalline Limestones
Limestones are known commercially as “ marble ” when
they can be effectively polished and used for decorative pur-
poses. Corals and the stems of sea-lilies give a variegated
pattern commonly seen on polished slabs cut from the grey
limestones of the Pennines. The pink and grey limestones of
Torquay have been fissured by earth movements, and per-
57
Metamorphism of Carboniferous limestone by the Great Whin Sill,
Falcon Clints, Teesdale, Co. Durham
meating solutions have mottled the stone in tints of red and

filled the cracks with white veins of calcite. This example
illustrates the ease with which limestone becomes changed by
natural processes into “ marbles ” of endless variety. Lime-
stone is chemically a very sensitive rock, and the walls of many
of Lyons’ restaurants show in spectacular fashion how readily
it has responded to the effects of heat, pressure and percolating
waters.
To the geologist, however, the term marble is restricted to
limestones which have been completely recrystallized by meta-
morphic processes during their burial in the earth’s crust.
Under the influence of heat from igneous intrusions the calcium
carbonate of shells and finer particles alike is gradually re-
constructed into crystals of calcite of roughly uniform size.
All traces of fossils are destroyed, and the rock, when pure,
becomes a white granular rock like the well-known statuary
marble from Carrara in Italy. The great sheets of basalt in
58
:
METAMORPHISM OF LIMESTONES
Antrim over a surface of Chalk (Plate 17 a), and the chalk,
lie
instead of being soft and friable like that of southern England,

has become hard and indurated by the heat from the over-
lying lavas. And against the channels through which the lava
reached the surface, the chalk is transformed locally into
saccharoidal marble.
It may be asked how it is that carbon dioxide did not
escape under such conditions, as it does when limestone is
heated to make lime. The explanation is that when the heating
takes place under pressure, as when the limestone is confined
under a load of overlying rocks, the carbon dioxide is not
liberated as a stream of gas, but only as dispersed molecules,
temporarily freed. These mobile molecules may be thought
of as acting like tiny ball-bearings, lubricating the rock during
the recrystallization of its minerals. Conditions favourable to
marble formation were successfully imitated in 1805 by Sir
James Hall, an eminent friend of Hutton’s, who tested many of
the rival theories around which controversy was then raging
by devising experimental methods of attacking the problems.
Hall enclosed pounded chalk in a porcelain tube, which in turn
was fitted closely into a cylindrical hole bored in a solid block
of iron, and securely sealed at the open end. On heating this
“ bomb ” to the highest temperature at his command, the
chalk was transformed into a fine granular marble.
When the original limestone is not pure, but contains other
carbonate minerals such as dolomite, and impurities of sand or
clay, various chemical reactions take place between the ingre-
dients when they are heated up sufficiently. New minerals
are then developed, olivine and garnet being examples, and
the carbon dioxide thus liberated is driven off
2CaMg(C03 2 +Si0 2 = Mg2 Si0 4 + 2CaC03

) + 2CO»
Dolomite Quartz Olivine Calcite Carbon dioxide
Moreover, when a limestone or dolomite is invaded by granitic
or other magmas, part of the magmatic material may diffuse

into the surrounding rock, thus adding new constituents and
so causing the growth of still other types of new minerals. The
beautiful green serpentine-marble of Connemara, familiar to
59
all who have probably originated in this way.
visited Ireland,
In some localities in the United States and Sweden important
masses of iron ore have been formed at the expense of lime-
stone. Evidently magmatic fluids rich in iron migrated into
the limestones, leaving the iron there as black magnetite, and
carrying off in solution the replaced calcium carbonate.
Slate
The world-famous roofing slates that are quarried from the

rugged hills around Snowdon owetheir value to a structure
whereby they can be along planes that are not parallel
split
to the bedding, like the lamination of shales, but inclined to
often at a high angle. This structure is neither stratification

it,
nor lamination, but a fissility or cleavage, often of great per-

fection. It is distinguished as slaty cleavage (Fig. 19). It must
not be confused with the cleavage of crystals, for the latter
is a property depending on the orderly
arrangement of atoms,
whereas slaty cleavage depends on the orderly arrangement
of minute flaky minerals such as mica, clay, and chlorite within
the rock. A thin section of slate cut at right angles to the
cleavage shows all the flaky minerals lying parallel
to the
cleavage planes. It is due to this orientation
that the rock
splits along these planes and along
no others.
60
ORIGIN OF SLATE
Traces of the original bedding planes can be made out in

the quarry face wherever bands of contrasted colour or more
gritty material are present ; such bands are often badly
crumpled and contorted. If, by a rare chance, a fossil is found,
it, too, is deformed and squeezed out of shape. Clearly an
overpowering pressure has been exerted on the rock, and
observation shows that it must have been applied sideways.
Lateral earth-pressure, sufficiently intense to crumple the rocks
in most intricate fashion, has compressed a thick mass of com-
pact mudstone or laminated shale so that the flat surfaces
of all the flaky minerals shifted round into positions approxi-
mately at right angles to the direction from which the
pressure came.
In 1856 Sorby demonstrated the truth of this explanation
by experimenting with a stiff mixture of clay and tiny flakes
of iron ore. The flakes orientated themselves with their flat
surfaces perpendicular to the direction of the pressure. Tyndall
later developed cleavage in wax by pressure. In this case the
new structure was due to the flattening out of the original
globular particles of wax. So we find that slates may also be
made out of certain very fine-grained materials even if they
contained no flaky minerals to begin with. The silvery green
slates of the Lake District have been formed from beds of fine
volcanic ashes. Their cleavage and silvery sheen is due, how-
ever, not merely to the flattening of the original particles, but
also to the development of new minerals from the volcanic
materials. Microscopically small shreds of mica and wisps of
chlorite have grown in the rock, all with their film-like surfaces
lying parallel to the cleavage direction.
Kinds of Metamorphism
”
Slate thus an example of a rock on which a new “ grain
is
has been impressed, mainly by the effect of shearing caused by

severe compressional earth movements. To some extent
chemical changes promoted by heat and solution have co-
operated in facilitating the growth of new minerals. The
61
,

original shale or volcanic ash has responded by becoming a
slate, a new type of rock. Since the main process is dynamic,
slate is said to be a product of dynamic metamorphism.
The change from limestone to marble, on the other hand,
Isochnally
folded fTTT r« i Igneous Mefamorphic
Ordovician 1223 Rocks I
I Aureole
and Silurian beds
Fig. 20
Aureoles of contact metamorphism
around small batholiths and stocks,
Galloway, S.-W. Scotland
ismainly brought about by the

action of heat. It illustrates
he effects of thermal metamorphism.
The rocks in contact with
Sarin^fluids
migrating fluids,
S
^
and
C ° mmonl >'
metamorphosed by heat and
metamorphism of this kind is dis-
tinguished as contact metamorphism.
The zone of altered rock
surrounding the intrusion is described as the metamorphic
62
CONTACT AND REGIONAL METAMORPHISM
aureole (Fig. 20). Where shales or slates are in contact with
granite the mineral changes promoted by the rise of temperature
become visibly conspicuous. Traced from the edge of the
aureole towards the granite, the rocks begin to be variegated
by little spongy spots which as yet are hardly individualized
as definite minerals. Nearer the granite these commonly
develop into a glistening felt of tiny brown and white
flakes, which further on become larger, and can be recognized
as micas. Close to the contact other new minerals may
appear. Metamorphic rocks of this kind are called homfels
(plural, hornfelses).
When all the agencies of metamorphism operate together,
as theydo in the heated depths of a crustal belt where mountain
building movements are in progress, the rocks throughout an
extensive region are characteristically transformed, and the
metamorphism is then described as regional.
Crystalline Schists
When shale or slate is by regional meta-

recrystallized
morphism and migrating fluids lead to the
the effects of heat
development of mica and other new minerals on a visible
scale, as in contact metamorphism. At the same time the
effects of shearing or flowage give the rock a new structure,
due to the stream-lined arrangement of the platy and elongated
minerals. This structure is called foliation, a term based on its
similarity to that of the tightly packed leaves of leaf-mould.
Foliation may develop along an earlier cleavage, but more
commonly it follows the stratification of sediments that have
not been cleaved, and in some cases, when the shearing or
flowage failed to coincide with the earlier parting planes, it
follows a new direction altogether. The surfaces along which
a foliated rock can be divided may be plane to undulating,
wavy or contorted. When the foliation is closely spaced
throughout the body of the rock, so that almost any part of it
can be split into flakes or flat lenticles, the rock is called a
schist (Gr. schistos, divided). The schistosity of the rock that —
63
is,the property whereby it can be easily split is somewhat —

akin to slaty cleavage, but on a coarser, rougher, and less
uniform scale.
Schists are named mineral responsible for
after the chief
the foliation in any given case. When
this is mica, as fre-
quently happens, the rock is called mica-schist, this being the
type that develops from shales and slates. Another type is
hornblende-schist, which develops from basaltic rocks, the
hornblende being formed at the expense of the original augite.
Sandstone and limestone rarely form schists, and then only
crudely, the reason being that quartz and calcite are naturally
granular minerals. The usual metamorphic equivalents of
these rocks are quartzite and marble, respectively, both granu-
litic rocks.
Gneisses and Migmatites
One of the most characteristic types of metamorphic rocks

is that known by the old Saxon miners’ term gneiss. Gneisses
are found associated with schists and granites, particularly in
regions where very ancient rocks are exposed. Like schist,
gneiss is a foliated rock, but its open and interrupted.
foliation is
Highly micaceous layers alternate with bands or lenticles or

“ eyes ” that are granular, and more like granite in their
texture and composition. Many gneisses, indeed, have the
mineral composition of granite or granodiorite, and thus they
further differ from schists in containing felspar as an important
mineral, especially in their granular portions.
Some gneisses may be simply the squeezed-out and re-
crystallized equivalents of granitic rocks, but most examples
appear to represent transitions between schist and granite.
Ifwe examine areas where gneisses are well exposed, such as
the low, wave-swept islands of southern Finland (Plate
10),
we can see that the older schists
seem to have been impregnated
by granite in every conceivable
way. In some places veins and
tongues of granite magma appear to have run between
the
folia of the schists, like water seeping
between the pages of a
book. Close by, the schists are strewn with crystals of felspar
PLATE 10
Folded ,| m „, completely granted,
GRAMTIZATIOX, SOUTH COAST
OF FINLAND
—
SCHIST —MIGMATITE —GRANITE

(Plate 11a), again suggesting that granite magma soaked into
the rocks. Such mixed gneisses are called migmatites (Gr. migma,
a mixture). However, the first idea that migmatites result
from the mixing of schist and granite magma cannot be sus-
tained. Both the veined and the “ eyed ” varieties of migma-
tite pass into types with larger and more abundant felspars, so
that the rock looks like a granite, except for a shadowy back-
ground representing vague remnants or “ghosts” of the original
schists (Plate 11b). Finally, even the “ghosts” vanish, and
only granite remains. The schists have somehow been trans-
formed into granite. Where has the old material gone ?
Whence came the new and what was its composition ? These
are difficult questions, and they have not yet been fully
answered. But clearly the problem of the origin of granite is
not quite so simple as Hutton thought. Hutton read part of
the story correctly, but we know to-day that it is no longer
safe to assume that all granites and granodiorites crystallized
from a completely molten state. In the alchemy of nature,
migrating fluids from the depths soak into the schists, adding
certain new ingredients and carrying away some of the old
ones. As a result of this chemical interchange, the schists
themselves are changed in composition, migmatites are formed,
and the final product is a granitic rock. Moreover, there is
growing evidence that locally the newly born granitic material
became mobile and fluid. Consequently we have here a most
—
important clue as to one and perhaps the most important
of the ways in which granitic magmas have been generated.
The Cycle of Rock Change

According to the accepted theory that the earth was
originally in a fluid state, the first rocks of the crust must in-
evitably have been igneous. However, none of these primor-

dial rocks has ever been discovered. The oldest igneous rocks
that can be found are seen to be intrusive into still older rocks
which, though now metamorphic, were originally sedimentary.
There must, of course, have been still older igneous rocks to
65
provide the materials of these sediments, and to serve as a

foundation for their deposition, but such basement rocks appear
to have vanished long ago in successive waves of fusion
which
attacked the crust from below. It is for this reason that no
evidence of the earth’s beginning has yet been detected.
Hutton remarked a hundred and fifty years ago that he could
find “ no vestige of a beginning ” and to-day, despite world-
wide search, we are still obliged to say the same. All the
Addition of Water, Oxygen, Carbon dioxide, etc.

from Atmosphere and Groundwater
Addition of
Organic Matter
Racks af all kinds

.Loose Sediments
exposed anfteSurfac
Indurateifand
Lava Flows ' Lemented. Sediments
tn
:x:
Schists S
$
%
tgneb.us <gjgneous'
^^bneis:
‘
eisses/5"
/
Intrusions^ Ir^rusions'
\
Migmatifes
GENERg'®
'0 the crus'
t
Addition of Magmas Addition of Migrating Fluids
from the Interior from the Interior
Fig. 21
The metamorphic cycle of rock change
observed igneous rocks have consolidated either from magmas

that ascended from the depths, or (at least in part) from
magmas that were generated within the crust by the fluxing
action of heat and migrating fluids on the materials already
there.
Sedimentary rocks are those produced by external pro-
cesses from pre-existing rocks of all kinds, or from organic
material derived mainly from air and water. Many of these
66
ULTRAMETAMORPHISM
have remained undenuded and, together with the accumu-
lated salts in the oceans, they represent the results of nearly
2,000 million years of surface activities. Others, however,
have been deeply buried and metamorphosed.
Metamorphic rocks are those produced by internal pro-
cesses on pre-existing rocks of all kinds, with the limitation that
the rocks concerned remained essentially solid during their
transformations. When the action of heat, stress, and migrating
fluids became sufficiently intense, the rock materials became
increasingly mobile, until they finally liquefied, wholly or in
part. Thus, beyond a certain limit of intensity, metamorphism
reaches the extreme stage —
sometimes called uitrametamor-
—
phism at which new magmas are generated. The rock
formed by the subsequent consolidation of such magma can
no longer be regarded as metamorphic. It has been re-born
as an igneous rock, and the great cycle of rock change-
igneous -> sedimentary -> metamorphic -ligneous is completed —
(Fig. 21).
In the light of the above considerations we can now
attempt a more rigorous definition of an igneous rock. An
igneous rock is one that has consolidated either (a) from
magma or (b) from hot mobile material containing sufficient
;
fluid to be capable of flow, in the course of which the pre-

existing metamorphic structures are obliterated.

G. A. J. Cole
Common Stones. Melrose, London, n.d.
Rocks and their Origins. Cambridge University Press, 1913.
C. L. Fenton and M. A. Fenton

The Rock Book. Doubleday, Doran and Co., New York, 1940.
S. J. Shand
The Study of Rocks. Allen and Unwin (Murby), London, 1931.
J. A. Howe
The Geology of Building Stones. Arnold, London, 1910.
67
MATERIALS OF THE EARTH'S CRUST : COMMON ROCKS
G. W. Tyrrell
The Principles of Petrology. Methuen, London, 1 926.
F. H. Hatch and A. K. Wells

The Petrology of the Igneous Rocks. Allen and Unwin, London, 1
937 -
F. H. Hatch and R. H. Rastall ( revised by M. Black)

The Petrology of the Sedimentary Rocks. Allen and Unwin, London,
I938.
68
Chapter VI
ARCHITECTURAL FEATURES OF THE

EARTH’S CRUST
Fracture, Flowage, and Flow
Any mass of rock in the earth’s crust is subject to gravity and

earth pressure. The resulting stresses can always be repre-
sented by three principal stresses acting mutually at right
angles Pn the least (including
:
tension) ;
P 2 intermediate and P 3
, ; ,
the greatest. One of the three is

generally vertical and the other two
horizontal (Fig. 22). If the maxi-
mum stress difference, P3 —P 1;
exceeds the strength of the rock,

then the rock is strained, and it
either breaks if it remains brittle,
or suffers a change in shape or
size if it becomes plastic. Near Resolution of the stress conditions
the surface, rocks yield mainly by acting on a mass of rock into
stresses actingalong the three prin-
fracture ; but at greater depths,
cipal axes of stress, each of which
which vary according to the kind of is perpendicular to the other two
rock, change of form takes place by

solid flowage, that is, by plastic flow, the movements being
facilitated by microscopic fractures and/or recrystallization.
Thus for each kind of rock (under the given conditions of
stress, temperature, and migrating fluids) there is a zone of
flowage below a certain depth and a zone of fracture above it.
A glacier illustrates the distinction very clearly. Here a super-
ficial zone of fracture is demonstrated by the occurrence of
crevasses in the ice. The glacier creeps down its valley, how-
ever, by flowage, and in a glacier the zone of flowage is quite
near the surface, because ice, unlike ordinary rocks, is near its
melting point, and recrystallization is easy.
When the ice of a glacier melts, the water flows away freely.
( 396 ) 69 6
ARCHITECTURAL FEATURES OF THE EARTH’S CRUST
THE CHIEF STRUCTURAL RESULTS OF FRACTURE, FLOWAGE,

AND FLOW
Sedimentary !
Metamorphic Ignfous
i
Fracture ... 1 ]
Fractures along which practically no displacement

of the rocks has occurred
i
1 1
Fractures along which the rocks on one side have

been displaced relatively to those on the other
Fracture and
Flowage . ! Bending of rocks by
(a) relative movements along cracks and bedding
and other parting planes, in the case of the stronger
and more rigid rocks and ;
(b) flowage in the case of the weaker and more

plastic rocks
'
1
1
Flowage Slaty cleavage (p. 60)
1
Foliation (p. 63)
Modes of
Occurrence
Flowage and Flow
Flow
(throughfractures, etc.) Dykes, sills,
laccoliths, etc.
(at and near the surface) Lava flows and
volcanic plugs *
Fracture
(
by explosion) Explosion pipes,
pyroclastic
accumulations
1 1
* These and other volcanic products and structures

are dealt with in
Chapter XX.
70
TILTED STRATA
Similarly, when rocks become molten, the material can flow,

though it rarely flows freely until the surface is reached. Then
extrusion takes place, either quietly (relatively speaking) as
lava flows, or explosively, with production of pyroclastic (“ fire-
broken ”) rocks. When the pressure of a magma in depth is
sufficient to overcome the strength of the rocks or their over-
head weight, the magma is impelled to move in the direction
of least resistance, either bodily or through fissures and other
parting planes, thus forming intrusions of various kinds, some
of which extend to the surface. The form and size of an
igneous mass as a whole, and its relations to the adjacent rocks
(often referred to as the “ country rocks ”) are called its mode
of occurrence.
Folds
Stratification is a primary structure of sedimentary rocks,

due to the deposition of layer after layer of differing or alter-
nating types of sediment. Most sediments were originally
To illustrate the meaning of the terms Dip and Strike
deposited on flat, very slightly inclined surfaces. But in many

regions we find that great thicknesses of strata have been tilted,
so that they now lie in inclined positions, sometimes for many
miles. The beds generally represent one side of a very
tilted
broad fold. The attitude, or position in the ground of any
inclined bed, is accurately described by what is called its dip
(Fig. 23). The dip is both the direction of the maximum slope
7i
,
[H.M. Geol. Survey

Fig. 24
Sea cliff of Torridon sandstone, showing dip and strike, Cailleach Head,
Ross-shire
down a bedding plane and the angle between the maximum

slope and the horizontal. The direction is measured by its
true bearing, as so many

degrees east or west of north, the
compass reading being suitably corrected for magnetic varia-
tion. The angle is measured with a clinometer. The strike
of an inclined bed is the direction of any horizontal line along
a bedding plane the direction, for example, of the inter-
;
section of the bed with still water or level ground. It is at right

angles to the dip direction (Fig. 24).
When beds are upfolded into an arch-like form (with the
lower beds within the upper) the structure is called an anti-
because the beds then “ incline away ” from the crest on
cline ,
either side (Plate 12a). When the beds are down-folded into
a trough-like form (with the lower beds outside the upper)
the structure is called a syncline because in this case the beds
,
on either side “ incline together ” towards the keel (Plate 12b).

The two sides of a fold are described as its limbs the limb which
72
FOLDED STRATA
isshared between an anticline and its companion syncline being

called the middle limb. The plane which bisects the angle
between the limbs is the axial plane, and the axis of the fold
along any particular bed (about which the bed is folded) is the
line of intersection of the axial plane with the surface of that
bed. If the axial plane is vertical, the fold is upright and
symmetrical, and the crest coincides with the axis. If the axial
plane is inclined, the fold is also inclined ;
relative to the
vertical of course, unsymmetrical (Fig. 25).
it is,
As they are traced along their axes, folds sooner or later

die out ; and here, or wherever the axes are not horizontal,
they are said to pitch, pitch being simply the dip of the beds
along the axis of a pitching fold. Anticlines and synclines
are thus like more or less elongated canoes, upside down or
Fig. 25
Inclined folds, with the upper strata

removed to show the surface of one
particular bed. The trace of an axial
plane is marked by the dotted line
right way up. Domes and basins represent the limiting cases
in which the beds dip in all directions, outwards from, or
inwards towards, the centre of the structure. Between these

and folds that extend for many miles there is every gradation.
Such structures must not, of course, be confused with hills and
valleys, for they refer solely to the attitudes of the bedrocks
below the surface, and not to the relief of the surface itself.
Folds range in intensity from broad and gentle undulations
to steep-sided, highly compressed folds in which the beds may
be turned on end, and the limbs of anticlines (or, in other
words, the upper limbs of synclines) turned over even beyond
the vertical (Plates 13 and 81). In the Southern Uplands of
Scotland the rocks have been so intensely compressed that the
limbs of the closely packed folds have become practically
parallel. Folding of this “ concertina ” type is described as
isoclinal (Plate 69a). In this region, and also in the Lake
73
District, thepuckers and smaller folds are superimposed on

broad anticlinal and synclinal folds of a much larger order.
An anticlinal complex of folds of different orders is called an
Fig. 26
Schematic section across an anticlinorium and a synclinorium
anticlinorium, and the complementary synclinal complex a

synclinorium (Fig. 26 ).
When a pack of cards is bent into a fold, slipping between
the adjacent cards invariably occurs. Similarly, during the
growth of anticlines, the outer beds shear or creep along the
inner ones towards the axial plane, while in synclines the lower
beds move in the same way under the higher ones. Plastic
Recumbent Fold Recumbent Fold

Fig. 27
Section across the Carrick Castle recumbent anticline (Argyllshire). The strata
of the lower limb are upside down (1 to 9 =oldest to youngest), and have
themselves been folded into an anticline (the Cowal anticline)
or incompetent ” beds, like shale, tend to creep by flowage

more readily than stronger or “ competent ” beds, and for this
reason they become squeezed out and thinned in the limbs and
74
folds and joints
thickened towards the axes. It is during this flowage that

shales tend to become slaty, the resulting cleavage planes being
parallel to the axial planes. Thus, in an anticlinorium the
cleavages are disposed across the whole structure like the ribs
of a fan.
An overturned fold, or overfold, has one of its limbs inverted,
and if the latter approaches a horizontal attitude the overfold
is described as recumbent the beds of the middle limb (the
;
lower limb of the anticline and the upper limb of the under-
lying syncline) are then upside down. Parts of the Grampians
are carved out of a gigantic recumbent fold (Fig. 27). In
Diagrammatic section showing various types of folding and thrusting
structures of this kind part of the middle limb may be sheared

or squeezed out altogether. Further development of the
structure then results in the rocks of the upper limb being
pushed bodily forward along the plane of shearing. The latter
has become a thrust plane and the structure an overthrust fold
(Fig. 28).
Joints
Thestronger beds in a folded series of strata become

ruptured when they are bent, the cracks being known as
joints. Unfolded rocks are also commonly divided into blocks
by jointing. If we examine a quarry or cliff section of sand-
stone or limestone, we find that in addition to the bedding
planes the rock is traversed by fractures that are generally
approximately at right angles to the stratification, and there-
fore nearly vertical when the beds are flat (Plates 8a and 14a).
Joints frequently occur in sets consisting of two series of parallel
joints, one series, if the beds are inclined, approaching in
trend to the dip direction and the other to the strike. Such
75
joints are of great assistance to the quarryman in his task of

extracting roughly rectangular blocks of stone, especially the
“ master joints,” which are often remarkably persistent and
strongly developed. In most rocks, however, there are sub-
ordinate joints which cut across the main sets, thereby dividing
the rocks into irregular angular blocks.
Joints may be due either to shearing under compression
or to tearing apart under tension. Shear joints tend to be
clean cut and tightly closed (in unweathered rocks), whereas
tension joints are more irregular, rough, and open. At the
surface, joints of all kinds are very susceptible to the attack of
weathering agents ; they are readily opened up by the work
of rain, frost, wind, and plant roots. It is due to this opening
of joints and bedding planes that cliffs and mountain scarps
often resemble roughly hewn masonry (Plates 15a and 32a).
Along the shore waves attack the rocks selectively along joints
and the influence of the joint pattern is often clearly shown in
the outlines of inlets, caves, and skerries (Plate 67a). The joint
pattern may also control the course of rivers, the joint planes
themselves commonly forming the walls of steep-sided gorges
and canyons (Plate 1).
In igneous rocks tensile stresses are set up by contraction
during cooling. In granite masses three series of joints are
commonly developed, two being nearly vertical and the third
approximately parallel to the surface. The latter produces a
sheet structure, which becomes conspicuous when the rock is
exposed to weathering. The tors on granite moorlands are
divided up into gigantic blocks with a bold architecture like
that of a cyclopean fortress (Fig. 62).
A quite different architectural effect is produced by the
columnar jointing of certain basalts and other fine-grained
and compact igneous rocks (Plate 14b). Somewhat similar
sets of polygonal cracks can be seen in the dried-up
mud of a
marsh or river flat that has been exposed to the sun (cf. Plate
19). Mud cracks result from shrinkage due to loss of water
from the surface layer. The polygonal cracks of basaltic
sheets are due to contraction during cooling. When a hot
homogeneous rock cools uniformly across a plane surface, the
.
COLUMNAR JOINTING
contraction equally developed in all directions throughout
is
the surface. This condition is mechanically the same as if the

contraction acted towards each of a series of equally spaced
centres. Such centres ( e.g C, 1, 2, 3, etc. in Fig. 29a) form the
corners of equilateral triangles, and theoretically this is the only
possible arrangement. At the moment of rupture the distance
between any given centre C and those nearest to it (e.g. 1-6)
is such that the contraction along lines such as C 1 is just—
sufficient to overcome the tensile strength of the rock. A
tension crack then forms half-way between C and 1 and at
right angles to the line C — 1. As each centre is surrounded by
Fig. 29
To formation of an ideal hexagonal pattern of joints
illustrate the
by uniform contraction in a plane towards evenly spaced centres
six others (1 to 6 in Fig. 29a), the resultant system of cracks is

hexagonal. Once
a crack occurs somewhere in the cooling
layer the centres are definitely localized, and a repeated pattern
of hexagonal cracks spreads almost simultaneously throughout
the layer (Fig. 29b). As cooling proceeds into the sheet of
rock the cracks grow inwards at right angles to the cooling
surface, and so divide the sheet into a system of hexagonal
columns.
Neither the physical conditions nor the rocks are usually
sufficiently uniform to ensure perfect symmetry, and the actual
result is a set of columns with from three to eight sides, six,
however, being by far the commonest number. Vertical
contraction is relieved by cross joints, which are generally
77
eitherconcave or convex, and the columns are thus divided
into short lengths.The resulting appearance of well-trimmed
masonry is often remarkably impressive, as in the tessellated
pavement of the Giant’s Causeway, and the amazing archi-
tecture of Fingal’s Cave in Staffa (Plates 3 and 14b). Com-
parable, but usually cruder and less regular columnar jointing
develops during the cooling of sills and dykes. In a dyke the
cooling surfaces are the vertical walls, and the resulting columns
are therefore horizontal.
Faults
A fault is a fracture surface against which the rocks have

been relatively displaced. Vertical displacements up to
thousands of feet and horizontal movements up to many
miles are well known, but in no case is there any reason to
suppose that the total displacement occurred in a single
catastrophic operation. Earthquakes commonly result from
sudden movements along faults that are still active, but the
fault movements are rarely more than a few feet at a time.
The various types of faults depend on the relationships between
the three principal stresses referred to on page 69, assuming,
of course, that in the circumstances of each case the stress
difference is sufficient to bring about fracture and movement.
The vertical stress may be due to gravity or magmatic pressure,
and the horizontal stresses to lateral earth compression or
tension. Three types of faults occur most commonly, but the
details given below by no means exhaust the possibilities,
which may be tantalisingly complex.
In connection with the inclination of fault planes, it should
be kept in mind that shear planes curve towards the locus of
least resistance. This is well shown by landslides (Fig. 63).
When the Panama Canal was dug,
the materials supporting
the hillsides were locally removed, and there was much
downward slipping along concave surfaces that began steeply
on the heights and curved round to the horizontal at the level
of the canal floor. On the other hand, the direction of least
resistance may be far below the surface, just where it might
78
NORMAL FAULTS
be thought that the obstruction would be irresistible. This
when there are down-dragging movements
possibility arises
in progress in the depths.
The sides of may be polished or
a clean-cut fault plane
striated by between the moving blocks ; such surfaces
friction
are known as slickensides. Sometimes, instead of a single
fracture, there are two or more, forming a strip consisting of a
sheet of crushed rock of variable thickness. This is dis-
tinguished as a fault zone, and the shattered material within is
called a fault breccia.
(a) Normal Faults (Figs. 30 and 31 and Plate 15b). —The

vertical stress is the greatest of the three. The resulting fracture
is generally inclined at an angle between 45° and the vertical.
Fig. 30 Fig. 31
The relative movement involved Oblique -slip normal fault,
in a normal fault. Part of the illustrating the meaning of
fault plane is left unshaded the term slip
The beds abutting against the fault on its upper face or

“ hanging wall ” are displaced downwards relative to those
against the lower face or “ footwall.” The terms “downthrow”
and “ upthrow ” for the two sides are, of course, purely relative.
In faults recently active the footwall is exposed at the surface
as a fault scarp, and survey measurements sometimes show that
both sides were uplifted ; one, however, being heaved up more
than the other. More usually, the two sides have moved in
opposite directions. Normal faults involve an extension of
the faulted beds. A boring through the fault would fail to
penetrate some particular bed altogether, passing between
its ruptured ends. Such extension indicates that freedom of
79
;
movement was possible towards the direction of least resistance

the minimum stress in many cases appears to have been a
tension, involving stretching instead of compression.
(b) Reverse or Thrust Faults (Fig. 32). —The vertical stress

is the least of the three. When the resulting fracture is inclined
at an angle between 45° and
the horizontal, as it often is,
the corresponding fault is de-

scribed as an overthrust. High-
angle thrusts or reverse faults
are, however, far from rare.
But whatever the angle, the
beds on the upper side are
thrust up the fault plane rela-
Fig. 32
tive to those below. Shorten-
Reverse or thrust fault
ing of the faulted area is thus
involved, and the operation of powerful tangential compres-
sion is obviously indicated. In overthrusts the upper block
may be driven forward for many miles (Fig. 33). As already
pointed out (page 75) overthrusts commonly develop along
the middle limbs of recumbent folds. In the Alps and
Pre -Cambrian
lewisian Gneiss Maine Schisfs
Clencoi/f 1 Imbricate
Thrust Plane structure between
j
Fig. 33
Section across the North-West Highlands of Scotland to illustrate
overthrusting
and imbricate structure. For Moine Thrust see Fig. 180
other great mountain ranges enormous overthrust folds and

overthrust blocks have been propelled far from their original
“roots.” Such far-travelled rock-sheets are called nappes. Along
the thrust plane (Plate 16) of a great overthrust or nappe
the
crushing is generally very severe, and a hard, streaky, or
banded
rock, to which the name mylonite is given,
may ultimately be
produced from the pulverised and rolled-out materials (Fig. 34).
PLATE If!
K. U'i'h'k
(A) Dolerite dske cutting Chalk and Tertiary basalt la\as. Clave Hill, Belfast
TEAR FAULTS
Fio. 34
Photomicrograph of mylonite formed from Lewisian gneiss
along a thrust plane near Laxford, North-West Highlands
(c) Tear Faults (Fig. 35). —The vertical stress is inter-

mediate between the maximum and minimum horizontal
stresses.The resulting movement is predominantly horizontal,
the fracture being vertical or nearly so. Tear faults are
commonly developed in nappes, where they naturally arise if
one part of a nappe has been driven forward further than the
Fig. 35
Tear fault ; also known by the
terms transcurrent or strike-slip
fault
adjoining parts. Other tear faults appear to be due to rota-

tional movements (about a vertical axis) of certain crustal
blocks relative to their surroundings. Each individual move-
ment is a small jerk, like that of the hands of a clock placed
horizontally. A sudden jerk of this kind along a tear fault
8i
hundreds of miles long caused the earthquake that wrecked

San Francisco in 1906. Roads, fences, and water mains that
crossed the fault were cut through and the ends displaced by
several feet. During its history a cumulative movement of
many miles has taken place along this fault. On the Pacific
side the apparent movement is towards the north, as if the
Pacific floor were moving in an anticlockwise direction against

the North American continental block {cf. Fig. 184).
Modes of Occurrence of Igneous Intrusions

The forms and attitudes of igneous rocks which have been
injected into crustal rocks largely depend on their relation to
the parting planes of the invaded formations. This is seen
Fig. 36
Diagrams to illustrate dykes and their surface features :
(a) Dyke more resistant to erosion than wall rocks

(
b) Dyke and wall rocks equally resistant (cf. Plate 17a)
(c) Dyke less resistant than wall rocks
most clearly where the strata have remained horizontal or have

been only gently tilted or folded. One of the commonest signs
of former igneous activity is provided by the wall-like intrusions
called dykes(Fig. 36 and Plate 17). Here the magma has
ascended through approximately vertical fissures, forcing the
walls apart asit rose, and so, on cooling, becoming a vertical
sheet of rock with roughly parallel sides cutting across the

bedding planes. Such intrusions are said to be transgressive
or discordant. In certain circumstances the magma may open
82
SILLS AND DYKES
up a passageway along a bedding plane, making room for
itself by uplifting the overlying rocks. The resulting tabular
sheet of rock (Fig. 37) is called a sill (Anglo-Saxon, syl, a ledge).
Intrusions that are parallel to the adjacent stratification are
said to be concordant.
Ideal section through a sill (black) in inclined strata
Dykes vary greatly in thickness, from a few inches to

hundreds of feet, but widths of five to twenty feet are most
common. There is also great variation in length, as seen at
the surface, from a few yards to many miles. Dykes are very
numerous in some regions of igneous activity. Along a fifteen-
mile stretch of the coast of Arran, for example, a swarm of
525 dykes can be seen, the total thickness of the dykes being
5,410 feet. Here the local extension of the crust has been more
than one mile in fifteen. Farther north, focused on Mull,
there is another great swarm of dykes, some of which can be
traced at intervals across southern Scotland into the north
of England (Fig. 38). Most of these dykes are dolerites of
various kinds, though along the margins the rock is generally
of finer grain, owing to chilling by the walls, and may be basalt
or even tachylyte. The wall rocks themselves show the effects
of thermal metamorphism. Coal seams in contact with dykes
are reduced to a hard mass of natural coke. When a dyke is
more resistant to weathering and erosion than its walls, it
projects as a prominent ridge, sometimes running across the
country like a wall. On
the other hand, if the walls are more
resistant, the dyke is worn away
(e.g. by sea waves) into a long
narrow trench or cleft. In ordinary usage, the word “ dyke ”

its
may refer to either a ditch or a wall (cf. Fig. 36).

83
architectural features of the earth’s crust
Fig. 38
Map showing the Tertiary dyke swarms in the British Isles,
and their relation to For a
the Tertiary plutonic centres.
view of the major intrusions of Skye see Plate 18b
The classic example of a sill is the Great Whin Sill of the

north of England (Figs. 39 and 40). Whin or whinstone is a
quarryman’s term for any dark-coloured rock, such as basalt
or dolerite, which can be used as road-metal. Beginning in
Fig. 39
Section across Northern Pennines to show the Great Whin Sill and its
the
thinning and upstepping towards the west. Length of section equals 17 miles.
T = Triassic. G — Carboniferous. S = Silurian and Ordovician
the north of Northumberland, the surface outcrop of the edge

of the Whin Sill swings round to the castled crags of Bamburgh
and seawards to the Fame Islands. Appearing again on the
coast below the ruins of Dunstanburgh Castle, it can be fol-
lowed across Northumberland towards the River North Tyne.
84
[H.M. (jcol. burvey
(A) Mudcracks, lorside Reservoir, Longdendale, Cheshire
rn j * 1 ourvey
1 1,lcd - m m ^dcracks m
'
•
, , . L
mudstone of Old Red Sandstone age, Clairdon
shore, east of Thurso, Caithness
GREAT WHIN SILL AND DYKES
Fro. 40
Map of the Great Whin Sill and related dykes
t396) 85
Then for miles its and weather-worn scarp boldly faces
tilted
the north (Plate 18a), and here, with an eye for the best
defensive line, the Romans carried their famous Wall along
its crest. To the south-west it outcrops in the valleys that
notch the western margin of the Pennines. Inland from the
exposed edge, the Upper Tees cuts through it in the waterfalls
of Cauldron Snout and High Force. Near Durham it is en-
countered in depth, over a thousand feet below the surface.
The thickness of the sill varies from a few feet, through a
general average of nearly 100 feet, to more than 200 feet. In
places it divides into two or more sills and
at different levels,
locally it betrays its intrusive character by breaking obliquely
across the strata from one set of beds to another. This observa-
Fig. 41
Diagram to illustrate the form of an ideal laccolith
tion proves that the Whin Sill is not a lava flow, and further
evidence of this is provided by the fact that the rocks above it
are metamorphosed, as well as those below (Figs. 18 and 39).
Instead of spreading widely as a relatively thin sheet, an
injected magma, it is very viscous, may find it
especially if
easier to arch up
the overlying strata into a dome-like shape
(Fig. 41). Such intrusive forms are typically developed to the
east of the Rockies,
where they were first described by Gilbert,
who them laccoliihs (Gr. laccos, a cistern, lithos, a rock).
called
There are few good examples of laccoliths in Britain, though
many stocks have been wrongly called laccoliths. Stocks are
discordant intrusions, whereas laccoliths, like sills, are con-
cordant.
Intrusions, which on the whole are
concordant and have
a saucer-like form, are distinguished as lopohths
(Gr. lopas, a
86
CONE-SHEETS AND RING-DYKES
shallow basin) The best-known examples are of extraordinary

.
dimensions, the largest, that of the Bushveld in South Africa

(Fig. 42), being nearly as extensive in area as Scotland. The
sagging of these great sheets would seem to be an inevitable
consequence of the transfer of enormous masses of magma from
the depths to the upper levels of the crust.
When magmatic pressure is exerted upwards from a more
or less circular area in depth (like the top of a dome), the
resulting cracks in the overlying rocks may be either radial
or concentric, according to circumstances. Dykes arranged
radially occur around certain volcanic centres from which the
outer cover of lavas and pyroclasts has been stripped by
denudation. Ring-shaped intrusions are less common, but
Fig. 42
Diagrammatic section across the lopolith of the Bushveld, Transvaal.
Length of section about 315 miles
1. Crystalline rocks of the Basement Complex 2. Transvaal system
3. Rooiberg series 4. Sills of dolerite 5. Norite (a variety of gabbro)

6. Granite 7. Pilandsberg volcanic centre
are characteristically developed around the intrusive centres

of the west of Scotland and the north of Ireland (shown
in Fig. In the type thought to be due to magmatic
38).
pressure the fractures have the form of an inverted cone.
Uplift of an inner cone relative to an outer one involves
opening of the crack between them into a fissure. Injections
of magma then give a series of concentric shells, dipping in-
wards, as illustrated in Fig. 43. Cone-sheets, as these intrusions
are called, are remarkably well displayed in Mull, Ardna-
murchan, and Skye.
If, on the other hand, the upper rocks are left relatively
unsupported, they tend to break along concentric fractures like

those shown in Fig. 44. Subsidence then opens the way for
injection of magma. The resulting intrusions are called ring-
dykes ; they differ from cone-sheets in being nearly vertical,
87
ARCHITECTURAL FEATURES OF THE EARTH'S CRUST
Fig. 43
Block diagram to illustrate the form of an ideal series of cone-sheets,
and their probable relationship to an underlying magma basin
dipping outwards at steep angles, and also in being much

thicker. Now these two characteristics are obviously mutually
inconsistent, since sinking of a cylinder or steep-sided cone
could not possibly make room for the very thick ring-dykes
Fig. 44
Block diagram to illustrate the form of an ideal series of ring-dykes
and their supposed relationship to an underlying magma basin
88
BATHOLITHS
that are actually known. The ring-dykes must, therefore, have
been emplaced by either displacing or replacing the wall-rocks
of a ring-fracture, or the rocks lying between two or more
concentric ring-fractures. How this difficulty can be met we
shall consider in connection with batholiths, where the “ space
problem ” has to be squarely faced.
Batholiths (Gr. bathos, depth) are gigantic masses of essen-
tially igneous rocks, generally composed of granite or grano-
diorite, with highly irregular dome-like roofs, and walls that
plunge downwards so that the intrusions enlarge in depth and
Fig. 45
Block diagram to illustrate the characteristic features of a batholith
appear to be without visible foundations (Fig. 45). They

occur in the hearts of mountain systems of all geological ages
and are seen wherever denudation has cut down sufficiently
deeply. Although in detail their margins are markedly dis-
cordant towards the surrounding rocks, they are generally
elongated parallel to the general trend of the mountain
systems in which they are found. Some of the batholiths of
western America have been exposed over lengths of many
hundreds of miles, the width being usually about a tenth of
the length or less.
Smaller intrusions of similar type, but less elongated and

with areal dimensions of only a few square miles or less, are
called stocks. Many of these are probably offshoots from under-
89
lying batholiths, of which only the highest parts have been

exposed to view by denudation. Mining operations have often
shown this to be the case. When a stock has a roughly circular
outline, like the Shap granite of Westmorland, it is sometimes
referred to as a boss.
Some examples of stocks and bosses may have come into
position by a process called cauldron subsidence.Part of the
cylindrical block of country rock enclosed within a vertical
ring-dyke may founder into the underlying magma reservoir,
while volcanoes are active at the surface. The space between
the lava roof and the sunken block fills with magma which
consolidates as, say, granite. On denudation the boss of granite
is revealed by the removal of the volcanic rocks. In Glencoe
there is visible evidence on the hillsides that such subsidences
have actually occurred. In recent times a similar process,
involving the collapse of the roof itself (i.e. of part of a volcanic
cone), has been responsible for the formation of giant volcanic
calderas [cj. Fig. 240).
The Emplacement of Batholiths

Cauldron-subsidence cannot be the mechanism by which
batholiths have been emplaced, because many of them still
retain patches of their original roof of country rocks. More-
over, the mechanism 'implies the presence in depth of a magma
reservoir of batholithic dimensions which itself calls for ex-
planation. As in the case of thick ring-dykes, the question has
to be faced What has happened to the crustal rocks that
:
formerly occupied the space now taken up by the batholith ?

If the batholith was entirely formed by the consolidation of
magma that ascended bodilyfrom below, then the pre-existing
rocks must have been displaced upwards, sideways, or down-
wards. But only a part of the space required can be made
available in ways like these. Observations of roof and wall
show that outward displacements of the country rocks do occur,
but only on a relatively small scale, such as would be inevitable
around any large expanding mass.
Downward displacement seems at first sight to be more
• 9°
EMPLACEMENT OF BATHOLITHS
promising. This involves shattering of the roof rocks by
thermal expansion, dislodgment of the fragments by the
penetration of gases and tongues of magma into cracks, and
finally the engulfment and sinking of the blocks. This process
is called magmatic stoping. Inclusions of country rocks of all
sizes,more or less intensely metamorphosed, are often pre-
served in the upper levels of stocks and batholiths. They are
known as xenoliths (Gr. xenos, foreign). In depth, however,
they become gradually smaller and less numerous, and more
and more like the granitic rocks enclosing them, until finally
they disappear altogether. Evidently they have been in-
corporated into the granite and the space problem remains.
Moreover, long “ islands ” of country rocks like those of the
metamorphic aureole can sometimes be recognized in the
granite, their structures remaining in continuity with those of
the bordering country rocks, though they may often be no
more than vague shadowy outlines, seen through a veil of
granitization (Plate 11b). Black metamorphosed dolerite
dykes, older than the batholith, can sometimes be traced into
it as partly granitized “ ghosts.” Since the heavy dyke rocks
did not sink, there is no reason to suppose that the associated
metamorphosed sediments could have done so.
The
stoping hypothesis has to face another serious difficulty.
If granite magma ascends in great bulk towards the surface
in thisway, it should often have broken clean through the
crust toform gigantic volcanoes erupting rhyolite and obsidian
and the corresponding pyroclasts. But this has very rarely
happened, as the retention of the original roof clearly proves.
Moreover, if granite magma rose from the depths in quantities
corresponding to the enormous volumes of batholiths, it
would be by far the most abundant of all magmas. In this
case rhyolite should be the most abundant of all volcanic
rocks. But it is not. Although granite and granodiorite are
easily the commonest of plutonic rocks, it is basalt that is the
commonest of the volcanic rocks. This striking fact strongly
suggests that the volume of actual magma concerned in the
development of batholiths was relatively small in comparison
with their enormous bulk.
91
Since we cannot problem by sinking the
solve the space
pre-existing rocks out of the way or by pushing them upwards
or sideways, it follows that a high proportion of the material
of the original rocks must still be there, though now trans-
formed into granite and associated types of igneous rocks.
Every positive line of evidence supports this view, including
that already summarized on p. 65 . Apparently the original
rocks were not invaded by granite magmaas such, but by
hot granitizing fluids rich in gases. These soaked through the
rocks, changing their composition, metamorphosing them with
an ultrametamorphism, thereby leading
intensity that reached
to the generation of granitic magma in situ, and rendering the
whole mass mobile, so that it became igneous.

J. Geikie ( revised by R. Campbell and R. M. Craig)
Structural and Field Geology. Oliver and Boyd, Edinburgh, 1940.
C. B. Brown and F. Debenham
Structure and Surface. Arnold, London, 1 929.
B. Willis and R. Willis
Geologic Structures. McGraw-Hill, New York, 1934.
E. S. Hills
Outlines of Structural Geology. Methuen, London, 1 940.
C. M. Nevin
Principles of Structural Geology. Wiley and Sons, New York, 1942.
F. H. Lahee
Field Geology. McGraw-Hill, New York, 1941.
R. A. Daly
Igneous Rocks and the Depths of the Earth. McGraw-Hill, New York,
1933 -
Architecture of the Earth. Appleton-Century, New York, 1938.

E. M. Anderson
The Dynamics of Faulting and Dyke Formation with Applications to Britain.
Oliver and Boyd, Edinburgh, 1942.
92
Chapter VII
ROCKS AS THE PAGES OF EARTH HISTORY

The Key to the Past
So far we have dealt with rocks as materials which have been
formed from pre-existing materials by the action of geological
processes. Rocks are also the pages of the book of earth
history, and the chief object of historical geology is to learn
to decipher these pages, and to place them in their proper
historical order. The fundamental principle involved in
reading their meanings was first enunciated by Hutton in
1785, when he declared that “ the present is the key to the
past.” Rocks and characteristic associations of rocks, with
easily recognizable peculiarities of composition and structure,
are observed to result from processes acting at the present day
in particular kinds of geographical and climatic environments.
If similar rocks belonging to a former geological age are found
to have the same peculiarities and associations, it is inferred
that they were then formed by the operation of similar pro-
cesses in similar environments.
We have already, as a matter of ordinary common sense,
had occasion to apply this principle. The presence in a lime-
stone of fossil corals or ammonites, or of the shells of other
marine organisms, indicates that the limestone was deposited
on the sea floor, and that what now is land once lay beneath
the waves. The limestone may pass downwards or laterally into
shale, sandstone, and conglomerate. The last of these represents
an old storm beach, and it indicates the shore line where land
and sea came Elsewhere, old lava flows represent
together.
the eruptions of ancient volcanoes beds of rock salt point to
;
the former existence of inland seas that evaporated in the

sunshine ; seams of coal, which are the compressed remains
of accumulations of peat, suggest widespread swamps and
luxuriant vegetation and smoothed and striated rock surfaces
;
associated with beds of boulder clay prove the former ex-

93
tension of glacial conditions. In every case the characters of

older formations are matched with those of rocks now in the
making.
Even the weather may be recorded in the structures of the
rocks. A brief rain shower falling on a smooth surface of fine-
grained sediment leaves its marks as little crater-like pittings
known as rain prints. Sun cracks develop in the mud flats of
tidal reaches or flood plains when the mud dries up and
shrinks (Plate 19a). Occasionally it happens that the poly-
gonal cracks become filled with wind-blown sand before the
next tide or flood sweeps over the area. Then, instead of being
obliterated, they have a chance to become permanent. Tims
it comes about that similar structures are preserved in older
beds of corresponding origin (Plate 19b). As far back as

geological methods can be applied to the earth’s history, such
relics of “ fossil weather ” prove that wind, rain, and sunshine
have always been much the same as they are to-day. Never-
theless, the distribution of climates over the earth’s surface
has varied in a most astonishing way.
In our own country the work of former ice sheets and
glaciers is still written conspicuously in the landscapes and
superficial deposits left behind when the ice began to retreat
about 20,000 years ago. In striking contrast, the very much
older clay through which the London tubes are bored contains
remains of vegetation and shells and reptiles like those of the
modem tropics. In still earlier periods there is evidence of
desert conditions. Elsewhere the vicissitudes of climate are
equally startling. In India and in central and southern
Africa there is clear proof that while Britain was a land of
swampy, tropical jungles (the time of coal formation) these
lands were buried under great ice sheets like those of Greenland
and Antarctica at the present day. In Greenland, however,
the older rocks contain remains of vegetation that could have
grown only in a warm climate, while near the South Pole
Captain Scott found beds of coal, pointing to conditions very
different from those of to-day.
We may next turn to certain examples of Hutton’s principle
which can be applied to problems connected with earth move-
94
CURRENT BEDDING
ments. When sand is deposited from currents of shallow water,
shoalsand sandbanks are often built up. The bedding of a
growing sandbank follows the gently curved slopes on which
the sand is dropped, giving a pattern in cross section which,
under resembles that shown in Fig. 46a.
ideal conditions,
With changing conditions, possibly during a storm, the upper
part of the sandbank is swept away, and the bedding planes
are sharply truncated by an erosion surface such as AB.
Later on, another group of sandbanks is likely to be de-
posited on the flat surface thus provided. So, in a quarry or
Fig. 46
Sections to illustrate current bedding and its value in determining whether a

bed is right way up or upside down
(a) The current-bedding structure is complete
(b) The upper part has been eroded off down to AB. The
truncated current-bedding pattern is the right way up
(c) The truncated current-bedding pattern is upside down
cliff exposure of an old sandstone, we may find that within
certain bands the bedding is oblique and variously inclined
to the general “ lie ” of the formation as a whole (Plate 20a).
This structure, which is original, and not due to tilting or
folding, is called cross bedding or current bedding. Sand dunes,
accumulating from wind-blown sands, also exhibit cross
bedding, the pattern of which reproduces, wholly or in part,
the characteristic outlines of dunes (Fig. 137).
Now we may take a further step. On page 75 it was stated
that some of the rocks of Scotland are upside down for many
miles, and the reader may well have wondered how such a
95
thing can be known. Hutton’s principle cannot be applied

directly,because no-one has ever witnessed rocks being turned
upside down by overfolding. Here, however, cross bedding
comes to our aid. Fig. 46b shows that when the upper part
of a sandbank is removed by erosion the cross bedding that
remains is abruptly truncated at the top, whereas at the bottom
it curves gently into the main stratification. The truncated
top and the original floor of the band of sandstone are clearly
distinguishable. At Kinlochleven and other places in Argyll-
shire observations of cross bedding show that the floor is now
above the truncated top (as in Fig. 46c). The beds in which
this inverted structure appears are therefore known to be upside
down.
Ripple marks, like those seen on a beach after the tide has
gone down, are often preserved in ancient sandstones. Desert
sands are often beautifully rippled by the wind (Plate 5a), but
from the nature of the case wind ripples are very rarely pre-
served. Ripple marks formed by the to-and-fro movements of
water have sharp crests and rounded troughs, and conse-
quently the top and bottom of any bed of sandstone in which
they occur can easily be recognized. Where such ripples
remain in formations that have been disturbed by severe fold-
ing, they, too, can be used to determine whether a particular
bed is upside down or not (Plate 20b).
There are, of course, certain processes and rocks to which
Hutton’s principle cannot be applied. It is impossible to
observe granite in the making, nor has it been found possible
to make granite experimentally. Precisely for these reasons
the problems associated with the origin of granite and other
plutonic rocks have long remained matters of speculation, and
therefore of controversy. In the circumstances of the case the
present provides no key to the past, and all we can do is to
suggest processes of diffusion, magma formation and crystal-
lization, which are consistent with the structures and relation-
ships of the rocks in question. Since the available evidence is
often difficult to interpret, there is often a choice between rival
possibilities. Moreover, the earth’s internal behaviour is not

limited by our present knowledge of physics and chemistry.
96
SUCCESSION OF STRATA
We have certainly no right to assume that all the possibilities

have yet been discovered. It is therefore not surprising that
attempts to reverse Hutton’s principle and use the past as
the key to the present have so far been only partially successful.
The Succession of Strata
To place all the scattered pages of earth history in their

proper chronological order is by no means an easy task. The
stratified rocks have accumulated layer upon layer, and where
N.W. Ingleborough S.E

Millstone Grit _
Yoreda V '’rfV*
Carboniferous
Limestone
Limestone, '..<
I L
T"*
...
»
*'£-')
»•
v
^
LZ 'Z!*
;
Lower Palaeozoic (.
- ’
Slates and Grits •
Fig. 47
Section Ingleborough and its foundations, showing the unconformity
across
between Carboniferous beds above and the intensely folded Lower
the
Palaeozoic strata below. Length of section equals 4 miles. (After D. A. H’rqy)
a continuous succession of flat-lying beds can be seen, as on

the slopes of Ingleborough (Fig. 47), where there has been no
inversion of beds by overfolding or repetition of beds by over-
thrusting, it is obvious that the lowest beds are the oldest, and
those at the top of the series the youngest. Where a series of
Mat tern Hills Cotswid Hills ChJtern Hi’s
SevernValfey
Section from the Malvern Hills to the Chiltern Hills

1. Pre-Triassic Pre-Cambrian and Cambrian) -. Triassic 3. Lias
i
4. Lower 5. Oxford Clay

Oolites (>. Corallian 7. Kimmeridge Clay
8. Portland Beds 9. Gault and Upper Greensand 10. Chalk
beds has been tilted, as between Gloucester and London, the

worn-down edges of lay'er after layer come in turn to the surface
and it becomes possible to place a long succession of beds in
their proper sequence (Fig. 48).
97
Around Ingleborough the great limestone platform of the
Pennines can be traced over a wide stretch of country, but
where the streams have cut through to its base, the limestone
is found to lie on the upturned edges of strongly folded and
Deposition of
Sediments
Gentle Folding
More intense Folding
Uplift and
Sculpturing of
Surface by Denudation
Reduction of Surface to
a Plain by Denudation
Subsidence and Deposition

of a new series of Sediments
—UNCONFORMITY
Fig. 49
Diagrams to illustrate successive stages in the development of an unconformity
crudely cleaved beds, as shown in Fig. 47 and Plate 21a. Here

there is evidently a sudden break in the continuity of the
record. Such a break, which may represent a very long
interval of geological time, is called an unconformity. The beds
above the break in the sequence are said to be conformable,
98
SIGNIFICANCE OF FOSSILS
and to restunconformably on the rocks below. The latter

consist of gritty sandstones and slaty shales. After their
deposition on the sea floor as newly formed sediments, they
were laterally compressed into folds in the heart of an ancient
—
mountain range -a range which extended throughout Scandi-
navia and much of the British Isles, and is known to geologists
as the Caledonian range because much of Scotland is carved
out of its contorted rocks. By denudation the folded grits and
slates were gradually uncovered and ultimately reduced to
an undulating lowland. Then the worn-down surface was
submerged beneath the sea, to become the floor on which the
horizontal sheets of the Pennine limestones were deposited.
Successive stages of the events which occurred during the time-
gap represented by the unconformity are shown by the dia-
grams of Fig. 49.
In general terms, every unconformity is an erosion surface
of one kind or another, representing a lapse of time during
which denudation (including erosion by the sea) exceeded
deposition at that place. If sediments were deposited ihere
during the interval, they must subsequently have been removed.
The time-gap is likely to be represented by strata somewhere
else (see Plate 21b), and our next problem is how to recognize
such strata if we find them.
The Significance of Fossils
The solution of this problem was the great achievement of

William Smith, a land surveyor who was born near Oxford in
1769. As a boy he collected fossils from the richly fossiliferous
beds near his home, and in later years he carefully collected
suites of fossils from each of the sedimentary formations repre-
sented in Fig. 48. He found that while some of the fossils in
any particular bed might be the same as some of those from
the beds above or below, others were definitely distinctive.
Each formation had, in fact, a suite of fossils peculiar to itself.
By 1799, when his duties had taken him further afield, he had
examined all the formations from the Coal Measures to the
Chalk, and everywhere he found the same types of fossils in
99
m i] I
1
rm
(
Trilobite
Tnnudeus
) U
Graptolite
Didymograptus
P
Graptolite
LJBrachiopod
Conchidium
ornatus murchisom Monograptus knighti
priodon
ORDOVICIAN SILURIAN
Coral Coral Brachiopod

Lithostrotion Zaphrentls Productus
basaltiform cornucopia semireticulatus
CARBONIFEROUS
Brachiopod
Splnftr
undulatus PERMIAN
Ammonite Ammonite
Ceratrtes cassianus Arietites bucklandl
TRIASSIC JURASSIC
/Dinosaur
(Reptile /4
feet high) M
Mollusc
Trtgoma /
urchin)
rlicraster cor-angumum
Belemnite
V Belemnitet
giganteus
JURASSIC / ~\ Iguanodon Ammonite

Hophtes
Mollusc
Spondylus
CRETACEOUS neocomlensis splnosts
Fig. 50
Some characteristic fossils (Cambrian to Cretaceous)
IOO
CORRELATION OF STRATA BY FOSSILS
the same formations and different suites of fossils in different

formations. Smith had discovered that the special assemblage
of fossils representing the organisms that lived during a certain
interval of time never occurred earlier, and never appeared
again. The relative age, or position in the time sequence, of a
formation could thus be ascertained from its distinctive fossils.
The principle of identifying the ages of strata by their
fossils has now been firmly established all over the world. Strata
in Europe and Australia, for example, are now known to be

practically contemporaneous if they contain similar suites of
fossils. The time
required for the migration of a particular
species to another does not introduce any
from one region
practical difficulty, because the intervals represented by even
the smallest divisions of geological time run into hundreds of
thousands or even millions of years, whereas the time required
for migration is relatively short. Everywhere the sequence of
fossils reveals a gradual unfolding of different forms of life, and
thus it becomes possible whole of the fossiliferous
to divide the
stratified rocks into appropriate divisions, each division having
its distinctive fossils and a definite chronological position.
Examples of some characteristic fossils are illustrated in
Fig. 50.
The Geological Time Scale

As the book of earth history is immensely long, it has been
found convenient to classify its contents in much thesame way
as a long book is divided into volumes, chapters, sections, and
paragraphs. Two sets of terms are employed for each division,
because it is essential for clarity of thought to discriminate
between the strata themselves and the time-intervals they
represent :
Divisions of a Book :
Volume Chapter Section Paragraph

Divisions of Strata :
Group System Series Formation

Corresponding Intervals of Time :
Era Period Epoch Age

(396) IOI 8
rocks as the pages of earth history
A coal seam, for example, is a. formation. A

great number of
coal seams, together with the associated shales and sandstones,
make up the series known as the Coal Measures. Below this
series comes the Millstone Grit, a series of massive sandstones
and grits and below this in turn a series characterized by
;
limestone formations. The three series together constitute the

Carboniferous system, the rocks of which came into existence
during the Carboniferous period. The latter, with five other
periods, make up the Palaeozoic era.
The on pages 104 and 105 shows the general scheme

table
of classification by eras and periods which was gradually built
up by the pioneer workers of the last century. It will be
noticed that the eras have names which broadly express the
relations of the forms then flourishing to those of the
life
present day. Beyond the

Palaeozoic era there are enormously
thick systems of rocks passing down into widespread areas of
schists, gneisses, and granites, which form the “ basement
complex ” of any given area. Only rare and obscure remains
of life have been found in the less altered sediments of these
ancient rocks, of no value for defining world-wide systems.
The only collective name for them all is Pre-Cambrian.
Igneous Rocks and the Geological Time Scale
To determine the geological ages of igneous rocks it is

necessary to observe how they occur in relation to associated
stratified rocks of which the ages are known. Volcanic
activity of former periods may be represented by volcanic
tuffs or lava flows, either of which may be found on old land
surfaces or interbedded with sedimentary beds. In the latter
case the age of the activity that of the strata containing the
is
tuffs or lavas. It is less easy to fix the age of an intrusion. The

intrusion must, of course, be younger than the youngest of the
invaded and it must be older than any beds subsequently
strata,
deposited on worn-down surface. A closer upper limit can
its
sometimes be fixed by applying the axiom that a pebble must

be older than the conglomerate in which it occurs. Pebbles of
Shap granite, for example, easily identified by the flesh-coloured
102
v
’
m
'" A),
s Sandstone
• in
<lCo1 boundary
-
Red
u unrepresented
lL
\ Old
dine
of
-Kinc.u
period
conglomerate
Angus
dining
l-.sk,
in
rut
North
(deposited
Cmrge
R
age
(B)
rev
m rioting Bridge,
,S
(,col.
M.
limestone
Tlclwilh
II.
(kulxmifeimis flagstones,
Yorkshire
:
Silurian
'nconforinity
inclined
l
uii
(A)
MEASUREMENT OF GEOLOGICAL TIME
porphyritic felspars {cf. Fig. 12), occur in the conglomerates at
the base of the Carboniferous system in Westmorland. The
intrusion is therefore pre-Carboniferous. Since the granite
itself invades Ordovician volcanic rocks and Silurian sediments,
it must have been emplaced either late in Silurian times or
during the Devonian period. The Shap granite is only one of
an immense number of intrusions associated with the develop-
ment of the Caledonian mountain range, to which reference
is made below (Fig. 51). In the case of the Whin Sill, the time
of intrusion can be fixed as late Carboniferous or early Permian,
because it locally cuts Coal Measures, and pebbles of it occur
in a Permian conglomerate found in the Vale of Eden not far
from the edge of the Pennines.
Although igneous rocks contain no fossils by which they
can be relatively dated, they sometimes contain rare radio-
active minerals ;
and these, in favourable circumstances, have
preserved within themselves a record of the actual period
which has elapsed since they crystallized. Radioactivity is the
process wjhereby the atoms of certain unstable elements (of
which uranium and thorium are the chief) break down into
atoms of other elements ;
the final stable end-products being
the gas helium, and the inert metal lead. Helium, being a
gas, tends to escape, but the lead accumulates. Thus a radio-
active mineral such as uraninite, which begins its existence as
UO a has been
,
engaged ever since in keeping a material register
of time, after the manner of an hour-glass. Uraninites from
comparatively recent igneous rocks contain very little lead,
but in those from very old rocks as much as 10 or even 15 per
cent, of lead may have accumulated at the expense of the
uranium. Since the rate of production of lead from uranium
is known, it is possible by making a chemical analysis of a
uraninite to determine its absolute age in millions of years.

If thorium is also present in the radioactive mineral, as it
often is, its output of lead must also be taken into consideration.
+
The expression Pb/(U 0-36Th) is called the lead-ratio of the
mineral, the symbols representing percentages of the elements
concerned ;
the corresponding age in millions of years is given
approximately by multiplying the lead-ratio by 7,600.
103
THE GEOLOGICAL
Eras Periods and Systems Derivations of Names
( QUATERNARY*
Recent or Holocene )
Holos = complete '
Glacial or Pleistocene 1 Pleiston = most

‘cene’
TERTIARY* from
Pliocene Pleion = more Kainos
Miocene Meion = less = recent
Oligocene Oligos = few
Eocene Eos = dawn ,
The above terms refer to the proportions of modern shells

occurring as fossils in each sub-system
U o~ f CRETACEOUS Creta = chalk

JURASSIC Jura Mountains
5
N £ « I TRIASSIC Y Threefold division in Germany
O « H i
«« i is
NEW
I
**! T j
\
RED SANDSTONE
PERMIAN Permia, anc. kingdom E. of Volga

y )
CARBONIFEROUS Coal-bearing
q S
DEVONIAN or Devon (marine sediments)
O « | OLD RED SANDSTONE (Land sediments of same period)
g i !
SILURIAN Silures, anc. tribe of Welsh borders
*
d | ORDOVICIAN Ordovices, anc. tribe, N. Wales
CL CAMBRIAN Cambria = Wales
PRE-CAMBRIAN ERAS: sometimes described as
PROTEROZOIC Proteros = Earlier

ARCHEOZOIC Archaeos = Primaeval
EOZOIC Eos — Dawn
The term ARCH/EAN refers to the oldest Pre-Cambrian crystalline
rocks of a given region
UNRECORDED INTERVAL (duration unknown)
ORIGIN OF THE EARTH
* In 1760 strata were classified as Primitive (crystalline), Secondary’ (consolidated),

and Tertiary (loose). By an extension of this nomenclature, the period in which we
104
TIME SCALE
Maximum known Thicknesses Approximate Distinctive Life
of Strata in feet dates in years
Present day
Recent Modern Man
| 4,000 25 000 1
Pleistocene Stone-Age Man
J
I.OOo'oOO J
15.000.
Pliocene 13,000 35.000.
Miocene 50.000. 000

21,000 |
Mamma s l and
70.000. 000 -
Oligocene 15,000 Rowerin S Plan «

000 I
Eocene 14,000
000
150.000.
Cretaceous 64,000 190.000.
120,000,000 |
Jurassic 20,000 Reptiles

000
Triassic 25,000 '
|
000
Permian 13,000 Amphibians and

220,000,000
Carboniferous 40,000 Primitive Plants
350.000.
280,000,000
Devonian or Old ) 400.000.
37,000
Red Sandstone 320,000,000
500.000. Fishes
J
Silurian 15,000
000
Ordovician 40,000 Invertebrates
|
000
Cambrian 40,000 I First appearance of
000 }
abundant fossils
Proterozoic
c Unknown "j Scanty remains of
1C \
! in detail
Archaeozoic V Sponges and Seaweeds
but immensely j
Eozoic great at least No direct fossil

1,750,000,000 evidence of Life
Unrecorded Interval
Origin of the Earth at least 2,000,000,000
are now living is called the Quaternary. The Mesozoic era is still sometimes referred
to by the term “ Secondary.”
105
ROCKS AS THE PAGES OF EARTH’S HISTORY
Provided that the radioactive minerals have remained

unaltered since they crystallized, it follows that all minerals
with the same lead-ratio are of the same age. When the
mineral used is known to have crystallized in an igneous rock
or mineral vein at the time when the rock or vein came into
existence, its lead-ratio serves to determine the age of the rock
or vein. If, in turn, the geological age of the latter is known
from its relations towards the associated fossiliferous strata,
then the absolute age corresponding to the geological age is
also known fairly closely. The time-scale based on the most
reliable of the lead-ratios so far determined is given in the table
on page 105.
Earth Movements and the Geological Time Scale

As indicated in Fig. 51, great mountain-building movements
have been particularly active at certain times in the earth’s
history, during which the rocks of long belts of the crust were
intensely compressed. Earth movements of this kind, and also
the resulting folded belts, are described as orogenic (Gr. oros,
—
a mountain). Mountain building in the purely structural
sense, without reference to the subsequent effects of denudation
— is also referred to as orogenesis, and a time of mountain
building is often called a tectonic or orogenic revolution. Nine

such revolutions have already been recognized, three since the
Cambrian and six before (see Fig. 52), with long intervals of
relative quiescence between each pair, during which thick
masses of sediments were accumulated from the denudation
of the lands then exposed. We are now living near the close
of the orogenic revolution of the Cainozoic era. Ideally, an
era can be thought of as a cycle consisting of one of these long
intervals (represented by the sediments then deposited) to-
gether with the orogenic revolution that brought it to an end
(represented by folding of the sediments) . Actually, however,
the revolutions are not found to be strictly contemporaneous
in different parts of the world, and consequently the time
divisions so defined for one continent would not coincide
exactly with those for other continents.
106
)
THE DATING OF EARTH MOVEMENTS

Orogenic movements are essentially tangential to the crust.
Radial movements, in which continental regions are raised or
lowered, with little folding, if any, are distinguished as epeiro-
genic (Gr. epeiros, a continent).Emergent lands and plateaus
result from epeirogenic movements of uplift. Widespread
continental areas are uplifted towards the close of each revolu-
tion, and the lands then become extensive and locally high,
as they are at present. Epeirogenic movements of depression
lead to the development of sunken regions, of which the Black
Sea and the Mediterranean basins are modern examples.
The marine sediments on the land also record fluctuating
changes of level, which no doubt must be largely ascribed to
epeirogenic movements. A
geological period is characterized
by one or more invasions of the land by the sea, during which
the marine beds of that particular system were deposited. Each
invasion can be divided into (a) an advancing phase, as the
sea overflows the slowly subsiding lands ;
culminating in ( b
the phase of maximum flooding of the lands ;
and (c) the phase
of retreat of the sea. As an era consists of several periods, it is
evident that relative to sea level the lands may rise and fall
many times during the interval between successive revolutions.
Occasionally it happens, either within a period or towards its
close, that retreat of the sea is brought about by orogenic move-
ments on a smaller scale than those referred to as revolutions.
Each revolution or minor orogenesis is recorded in the
geological time scale by folding of the rocks, and by the
presence of an unconformity between those rocks and the im-
mediately overlying sediments. The geological age of the
folding is obviously later than that of the youngest of the
folded beds and earlier than that of the oldest beds above the
unconformity. Only the last three or four revolutions can be
effectively dated in this way. The earlier ones involved beds
which cannot be dated by means of fossils, and their ages can
therefore be known only from the lead-ratios of radioactive
minerals. Fortunately, the latter most commonly occur in
granitic rocks that crystallised in the orogenic belts towards the
close of each revolution. Consequently, when they do occur,
they serve to date the revolution with which they are associated.
i°7
£
s Orogenic Igneous Activity
Millions
>. Periods Revolutions Volcanic Rocks and Minor Intrusions

*5 (See Fig. 52) Major Intrusions ( mainly granitic )
Dyke swarms (Fig. 38)

Mourne, Arran, Skye, etc. (Fig. 38)
Plateau Basalts r Antrim — Skye— Faroes
— Iceland
'Granites of Cornwall and Dartmoor

Great Whin Sill and Dykes, N. of England
(Fig. 40)
Midland Valley of Scotland (continued
into Permian)
Midland Valley of Scotland (e.g. Clyde
• Plateau, Campsie Fells, Bathgate and
Garlton Hills, Arthur’s Seat), Anglo-
Scottish Border, Derbyshire
Submarine lavas of Cornwall and Devon
(continued from Devonian)
/•Ben Nevis, Glen Coe, Cheviot. Ben Nevis,

Glen Coe, Oban Sidlaw, Ochil, Pentland
;
|
and Cheviot Hills
Highlands of Scotland ( see below*),
Galloway (Fig. 20)
Lake District, Donegal, Newry, Leinster
Southern Uplands of Scotland

Lake District (Borrowdale Volcanic
Series), Shropshire, Wales (e.g. Snowdon,
{ Cader Idris)
* The Caledonian granitic intrusions of

the Highlands include : Helmsdale,
Peterhead, Hill of Fare, Mt. Battock
( Kincardineshire), Lochnagar, Cairngorm,
Foyers, Rannoch Moor, Etive, Strontian

and Ross of Mull
Wrekin, Caer Caradoc, Charnwood

Forest, Malvern Hills
Fig. 51
Diagram to show the ages of the chief European orogenic movements
and the chief
British occurrences of igneous rocks
(since the late Pre-Cambrian)
108
' — ,
CyclM North Mainly Millions

America Europe of Years
Present Day —
CIRCUM-PACIFIC _ALPINE (including Asiatic
IX 1-70' 20-70 extensions)
Younger APPALACHIAN 200

HERCYNIAN (Central Europe
VIII 200-230 200-250 — S. of Ireland)
Older APPALACHIAN . CALEDONIAN (Norway

350 310-320 British Isles) 400 -1
VII Base of
Cambrian
Lake Superior CHARNIAN (England)
550 (Katanga, Central Africa, 600)
600 H
VI
KILLARNEAN KARELIAN (Lapland — L. Ladoga 800

'
(Lake Superior) 750 850 and S. E. Norway)
1000 H
LAURENTIAN SVECOFENNIAN (S. Finland
(St. Lawrence) 1 050 ’
1050 —Stockholm)
IV
1200 -J
Great Bear Lake

(Canada) and Black (Western Australia, 1250)
Hills of S. Dakota
I35C
1400 J
III
Older Black Hills MAREALBIAN (White Sea) 1600 _|

1600 1600
Manitoba
1750 1800 -I
Conglomerates containing pebbles of oldest known granite

Base not exposed
2000 -I
Fig. 52. Diagram to show the chief otogenic revolutions of geological time,
with approximate dates in millions of years
109
ROCKS AS THE PAGES OF EARTH’S HISTORY

A. K. Wells
Outline of Geological History. Allen and Unwin, London, 1938.
L. J. Wills
The Physiographical Evolution of Britain. Arnold, London, 1929.
C. Schuchert and C. O. Dunbar

Outlines of Historical Geology. Wiley and Sons, New York, 1941.
W. T. Lee
Stories in Stone. Chapman and Hall, London, 1926.
A. Holmes
The Age of the Earth. Nelson and Sons, Edinburgh, 1937.
H.M. Geological Survey

British Regional Geology. (Illustrated accounts of the geology of
the natural regions of Britain.) H.M. Stationery Office,
London and Edinburgh.
J. Phemister Scotland : The Northern Highlands. 1 936.
H. H. Read— The Grampian Highlands. 1935.
J.E. Richey — The Scotland: Tertiary Volcanic Districts. 1935.

M. Macgregor and A. G. MacGregor — The Midland Valley
of Scotland. 1936.
J.
— The
Pringle of South1935. Scotland.
T. Eastwood — Northern England. 1935.

D. A. Wray — The and Adjacent
Pennines Areas. 1936.
F. H. Edmunds and K. Oakley— The P. Central England District.
936
I -
C. Chatwin— East
P. 1937. Anglia.
R. W. Pocock and T. H. Whitehead— The Welsh Borderlands.

»
935 -
B. Smith and T. N. George. North Wales. 1935.
J. Pringle and T. N. George South Wales. — 1937.

F. B. A. Welch and R. Crookall — The Bristol and Gloucester
District. 1935.
no
—
BRITISH STRATIGRAPHY
H. Dewey— South-West England. 1935.

C. Chatwin— The Hampshire Basin.
P. 1936.
R. L. Sherlock London and the Thames Valley. 1935.
F. H. Edmunds — The Wealden District. 1935.
For the study of British Stratigraphy the following map will be

found invaluable :
H.M. Geological Survey — Geological Map of the British Isles.
Scale twenty-five miles to one inch. H.M. Stationery

Office, 1939.
Ill
PART II EXTERNAL PROCESSES AND
THEIR EFFECTS
Chapter VIII
ROCK WEATHERING AND SOILS
Weathering and Climate

Weathering is the total effect of all the various sub-aerial
processes that co-operate in bringing about the decay and
disintegration of rocks, provided that no large-scale transport
of the loosened products is involved. The work of rain-wash
and wind, which is essentially erosional, is thus excluded. The
products of weathering are, however, subject to gravity, and
there is consequently a universal tendency on the part of the
loosened materials to fall or slip downwards, especially when
aided by the lubricating action of water. It is, indeed, only
through the removal of the products of weathering that fresh
surfaces are exposed to the further action of the weathering
processes. No clean-cut distinction between weathering and
erosion can therefore be attempted.
The work accomplished by weathering is of two
geological
kinds : (a) physical or mechanical changes, in
which materials
are disintegrated by temperature changes, frost action, and
organisms and ( b ) chemical changes, in which minerals are
;
decomposed, dissolved, and loosened by the water, oxygen,

and carbon dioxide of the atmosphere, and by organisms and
the products of their decay. The physical, chemical, and bio-
logical agents actively co-operate with one another. Shattering
requires stresses powerful enough overcome the strength of
to
the materials, but the latter is generally greatly reduced by the
preliminary action of decomposition. Shattering in turn
provides increased opportunities for the further penetration of
the chemical agents. Everywhere full advantage is taken of
1 12
DISINTEGRATION OF ROCKS
the joints and bedding planes which, together with the cracks
newly formed, admit air, water, and rootlets down to quite
considerable depths. Thus, although the processes of weather-
ing may be considered separately, it must not be forgotten
that the actual work done is the resultant effect of several pro-
cesses acting together in intimate co-operation.
The materials ultimately produced are broken fragments
of minerals and rocks residual decomposition products,
;
such as clay ; and soluble decomposition products which are

removed in solution. The products of weathering differ widely
in different places according to the climatic conditions and the
relief and configuration of the surface. In general may be
it
said that disintegrationfavoured by steep slopes

is and by the
conditions of frost-ridden or desert regions,
characteristic
while decomposition and solution are favoured by low relief
and by humid conditions, especially in tropical regions. In
the temperate zones the weather is widely variable, and most
of the leading processes are to be found in operation during
one part of the year or another.
Disintegration by Temperature Changes
Frost is an irresistible rock breaker. When water fills

cracks and pores and crevices in rocks and then freezes, it
expands by ten per cent, of its volume, and exerts a bursting
pressure of about 2,000 lb. to the square inch. The rocks are
ruptured and fragments are wedged apart, to become loose
when thaw sets in. Steep mountain slopes and cliffs are par-
ticularly prone to destruction in this way, especially where
joints are plentiful. The frost-shivered fragments fall to lower
levels, and accumulate there as screes of angular debris (Plates
22 and 23a). Above, the ragged sky-line rises out of the ruins.
The screes of Wastwater, the flying buttresses of Snowdon
(Plate 54b), and the great pyramid of the Matterhorn (Plate 50)
are familiar witnesses to the quarrying power of frost. For a
time the screes protect the lower slopes, but their permanent
accumulation is prevented by landslides and avalanches,
”3
transport by rivers and glaciers, and coast erosion by the sea.

In sub-arctic regions wide shore-platforms are produced by the
co-operation of the waves with melting snow and frost in their
ceaseless attack on the cliffs (Fig. 154).
In arid climates the rocks exposed to the blazing sun
become intensely heated, and in consequence an outer shell
expands and tends to pull away from the cooler layer a few
inches within. If the rocks are appropriately bedded or
jointed, actual separation of a curved shell readily takes place.
If the rocks are massive, they must first be weakened by
chemical weathering, but sooner or later rupture occurs.
When the rocks cool down again, the resulting contraction is
relieved by the development of cracks at right angles to the
surface. This part of the process is facilitated by rapid chilling
due to sudden rainstorms, for in the rare downpours of desert
regions the rain may be near freezing-point, and even hail-
stones are not unknown. Shells and flakes of rock are thus
set free and broken down into smaller fragments. The dis-
integration of pebbles is often conspicuous (Plate 23b). In-
dividual minerals swell and shrink and gradually crumble
apart, especially in coarse-grained rocks like granite. Even in
temperate regions the effect of the sun is far from negligible.
Building stones exposed to the sun are found to decay much
more rapidly than those facing north or otherwise in the shade.
In desert and semi-arid regions and in monsoon lands with
a marked dry season a characteristic effect on the outlines of
upstanding hillocks and peaks, especially where they are made
of crystalline rocks, is produced by exfoliation, the peeling off
of curved shells of heated rock. At edges and corners the
ruptures are particularly curved, because there the increase
of temperature penetrates more deeply into the rock than
where flat surfaces are exposed. Sharp corners and projecting
knobs are the firstaway, rounded outlines are developed,
to fall
and the hills become dome-shaped. On convex slopes, suc-

cessive shells may be seen, overlapping like the tiles on a roof,
each ready to fall away as soon as it is liberated by the forma-
tion of radial cracks. This effect is well seen in the inselbergs
(isolated “ island ” mounts) of Mozambique and other parts
1
14
GEOLOGICAL WORK OF ORGANISMS
of Africa (Fig. 144 and Plate 65a). Sometimes after sunset
the loud report of a splitting rock and the noise of its fall
down the mountain side can be heard.
The Role of Animals and Plants
Earthworms and other burrowing animals such as rodents

and termites play an important part in preparing material
for removal by rain- wash and wind. Worms consume large
quantities of earth for the purpose of extracting food, and the
indigestible particles are passed out as worm-casts. In an
average soil there may be 150,000 worms to the acre, and in
the course of a year they raise ten to fifteen tons of finely com-
minuted materials to the surface.
The growing rootlets of shrubs and trees exert an almost
incredible force as they work down into crevices. Cracks are
widened by expansion during growth (Plate 4b) and wedges
of rock are forcibly shouldered aside. Plants of all kinds,
including fungi and lichens, also contribute to chemical
weathering, since they abstract certain elements from rock
materials. Moreover, water containing bacteria attacks the
minerals of rocks and soils much more vigorously than it
could do in their absence. The dead remains of organisms
decay in the largely as a result of the activities of bacteria
soil
and fungi. In this way carbon dioxide and organic acids,
together with traces of ammonia and nitric acid, are liberated,
all of which increase the solvent power of soil-water. The chief
organic product is a “ complex ” of brown jelly-like substances
collectivelyknown as humus. Humus is the characteristic
organic constituent of soil, and water containing it can dis-
solve small amounts of certain substances, such as Iimonite,
which are ordinarily insoluble.
Another effect of vegetation, one which is of vital import-
ance in the economy of nature, is its protective action. Root-

lets bind the soil into a woven mat so that it remains porous
and able to absorb water without being washed away. The
destructive effects of rain and wind are thus effectively re-
115
strained. Forests break the force of the rain and prevent the
rapid melting of snow. Moreover, they regularize the actual
rainfall and preclude the sudden floods that afflict more sterile
lands. For these reasons the reckless removal of forests may
imperil the prosperity of whole communities. Soil erosion is
intensified, agricultural lands are impoverished and lost, and
barren gullied wastes, like the “ badlands ” of North America,
take their place (Plates 32b and 63a). Except after heavy
rainfall, the rivers run clean in forested lands, but after de-
forestation their waters become continuously muddy. Destruc-
tion of the natural vegetation by land clearing and ploughing,
and the failure to replace forests cut down for timber or
destroyed by fire, have had disastrous economic consequences
in many parts of Africa and America. Man himself has been
one of the most prodigal of the organic agents of destruction.
Chemical Weathering
The and solution of rock material by chemical

alteration
processes is by rain-water acting as a
largely accomplished
carrier of dissolved oxygen and carbon dioxide, together with
various acids and organic products derived from the soil. The
degree of activity depends on the composition and concentra-
tion of the solutions so formed, on the temperature, on the
presence of bacteria, and on the substances taken into solution
from the minerals decomposed. The chief changes that occur
are solution, oxidation, hydration, and the formation of car-
bonates. Only a few common minerals resist decomposition,
quartz and muscovite (including sericite) being the chief
examples. Others, like the carbonate minerals, can be entirely
removed in solution. Most silicate minerals break down into
insoluble residues, such as the various clay minerals, with
liberation of soluble substances which are removed in solution.
Limestone is scarcely affected by pure water, but when
carbon dioxide is also present the CaC0 3 of the limestone
is slowly dissolved and removed as calcium bicarbonate,
Ca(HC0 3 ) 2
. The harsh limestone platforms around Settle
116
In the humid conditions of the temperate zone aluminium

hydroxide is not liberated to any important extent, but during
the dry season of tropical and monsoon lands it is precipitated
in a highly insoluble form, and so accumulates at or near the
surface as bauxite (see page 120).
The decomposition of the ferromagnesian minerals may be
illustrated by reference to the simplest type of pyroxene :
Water + carbon dioxide + Ca(Mg,Fe)(Si0 3 )

|-2Si0 2 .2H
?
0 “acid”)
(silicic
2 Soluble bicarbonates of Ca,
-j
Diopside
(. Mg, and Fe.
When AJ 2 0
and Fe 2 0 3 are also present (as in biotite and all
3
varieties of augiteand hornblende) clay and chloritic minerals

and limonite remain as residual products. In the presence of
oxygen limonite is also precipitated from solutions containing
Fe(HC0 3 ) 2 For this reason weathered rock surfaces are
.
commonly stained a rusty brown colour. Ordinary rust is,

in fact, the corresponding product of the action of water and
air on iron and steel.
Chemical weathering contributes to the disintegration of
rocks (a) by the general weakening of the adhesion between
minerals, so that the rock more readily succumbs to the attack
of the physical agents (b) by the formation of solutions which
;
are washed out by the rain, so that the rock becomes porous
and ready to crumble ( e.g the liberation of the grains of a
.
sandstone by solution of the cement) ; and ( c ) by the formation

of alteration products with a greater volume than the original
fresh material, so that (as in exfoliation) the outer shell swells
and away from the fresh rock within.
pulls
Theseparation of shells of decayed rock is distinguished as
spheroidal weathering (Plate 24 b ). It is best developed in well-
jointed rocks which, like many basalts and dolerites, are
readily decomposed. Water penetrates the intersecting joints
and thus attacks each separate block from all sides at once. As
the depth of decay is greater at corners and edges than along
flat surfaces, it follows that the surfaces of rupture become
rounded in such positions. As each shell breaks loose, a new

surface presented to the weathering solutions, and the process
is
is repeated again and again, aided in the appropriate regions

1 18
WEATHERING RESIDUES
by frost. Each successive wrapping around the fresh core
becomes more nearly spheroidal than the one before, until
ultimately the angular block is transformed into an onion-like
structure of concentric shells of rusty and thoroughly rotted
residual material, with perhaps a round and coherent core of
fresher rock still left in the middle. Such cores may stand out
like boulders when their soft outer wrappings have been
washed away by the rain (Plate 25).
Weathering Residues
In dry climatic regions, on steep rock-slopes, and over
massive crystalline rocks, the coating of chemically weathered
material may not become more than a thin film. But where
rain- and soil-water can soak deeply into the rocks, weathering
may proceed to a considerable depth. In the Malay States,
where the rainfall is heavy and evenly distributed, granite has
been converted into soft friable earth to depths of as much as
fifty feet. In tropical regions which have a heavy rainfall
during the wet season, succeeded by a dry season, when the
temperature is high and evaporation rapid, the weathering
residues may be very different. Soil-water is removed by
plants, and water from below is drawn up to make good the
loss so long as the supply holds out. The weak solutions
produced by leaching of the rocks during the wet season
thus become concentrated by evaporation, and the dissolved
materials are deposited, the least soluble being the first to be
precipitated. The products include hydroxides of aluminium

and iron, silica, and various carbonates and sulphates. Most
of these are re-dissolved by the rains of the next wet season,
but as the hydroxides of aluminium and iron are left in a highly
insoluble state, they remain at or near the surface, and gradu-
ally accumulate as a reddish brown deposit to which the name
laterite has been given (L. later, a brick). In depth the material
is variegated and paler in colour, and it is here that alumina
tends to be specially concentrated. At greater depths the bed-
rocks may be intensely decomposed, with abundant develop-
ment of clay minerals.
119
Sheets of rarely become thick — 30

laterite exceptional
feet is
—because, being impervious, their very formation puts an end

to the drainage which is essential to their continued growth.
For this reason, too, the overlying soil becomes infertile, and
then, no longer held together by roots, it is easily scoured away
by torrential rains, leaving large sterile exposures of laterite.
In certain regions, as in parts of India, laterite is found to be
quite soft below its hard and slag-like crust. The soft laterite
can be readily cut into bricks which set hard on exposure to
the sun. This easily worked and valuable building material
was called laterite not because it resembles brick, but because
bricks are made from it.
Quartz, and clay minerals that have not suffered the
ultimate loss of silica, remain cemented in the deposit in various
proportions. There is every possible gradation between
quartz (the chief mineral of sand), clay (mainly composed of
clay minerals), and laterite, as indicated by the following
scheme :
LATERITE
Ofthe laterites themselves, two important varieties are

distinguished: those rich in iron and those rich in aluminium.
The latter, when of high grade, constitute bauxite, the only
ore of aluminium from which it is practicable to extract the
metal on a commercial scale.
120
SUPERFICIAL DEPOSITS
The Mantle of Rock-Waste
The superficial deposits which lie on the older and more

coherent bedrocks form a mantle of rock-waste of very varied
character. In many places the mantle lies directly on the
bedrock from which it was formed by weathering. In this case
quarry sections and cuttings of all kinds generally show a
surface layer of soil, passing gradually downwards through a
zone of shattered and partly decomposed rock, known as the
subsoil, to the parental bedrock, still relatively fresh and un-
broken by weathering agents. In the soil vegetable mould
and humus occur to a varying extent, and under appropriate
conditions they accumulate to form thick beds of peat which
must also be regarded as part of the mantle. Soils develop,
however, not only on bedrock, but also on a great variety of
loose deposits transported into their present positions by
gravity, wind, running water, or moving ice. Although these
deposits will not be considered in detail till later, it is con-
venient to summarise them here, according to their mode of
origin, together with the untransported or sedentary deposits of
the mantle.
THE MANTLE OF ROCK-WASTE

(Continental deposits)
Mode of Origin Characteristic Deposits

Sedentary
Essentially inorganic Residual Gravels, sands, and clays. Terra rossa,
lateritic earths and laterite
Inorganic and organic Soils Including soils on bedrock and on mantle
deposits
Essentially organic Cumulose Vegetation residues : swamp deposits and
peat
Transported
By Gravity Colluvial Screes and landslip deposits
,, Wind /Eolian Sand dunes, sand wastes, and loess
,, Ice Glacial Boulder clay, moraines, and drumlins
,, Melt-water from ice Glaciofluvial Outwash fans, kames, and eskers
,, Rivers (deposited in Lacustrine Alluvium and saline deposits
lakes)
„ Rivers Fluviatile Alluvium, passing seawards by way of
Estuarine or Deltaic deposits into Marine
deposits
121
The Growth and Nature of Soils
As Grenville Cole has finely said :

“ The soil, considered
as a rock, links common stones with the atmosphere, and the
dead dust of the earth with the continuity of life.” The purely
mineral matter of the residual or transported deposits is first
colonised by bacteria, lichens, or mosses. By the partial decay
of the dead organisms mould and humus (see p. 115) begin to
accumulate ;
lodgment is afforded for ferns and grasses ;
berries and winged seeds are brought by birds and the wind ;
and finally shrubs and trees may gain a footing. The rootlets
work down, burrowing animals bring up inorganic particles,
and the growing mass becomes porous and sponge-like, so that
it can retain water and permit the passage of air. Frost and
rain play their parts, and ultimately a mature soil, a complex
mixture of mineral and organic products, is formed. But
though the soil is a result of decay, it is also the medium of
growth. It teems with life, and as the source of supply of
nearly all food it is for mankind the most valuable and least
dispensable of all his natural assets.
Soil may be defined as the surface layer of the mantle of

rock-waste in which the physical and chemical processes of
weathering co-operate in intimate association with biological
processes. All of these processes depend on climate, and in
accordance with this fact it is found that the resulting soils
also depend on the climate in which they develop. Other
factors are also involved : particularly the nature of the bed-
rock or other deposit on which the soil is generated, the relief
of the land, the age of the soil (that is, the length of time during
which soil development has been in progress), and the super-
imposed effects of cultivation.
The influence of the parental material is easily understood.
Sand makes too light a soil for many plants, as it is too porous
to hold up water. Clay, on the other hand, is by itself too
impervious. A mixture of sand and clay makes a loam, which
avoids these extremes and provides the basis of an excellent
soil. A clay soil may also be lightened by adding limestone,
122
CLIMATIC SOIL TYPES
and the natural mixture, known as marl, is also a favourable

basis. Limestone alone, as we have seen, cannot make a soil
unless it In most climates granite de-
contains impurities.
composes slowly and up its store of plant foods very
yields
gradually. Basaltic rocks, on the other hand, break down
much more quickly. Lavas provide highly productive soils
which, even on the flanks of active volcanoes, such as Etna,
compensate the agriculturist and vine-grower for the recurrent
danger and possible destruction.
risk of
These differences are most marked in young soils and in
temperate regions. As the soil becomes older, and especially
when the climate is of a more extreme type, the influence of
long continued weathering and organic growth and decay
makes itself felt more and more. Certain ingredients are
steadily leached out, while others are concentrated. Humus
accumulation depends on the excess of growth over decay, and
this in turn depends on climatic factors. The composition of
the evolving soil thus gradually approaches a certain char-
acteristic type which is different for each climatic region. The
black soil of the Russian steppes, for example, is equally well
developed from such different parent rocks as granite, basalt,
loess, and boulder clay. Conversely, a single rock type, like
granite, gives grey soils in temperate regions ( podsol ), black
soils in the steppes (
chernosem ), and reddish soils in tropical
regions of seasonal rainfall ( lateritic earths ). The colours of
soils are almost wholly due to the relative abundance (or
paucity) of various iron compounds and humus.
Deeply cultivated soils may be more or less uniform
throughout, but this is not the case in purely natural soils. A

vertical cutting through an old natural soil reveals a character-
istic layered arrangement which is called the soil profile. The
latter is clearly developed in the grey soils of the more or less
forested north-temperate belt of Canada, Northern Europe,
and Asia. As the drainage is dominantly descending, iron
hydroxides, and humus derived from the surface layer of
vegetable mould, are carried down in colloidal solution. Thus
a bleached zone is developed, and for this reason the soil type
is called podsol (Russian, ashy grey soil) By the accumulation
.
123
DEPOSITS
ASSOCIATED
AND
TVPES
SOIL
CLIMATIC
124
PODSOL AND CHERNOSEM
of the ferro-humus material at a depth of a few inches or a foot,

accompanied by particles of silt and clay washed down
mechanically, a deep brown or nearly black layer of variable
thickness is formed. This may develop into a hard, well
cemented band, impervious to drainage, which is known as
hard pan. One of the objects of ploughing is to prevent the
growth of hard pan. Otherwise, water-logged conditions may
set in,and there will then be a marked tendency for peat to
accumulate.
Farther south in the grasslands of the steppes and prairies,
summer drought and winter frosts favour the accumulation of
humus, largely provided by the grass roots which die each
year. During the dry season ground-water is drawn towards
the surface, and CaC 0 3 is precipitated, often in irregular
nodules, at a depth of two or three feet. Under the influence
of the ascending calcareous solutions the humus becomes black
and insoluble. Iron hydroxides are therefore not leached out
as in the podsol. The upper layer of the soil profile is black,
becoming brown in depth where the humus content is less.
For this reason the soil type is called chernosem (Russian, black
earth). The black cotton soils of India and the “ black
bottoms ” of the Mississippi flood plains are of similar origin.

G. W. Robinson
Soils, their Origin, Constitution, and Classification. Allen and Unwin
(Murby), London, 1936.
E. Ramann ( translated by C. L. Whittles)
The Evolution and Classification of Soils. Heffer and Sons, Cambridge,
1928.
H. Jenny
Factors of Soil Formation. McGraw-Hill, New York, 1941.
G. V. Jacks and R. O. Whyte
The Rape of the Earth. Faber and Faber, London, 1939.
125
Chapter IX
UNDERGROUND WATERS
Sources of Ground-water
There is abundant evidence for the existence of important
underground supplies of water. At least from the time of the

Babylonians there was a widespread and firmly held belief
that not only springs and wells, but also rivers, were fed and
maintained by water from vast subterranean reservoirs. The
underground streams of limestone caverns supported this
belief, and so did the spurting up of “ the fountains of the
deep ” through fissures riven in alluvial flats by earthquakes.
Moreover, in arid countries like Mesopotamia and Egypt it
was far from obvious that rivers could be maintained by rain-
fall. The author of Ecclesiastes remarked that although “ the
rivers run into the sea, yet the sea is not full,” and inferred that
the balance was restored by a return circulation, underground
from the sea floor, back to the sources of the rivers. How the
sea water rose to such high levels was not explained, and how
it lost its salt before emerging as springs of freshwater remained
an unsolved problem. It was only late in the seventeenth

century that it first came to be realized, notably by Halley,
that the circulation from sea to rivers was not underground,
but through the atmosphere by way of evaporation and rain-
fall. Aristotle’s erroneous conviction that the rainfall was
quite inadequate to supply the flow of rivers was not dispelled
until accurate measurements took the place of mere opinion.
In 1674 Pierre Perrault completed the first quantitative
investigation of the relation between rainfall and stream
flow. He found that in the upper valley of the Seine the
rainfall was actually several times greater than the stream
flow, and so demonstrated for the first time a relationship that
in humid climates seems now to be almost a matter of common
sense.
126
ORIGIN AND OCCURRENCE OF GROUND-WATER
The following scheme shows the various ways in which
rainwater is distributed (see also Fig. 7, page 22) :
Run-off Direct flow down surface slopes

Stream Flow
["Superficial flow through soil and j
sub-soil to streams ]
Downward infiltration into bedrocks \
Percolation - to replenish the ground-water and j- Ground-water

maintain its circulation }
I Absorption by soil and vegetation, t

(.subsequently evaporated y Evaporation
Direct Evaporation J
(During weathering a relatively trifling amount of water is fixed
by hydration in clay minerals and other weathering products.)
Ground-water supplied by rain or snow or by infiltration

from rivers and lakes is described as meteoric. Fresh or salt
water entrapped in sediments during their deposition is dis-
tinguished as connate. During burial and compaction of the
sediments, much of this fossil water is expelled, and during
metamorphism most of it is driven out, carrying with it dis-
solved material which helps to cement the sediments at higher
levels. Steam and hot mineral-laden water liberated during
igneous activity, and believed to reach the surface for the first
time, is known as juvenile water.
The Storage and Circulation of Ground-water

Below a certain level, never far down in humid regions, all
porous and fissured rocks are completely saturated with water.
The upper surface of this ground-water is called the water table.
The water table is arched up under hills, roughly following the
relief of the ground, but with a more subdued surface. In
general, three successive zones may conveniently be recognised
(Fig. 54) :
(a) The zone of non-saturation, which is never completely

filled, but through which the water percolates on its way to
the underlying zones. A certain amount of water is retained

by the soil, which yields it up to plant roots.
(b) The zone of intermittent saturation, which extends from

the highest level reached by ground-water after a period of
127
UNDERGROUND WATERS
Fig. 54
To water table to the surface and its variation of
illustrate the relation of the
level from the top to the bottom of the zone of intermittent saturation after
prolonged periods of wet and dry weather respectively
prolonged wet weather, down to the lowest level to which the

water table recedes after drought.
The zone of permanent saturation, which extends down-
(c)
wards beneath which ground-water is not en-

to the limit
countered. The depths in mines and borings at which the
rocks are found to be dry vary very considerably according to
the local structures, but a limit of the order 2,000 to 3,000 feet
is not uncommon. Juvenile and expelled connate water may,
of course, ascend from much greater depths.
Wherever the zone of permanent saturation rises above
ground level, seepages, swamps, lakes, or rivers occur. When
the zone of intermittent saturation temporarily reaches the
surface, floods develop and intermittent springs appear.
Conversely, many springs and swamps, and even the rivers
of some regions, go more or less dry after long periods of dry
weather when the water table falls below its usual level.
Rocks through which water can pass freely are said to be
pervious. They may be porous and and
permeable , like sand
sandstone or they may be
;
practically non-porous and
impermeable, like granite, but nevertheless pervious because
of the presence of interconnected open joints and fissures
through which water can readily flow. Impervious rocks are
those through which water cannot easily soak ; they may be
of two kinds : porous, like clay, or relatively non-porous, like
massive unfissured granite. It should be noticed that although
porosity is essential in order that a formation can be readily
permeable by water, it is not a sufficient condition. The size
and arrangement of the openings must also be such that
continuous through-channels for the free passage of water
128
CIRCULATION OF GROUND-WATER
In clays there are no continuous passage-ways

are available.
wide enough to permit the flow of water, except by slow
capillary creep.
Ordinary sand or gravel has a porosity of about 35 per cent.
”
(i.e. the material in bulk is made up of 35 per cent, of “ voids
and “
65 per cent, of solids ”), but this drops to about 15 per
cent, in common sandstones, according to the degree of com-
paction and the amount of cementing material. Clay, al-
though it is impervious, may have a porosity of over 45 per
cent. By compaction under pressure and the squeezing out
of water the porosity drops gradually, falling to as little as
5 per cent, in some shales, and to 3 per cent, in slates. In
limestones the porosity ranges from 30 per cent, in friable chalk
to 5 or less in indurated and recrystallised varieties. Lime-
stone, however, may carry a great deal of water in joints and
other channels (including caves) opened out by solution. The
porosity of massive igneous and metamorphic rocks is generally
less than 1 per cent., but here again water may circulate in
appreciable quantities through the passage-ways afforded by
interconnected joints and fissures.
Alternations of pervious and impervious strata, especially
when folded, faulted, and jointed, form underground reservoirs
and natural waterworks of great variety. Where the catchment
area is sufficiently high the water slowly migrates through the
most pervious formations towards places at a lower level where

the water can escape to the surface. It may emerge through
natural openings (seepages and springs) or through artificial
openings (wells), or it may feed directly into rivers or lakes or
even discharge through the sea floor. The sustained flow of
rivers which, like the Nile, successfully cross wide stretches of
desert is to some extent due to supplies received from under-
ground sources.
Springs and Wells
When rainwater sinks into a pervious bed, such as sand-

stone, it down until it reaches an underlying impervious
soaks
bed, such as clay or shale. If the surface of the junction is
129
UNDERGROUND) WATERS
inclined, the water flows down the water-tight slope, to emerge

where the junction is intercepted by a cliff or valley side
(Fig. 55a). A general oozing out of the water along the line
of interception is called a seepage. More commonly a line of
Fig. 55
To illustrate various conditions giving rise to springs (see text)
localised springs The other diagrams of Fig. 55

appears.
illustrate variousexamples of other structures favouring the
development of springs. In (Z>) a fault brings pervious sand-
stone against shale which, being impervious, holds up the
water. Springs are localised along the line of the fault, and the
low ground on the left is marshy. In (c) water enters the joints
130
SPRING? AND WELLS
in a massive rock, such as granite, and issues in appropriate

places. In ( d ) water impounded by a dyke escapes along the
outcrop of the junction. In (e) the upper spring is thrown out
by a conformable bed of shale, as in (a) the Jower spring
;
appears at the outcrop of an unconformity, the underlying

folded rocks being impervious. In (f) water enters jointed
limestone, widens the joints by solution, forming caves and
underground channels down to the impervious base of the
formation. The latter holds up the water and allows it to
drain out, sometimes as an actual stream, where a valley has
To illustrate the structural conditions favourable to artesian wells
been excavated through the limestone into the underlying

rocks.
Wells are simply holes dug or bored or drilled into the
ground to a depth at which water-bearing permeable formations
or fissured rocks are encountered. Shallow wells, as shown in
Fig. 54, may dry up at certain seasons, unless they tap the zone
of permanent saturation. Ground-water percolates into the
bottom of the well, and rises to a level that depends on the
head of pressure behind it. Pumping or lifting may be neces-
sary to bring the water to the surface. In selecting sites for
shallow wells special precaution is necessary to preclude
contamination by germ-laden water which might drain into
I
3I
UNDERGROUND WATERS
the source of supply from farmyards and
cesspools. The ground-water from more
deep-seated formations is preferable for
human consumption, as it is more likely
to be free from the dangers of surface
contamination.
Artesian wells are those in which the
water encountered in depth is under a
sufficient hydraulic pressure to force it
to overflow at the surface. The neces-
sary conditions are : (a) an inclined or
broadly synclinal water-bearing forma-
tion, or aquifer, enclosed on both sides by
watertight beds (b) exposure of the rim
;
of the aquifer over a catchment or intake

area at a sufficient height to provide a
hydraulic head at a level above the
ground where the wells are sunk (c) a ;
sufficient rainfall to furnish an adequate

supply of water and ( d ) absence of a
;
ready means of escape of the water except

through the wells. The term artesian is
sometimes extended to include deep wells
in which the water approaches the surface
but does not actually reach it.
The London Basin (Fig. 57) exem-
plifies these conditions very clearly. The
aquifer is the Chalk, with sandy beds
above and, locally, below. The enclos-
ing impervious formations are the Lon-
don clay above and the Gault clay
below. Water falling on the Chalk,
where it is exposed along the Chilterns
to the north, and the North Downs to the
south, sinks into the basin and accumu-
lates there —or did so until the original
resources became impoverished by the
insatiable thirst of London. The water
132
YORKSHIRE
HOLES,
SWALLOW
ARTESIAN WELLS
in the tapped in the London area by hundreds of

Chalk is
wells sunk to depths up to COO or 700 feet. Up to a century

ago the Chalk was saturated, and when the fountains in
Trafalgar Square were first constructed the water gushed out
well above the surface. In more recent years the enormous
supplies which have been drawn from the Chalk reservoir
have exceeded replenishment by rainfall on the rims of the
basin. The level of the water in the Chalk has therefore
fallen, and water can now be raised to the surface only by
pumping.
In North and South Dakota an important aquifer dips off
the edge of the Black Hills, and carries a copious supply of
water beneath the plains to the east. Over an area of 15,000
square miles the water can be tapped by artesian wells. The
largest artesian basin in the world is that of Queensland and
adjoining parts of New South Wales and South Australia.
The catchment area is in the Eastern Highlands where a wide-
spread series of soft Jurassic sandstones come to the surface.
These sandstones, with their accumulated stores of water,
underlie an area of about 000,000 square miles. Without the
artesian wells, some of which are 4,000 to 5,000 feet deep, much
of this vast region would be a barren waste. In this area it
is suspected that only part of the water is meteoric. In some
of the wells the enormous pressure of the water, the abundance
of gases, and the composition of the dissolved constituents all
suggest that juvenile sources may also contribute in depth,
the pressure being partly due to gases and partly to the weight
of overlying rocks.
Many of the oases of the Sahara and other deserts owe their
existence to the local emergence of artesian water at the
surface. Fertilised by the escaping underground water,
vegetation flourishes amazingly and makes “ a paradise in a
and glaring rocks.” Between the Chad
setting of blazing sand
basin and the Sahara the highlands of Erdi and Ennedi con-
stitutean important catchment area. There the occasional
rains are readily absorbed by bare sandstones which continue
underground far across Libya and Egypt. Many a traveller
has died of thirst in the heart of the desert with water only a
(396) I
33 XO
UNDERGROUND WATERS
Cafrhmenf
Fig. 58
Idealized section across the Sahara to illustrate conditions favourable to the
development of oases
few hundred feet beneath him. Where this usually inaccessible

water emerges through fissures or artesian wells, or is brought
to the surface by anticlines, or where the desert floor itself
has been excavated by the wind down to the level of the
ground-water, oases occur (Fig. 58). South of Aswan, the Nile
taps part of this artesian water where its channel cuts into an
aquifer locally brought up by an anticline.
Swallow Holes and Limestone Caverns

The solution by rainwater charged with
of limestone
carbon dioxide has already been described (page 116). In
limestone districts water readily works its way down through
joints and along bedding planes until it reaches an impervious
layer, which may be within the limestone formation or beneath
it. The water then follows the natural drainage directions
until it exit, perhaps many miles away from the
finds an
intake. Once a through drainage is established, but not
until then, dissolved material is carried away and fresh
supplies coming
in from above continue the work of solu-
tion, localised along joints and bedding planes, until a laby-
rinth of interlacing channels and caves is dissolved out of the
limestone (Fig. 55 f).
The surface openings become gradually enlarged in places
where the contours of the ground favour a special concen-
tration of the flow-off, and funnel-shaped holes, known as
swallow holes or sink holes are developed. By continued solution
beneath these, and the falling in of loosened joint blocks, the
134
LIMESTONE CAVERNS
holes may be enlarged into roughly cylindrical shafts or chasms

communicating with great vaulted chambers, perhaps hundreds
of feet below (Plate 26a). One of the most impressive of these
giant swallow holes in Britain is Gaping Ghyll (Plate 26b), on
the south-east slopes of Ingleborough. The shaft goes down
for 365 feet into a chamber 480 feet long and 110 feet high.
The water escapes through an intricate system of passages into
Ingleborough Cave, whence it emerges as the Clapham Beck.
In the Ariege (French Pyrenees) two great chasms connected
by a long grotto, with a permanent stream from top to bottom,
have been explored down to a depth of 1566 feet, and severed
even deeper examples occur in Italy. The “ cavernous lime-
stone ” plateau of Kentucky has over 60,000 sink holes and
hundreds of caves, including the great Mammoth Cave, which
itself has over 30 miles of continuous passages. Another
famous American cave, the Carlsbad Cavern of New Mexico,
has a “ Big Room ” nearly 4,000 feet long, with walls over
600 feet apart, and a ceiling rising to a height of 300 feet
(Plate 27a).
Occasionally the roof of a cave collapses and leaves another
”
kind of “ sink at the surface. When the roof of a long
underground channel falls in, a deep ravine, floored with
limestone debris, further diversifies the irregular limestone
topography (Fig. 59). Sometimes one part of the roof
holds firm, thus forming a natural arch or bridge (Fig. 60).
Limestone regions such as those referred to above, having
a roughly etched surface, pitted with depressions due to
solution or roof collapse, and with underground drainage in
place of surface streams, are said to have a karst topography,
from the prevalence of these features in the Karst Plateau
north-east of the Adriatic coast between Trieste and Cattaro
(see Fig. 166).
In addition to the streams which flow through the under-
ground network of passage-ways, there is generally a slow
seepage of lime-charged water from innumerable joints and
crevices in the roofs and walls of caves. Calcium carbonate is
deposited when a hanging drop of such water begins to
evaporate or to lose part of its carbon dioxide. When the
135
UNDERGROUND WATERS
[A . Horner & Sons, Settle

Fig. 59
Trow Ghyll, on the slopes of Ingleborough, Yorkshire. A
dry valley due to
collapse of the roof of a former limestone cavern
drop falls on the dry floor of a deserted channel, the remaining

calcium carbonate is deposited. Thus, long icicle-like pen-
dants, called grow downwards from the roof; and
stalactites,
thicker columns, distinguished
as stalagmites, grow upward
from the floor (Plate 27). In time stalactites and stalagmites
unite into pillars, and these are commonly clustered together
in forms resembling organ pipes and other fantastic
shapes
that are often given fanciful names. Where the water trickles
out more or less continuously along a roof joint, a fluted
curtain or wavy screen may grow across the cave. When
the
water comes through a bedding plane it builds up encrustations
Natural Bridge, Virginia. Part of the roof of a
former limestone cavern
from wall which look like frescoes or “ frozen cascades.”

to floor
The internal decoration of caverns in which these varied
structures have grown in profusion produces an underground
scenery of weird and fascinating beauty.
Hot Springs and Geysers
Ground-water that has circulated to great depths in deeply

folded rocks becomes heated, and if a sufficiently rapid ascent
to the surface should be locally possible, it will emerge as a
137
UNDERGROUND WATERS
warm spring. Such conditions are rare, however, and really

hot springs generally occur in regions of active or geologically
recent vulcanism, where they owe their high temperature to
superheated steam and associated emanations which rise from
subterranean sources and mingle with the meteoric circu-
lation at higher levels. It is probable that some hot
springs may be a mixture of meteoric water with hot water
expelled from underlying rocks that are undergoing meta-
morphism.
There are three volcanic regions where hot springs and
geysers occur on an imposing scale Iceland, Yellowstone
:
Park, and the North Island of New Zealand. The waters are
highly charged with mineral matter of considerable variety.
The Mammoth Hot Springs of Yellowstone Park are rich in
calcium carbonate derived from neighbouring limestones.
This is deposited at the surface as mounds and terraces of
travertins (Plate 28a). In all three regions many of the springs
are alkaline and carry silica in solution, which is similarly
deposited as siliceous sinter or geyserite (Plate 28b). As to the
water itself, investigations show that about 80 to 90 per cent,
is
ordinary meteoric ground-water. In Iceland, for example,

much of it comes from melting snow. However, the minor
constituents carried in solution include many unusual elements,
a peculiarity which points to a juvenile source and suggests
that they must have been swept into the local meteoric water
by associated juvenile water or steam. The existence of such
steam, however, is not merely a matter of inference. Actual
borings through the rhyolite lavas of Yellowstone Park en-
countered vast quantities of high-pressure superheated steam.
In one case the temperature at a depth of 245 feet was found
to be 205°C. Moreover, where steam remains uncondensed
by admixture with cold ground-water, steam fumaroles dis-
charge at the surface.
Geysers are hot springs from which a column of hot water
and steam is explosively discharged at intervals, spouting in
some cases to heights of hundreds of feet (Plate 29). The term
comes from Geysir, the Icelandic name for the Great Geyser,
which is the most spectacular member of a group situated in
138
[Cr. .4. (rnint, L'.S. (rcol. Surrey
Gcvser in eruption. Old Faithful. Yellowstone National Park
THE GREAT GEYSER OF ICELAND
a broad valley north-west of Heckla (Fig. 61). It will serve
as a typical example. A mound of geyserite, built up by
deposition from the overflowing water, surrounds a circular
basin, about 69 feet across and 4 feet deep, filled to the brim
with siliceous water at a temperature of 75° to 90° C. From
Geysir
Fig. 61
Schematic section through Geysir (the Great Geyser of Iceland) to illustrate the
conditions appropriate to intermittent eruption showing subterranean reservoirs
;
fed by ground-waters heated from below by the ascent of superheated steam

(After T. F. W. Barth)
the middle of the basin a pipe, also lined with geyserite, goes
down about 100 feet. At the bottom the temperature of the
water is well above that at which the water would boil if it
were not for the pressure due to the weight of the water column
above it. But the continued accession of superheated steam
through cracks in the pipe gradually raises the temperature
until eventually the boiling-point is reached far down in the
139
UNDERGROUND WATERS
pipe. A certain amount of water then suddenly expands into
steam which heaves up the column and causes an overflow
from the basin. This so relieves the pressure on the superheated
water in depth that it violently flashes into a vast volume of
steam which surges up with irresistible force, hurling the
water into the air, sometimes as much as 200 feet.
In some geysers the amount of water discharged is many
times greater than that contained in the pipe and basin. In
these cases the pipe must therefore communicate with a
neighbouring underground chamber into which continuous
supplies of both meteoric water and juvenile steam have access.
The caves and tunnels which sometimes occur in lava flows
would provide the sort of reservoir required. During each
—
period of quiescence the whole system underground reservoir,
—
communicating channels, pipe, and basin rapidly fills up
and gradually rises in temperature, until the quiet phase of
the cycle is terminated by the paroxysm of high-pressure boiling
which brings about a roaring eruption of water and steam.
Deposition from Ground-waters
As indicated by the discussion on pages 116-18, the chief

ingredients carried in solution by ground-water are the bi-
carbonates of calcium, magnesium and iron, and colloidal
silica. Examples of deposition of calcium carbonate as stalac-
tite and stalagmite in limestone caverns, and as travertine
from the waters of hot springs, and of silica as siliceous sinter or

geyserite, also from the waters of hot springs, have already
been mentioned. These cases are easy to understand, as they
are clearly due to evaporation, loss of carbon dioxide, or cool-
ing. The return of dissolved material to the rocks through
which ground-water is circulating involves the operation of
much more complex and delicately balanced processes which
are still but little understood. Precipitation may be brought
about by such factors as loss of gases and consequent decrease
of solvent power ; cooling while waters are ascending changes
;
of pressure during circulation ; or the mingling of waters from

140
DEPOSITION FROM GROUND-WATERS
different sources. Moreover, as a result of reaction between

solutionsand the materials through which they pass, one sub-
stance may be precipitated and another taken into solution.
Only brief notes on some of the chief results of deposition from
ground-water can be given here.
Cementation of porous sediments occurs when deposition
takes place between the particles of the rock. Loose sands may
thus become calcareous, siliceous, or ferruginous sandstones,
according to the nature of the cement introduced between the
grains.
Replacement, or substitution of one substance for another, is
well exemplified by parts of certain limestone formations in
which calcite has been replaced by dolomite or chalybite or
silica. The change often takes place atom by atom, so that
the original structures are perfectly retained. Calcareous
fossil shells may thus be transformed to any of the materials
mentioned, as well as to others, less common, with complete
preservation of the original form and of the most intricate
structural details. Even organic matter can be replaced in
this way. Water-logged tree trunks buried in sand, or stumps
of trees overwhelmed by volcanic ash, may thus be petrified.
Such fossil wood, with all the tissues perfectly reproduced in
opaline silica or calcite, is notably abundant in Yellowstone
Park, and in parts of Burma and Queensland.
Nodules and concretions form in sediments by concentrated
cementation or replacement, where deposition is localised
around a nucleus of some particular mineral grain or fossil
which initiates the precipitation. Their composition is gener-
ally widely different from that of the formation as a whole.
Flint, for example, is a concretionary form of silica occurring
in the Chalk as scattered nodules of irregular shapes and also
as tabular sheets and vertical stringers (see Fig. 151). Ground-
water percolating through the Chalk at some stage after its
uplift from the sea floor picked up colloidal silica from minute
and easily dissolved opaline sponge spicules dispersed through
the formation. By replacement of calcium carbonate, wher-
ever conditions were favourable, the silica was then deposited
as flint which, being insoluble, was not again redissolved.
141
UNDERGROUND WATERS
Similar deposits of silica occurring in other limestones are
usually referred to as chert. Calcareous and ferruginous
nodules, characteristically of ellipsoidal shapes, are common
in some of the clays and shales of the Jurassic and Carboniferous
systems in Britain.
Small veins of common minerals, such as calcite and quartz,
may be formed from ground-waters in joints and fault fissures,
or in gashes across the limbs of folds. In tightly folded rocks
and in areas of regional metamorphism, irregular quartz veins
are locally very abundant. Many of these have been deposited
in tension clefts from siliceous water “ sweated out ” of the
original rocks during orogenesis. Most mineral veins, however,
and especially those containing commercially valuable ores,
have been deposited from hydrothermal solutions of juvenile
origin, generally in association with the expiring igneous
activity of the period concerned.

H. B. Woodward
The Geology of Water-Supply. Arnold, London, 1910.
F. Dixey
A Practical Handbook of Water-Supply. Allen and Unwin (Murby),
London, 1931.
C. F. Tolman
The Geology of Ground Water. McGraw-Hill, New York, 1937.
W. M. Davis
Origin of Limestone Caverns. Bulletin of the Geological Society of
America, Vol. XLI., pp. 475-628, 1930.
W. M. McGill
Caverns of Virginia. University of Virginia, 1933.
N. Casteret
Ten Tears under the Earth. Dent, London, 1939.
E. T. Allen and A. L. Day
Hot Springs of the Yellowstone National Park. Carnegie Institution of
Washington (Publication No. 466), 1935.
142
Chapter X
RIVER ACTION AND VALLEY DEVELOPMENT
Some General Considerations
The necessary conditions for the initiation of a river are an

adequate supply of water and a slope down which to flow.
As we have already seen, Perrault was the first to discover that
an adequate supply is provided by the rainfall. From the
result of his pioneer work in the valley of the upper Seine he
justly concluded that “ rain and snow waters are sufficient to
make Fountains and Rivers run perpetually.” Rivers are
partly fed from ground-waters, and some have their source in
the melt-waters from glaciers, but in both cases the water is
derived from the meteoric precipitation. In periods of drought
rivers may be kept flowing, though on a diminished scale,
entirely by supplies from springs and the zone of intermittent
saturation. When these supplies also fail, through the lowering
of the water table, as commonly happens in semi-arid regions,
rivers dwindle away altogether. However, even then water
may still be found not far below the surface where the floors
of such intermittent streams have a covering of alluvium.
The initial slopes down which rivers first begin to flow are
provided by earth movements or, more locally, by volcanic
accumulations. Many of the great rivers of the world, e.g.
theAmazon, Mississippi, and Congo, flow through widespread

downwarps of the crust which endowed them with vast ready-
made drainage basins from the start. The majority of rivers,
however, originated on the sides of uplifted regions where,
often in active competition with their neighbours, they gradu-
ally evolved their otvn drainage areas.
Most rivers drain directly into the sea. But in areas of
internal drainage permanent or intermittent streams terminate
in lakes or swamps having an area such that evaporation just
balances the inflow, the conditions being such that the water is
H3
unable to accumulate to the level at which it could find an

outlet. Notable examples occur in Central Asia and Australia.
The development of a river valley depends on the original
surface slope on the climate, which determines the rainfall
; ;
and on the underlying geological structure, which determines

the varied resistance to erosion offered by the rocks en-
countered. Where a newly emergent land provides an initial
seaward slope, the rivers which flow down the slope, and the
valleyswhich they excavate, are said to be consequent. The
down which tributaries
valley sides constitute secondary slopes
can develop these streams and their valleys are distinguished
;
as subsequent. Later, of course, other generations of tributaries

are added. A
main river and all its tributaries constitute a
river system, and the whole area from which the system
derives water and rock- waste is its basin. Weathering con-
tinuously supplies rock-waste, which falls or is washed by rain
into the nearest stream. latter carries away the debris
The
contributed to it, same time acquires still more by
and at the
eroding its own channel. Valleys develop by the removal of
material, all of which is carried away by the streams which
drain them. The load acquired by the main river is ultimately
transported out of the basin altogether or deposited in its
lower reaches. Deposits are, of course, dropped on the way at
innumerable places, but these are only temporary halts in the
journey towards the sea. Rivers are by far the chief agents
concerned in the excavation of valleys, not merely because of
their own erosive work, but above all because of their enormous
powers of transportation.
The excavation of valleys in turn involves the development
of residual landforms (Plate 30), such as peaks, hills, ridges,
and isolated pinnacles of resistant rocks (Fig. 62). The original
uplifted area thus gradually dissected into a varied and
is
slowly changing landscape. Sooner or later, as rock-waste

continues to stream away from every part of the area in turn,
valleys are widened and the intervening divides are reduced,
until the region may be worn down to a low-lying surface of
faint relief which is called a. peneplain (almost a plain). The
whole sequence of changes passed through during this long
144
CYCLE OF EROSION
Fig. 62
A Dartmoor tor summit of a residual hill

at the carved from well-jointed
;
granite by rain and wind. Bowerman’s Nose, Manaton, Devonshire
evolution is called a cycle of erosion. It is convenient to divide the

cycle into three successive stages which, by analogy with a
lifetime, are referred to as youth, maturity, and old age. Each
stage is characterised by distinctive types of landforms.In this
chapter we be
shall mainly concerned with the cycle of erosion
in humid regions, where the work of rain and rivers is in con-
trol. In other climatic regions, such as deserts, where the work
of the wind is chiefly in evidence, and frigid regions, where
glaciers and ice-sheets are the supreme agents, the correspond-
ing cycles of erosion involve the development of landforms which
are very different from each other and from those of humid
regions. The cycle of river erosion does not always run its
course without interruptions. Further uplift may take place
while a cycle is still uncompleted, or the area may be glaciated.
For both reasons many of our British rivers are still in their
infancy, and the landscapes associated with them are partly
inheritances from glacial and pre-glacial time-s.
145
Nevertheless, the valleys of the present day, whether in

Britain or elsewhere, provide representatives of all the stages
in the cycles of erosion now in progress. Weare thus enabled
to study by how they gradually develop.
direct observation
We shall have to consider in turn how the lands are eroded
and the rivers acquire their load of rock-waste how the rivers
;
themselves erode their channels and transport and deposit

their load ; how valleys are lengthened, deepened, and
widened ; how river systems develop in plan and compete
with their neighbours in the struggle for space how topography
;
varies according to the stage reached in the cycle of erosion ;
and how it is further varied by interruptions of the normal

cycle.
Rain Erosion
The chief mechanical effect of rain as it pelts down on the

mantle of surface materials is to wash loose particles to lower
levels. Eventually the rain-wash is swept into rills and streams.
During a sudden cloudburst a sheet of flowing water may be
locally produced which undercuts and removes the turf on
sloping ground, sweeping the underlying soil to the foot of the
slope, and leaving a long gash in the hillside. The gash is
gradually deepened into a gully by recurrent rains, and as
soon as the water table is tapped it begins to carry water and
becomes a rivulet. This is one of the ways in which tributaries
originate.
In semi-arid regions, where the occasional rains are often
exceptionally violent, rain-gashing reaches spectacular pro-
portions in sloping ground underlain by clay or soft earthy
deposits. Such land is sculptured into an intricate pattern of
gullies and small ravines, separated by sharp spurs and
buttresses. The gullies grow backwards into the adjoining
upland, and the intervening ridges in turn are further cut up
into smaller ribs and trenches (Plate 32a). Tracts of the
almost impassable country so developed are graphically
described as badlands in North America, where they are widely
scattered from Alberta to Arizona (Plate 32b).
146
RAIN EROSION AND SOIL CREEP
The curious structures known as earth-pillars develop

locally from spurs lefton the slopes of valleys carved in boulder
clay (Plate 33a). Wherever a boulder is encountered while
the surface is being worn down by rain erosion, it acts like an
umbrella over the underlying clay. Where the slope is sheltered
from strong winds, earth-pillars of surprising height, each sur-
mounted by a protective cap, may then be etched out by the
rain. Some of the pillars of Botzen in the Tyrol reach a height
of 69 feet, and smaller examples occur in favourable situations
in Scotland and other regions where boulder clay and easily
eroded conglomerates (Plate 33b) are exposed to the weather.
Soil-creep and Landslides
Between the extremes of rain-wash on gentle slopes and

from precipices there are various kinds of mass
rock-falls
movements of surface materials downslope, which result from
the action in different combinations of water, frost, organisms,
and gravity. All these processes co-operate with weathering
in widening valleys (Plate 31).
Slow downward movement of soil on hillsides, known as
evidenced by tilted fences, bulging walls, and the
soil-creep, is
outward curving of tree trunks near the ground. Interstitial

rain-washing, together with various less obvious processes, all co-
operate. Ice crystals heave up stones and particles of all sizes
during frost. The outward heave and
the subsequent drop
when thaw are both in a downhill direction. Imper-
sets in
ceptible movements due to expansion and contraction, or to
the wedging action of rootlets, take place under the control of
gravity, so that the cumulative effect is downwards. Even
the sub-soil and the upper parts of the bedrock share in this
movement. The upper ends of steeply dipping or cleaved
beds are prised apart by frost and rootlets until they gradually
curve over in the downhill direction. This results in
apparent dips which, exposed in cuttings and gullies, may
depart considerably from those of the undisturbed formation.
When soil becomes thoroughly water-logged, as happens
147
particularly in colder climates and at high altitudes after the

melting of snow, the downward creep passes into actual flow
and is then described as solifluction (soil-flow). Flows of mud
and peat, referred to as bog-bursts, occur periodically in
Ireland.
The same group of processes operating in screes leads to
similar results.Moreover, screes become bodily unstable as
fragments fed from above gradually steepen the slopes. After
a thorough soaking and lubrication by water from rain or
melting snow, the weight is increased, while friction and the
angle of repose (normally 25°-35°) are decreased. The scree
thereupon begins to slide. In the gorges of rivers in the Hima-
layas and other ranges of vigorous relief such debris-slides or
sometimes occur on a gigantic scale.
rock-avalanches The
valley below may be dammed across. A lake then forms on
the upstream side and, bursting suddenly through the wall of
debris, may cause a disastrous flood. In sub-arctic regions,
where frost and thaw are especially active, rock-glaciers or stone-
rivers spread outwards on suitable slopes. The mechanisms of
these phenomena are intermediate between these responsible
for soil-creep and debris-slides.
It is a matter of common observation that steep grassy
slopes above valley floors are often scored at intervals with
or “ sheep-tracks ” in the soil, particularly where
little terracettes
the stream undercuts and steepens the banks. These jfeatures

are due to small landslips which, as they are of frequent
occurrence, contribute largely to the removal of mantle
deposits. The essential conditions are lack of support in front
and lubrication behind. Similar conditions favour landslides on
a bigger scale, wherever slumping (Fig. 63) or sliding (Fig. 64)
can occur on the sides of undercut slopes, precipices, and cliffs,
or of road, railway, and canal cuttings, particularly where
heavy massive rocks ( e.g plateau basalt) overlie weak and
easily lubricated formations.
Slumping takes place on a
curved slip and a backward tilting of the surface often
plane,
results. Sliding occurs when bedding and cleavage planes,
fault fractures, or joints dip towards a valley or other depression
at a dangerous angle.
148
LANDSLIDES
Plate 34 illustrates a landslide (as seen in 1935) that

obstructed a tributary of the Ticino valley in 1927. Several
years previously a crack that appeared near the top of a hill
To illustrate slumping on curved surfaces in unconsolidated or other weak

formations ;
showing characteristic back-tilting at the surface
on the right-hand side had slowly developed into a gaping

fissure about 6 feet wide. Subsequent movements were care-
fully measured by setting up a line of stakes and recording
Fig. fit
To illustrate conditions favouring rock slides on lubricated bedding planes
their positions several times daily. One day in 1927 a sudden

movement of 8 or 10 feet occurred. A warning was im-
mediately telephoned, and the danger zone was evacuated.
<s9i.)
149 11
Forty-eight hours later the long-threatened slide took place,

fortunately without loss of life.
The sediments of the flat terrace seen in the foreground of
Plate 31 were deposited in a lake that formerly occupied a long
stretch of the Upper Rhine and its tributaries. The lake was
impounded by a gigantic prehistoric landslide that blocked
the main valley near Flims, about 30 miles downstream. The
landslide debris itself covers an area of some 20 square miles,
and formed a dam (since cut through by the river) not less
than 2,700 feet thick.
Erosion and Transport by Rivers
The work of river erosion is accomplished in four different

ways, all of which actively co-operate.
(a) Corrosion is the solvent and chemical action of the water
of the stream on the materials with which it comes into contact.

(
b ) Hydraulic action is the mechanical loosening and removal
of materials by water alone. Flowing water can sweep away
loose deposits and wash out particles from weakly resistant
sediment. A river may not acquire much new material by
sluicing its channel, but the coarser part of the load is likely to
be dropped over and over again during transit (Plate 35a),
and each time it has to be picked up afresh before transport
can proceed. Where the current is strong enough, as when a
river is in spate, water may be driven under jointed slabs with
sufficient force to hoist them up, turn them over, and so make
them available for transport.
(c) Corrosion is the wearing away of the sides and floor with
the aid of the boulders, pebbles, sand, and silt which are being
transported. By scour and impact even the hardest bedrocks
are excavated and smoothed. The is one of
drilling of pot-holes
the most potent methods of down-cutting. These develop in
the depressions of rocky channels or from hollows formed
where boulders and pebbles, acting like drilling tools, are
rapidly swirled round by eddies (Plate 35b). Vertical holes
are cut deeply into the rock as the water plunges in and keeps
the drilling tools in action by its spiral motion. As the boulders
150
RIVER EROSION AND TRANSPORT
wear away, and are swept out with the finer materials, new
ones take their place and carry on the work. In front of a
waterfall very large pot-holes may develop in the floor of the
“ plunge-pool.” This leads to deepening of the channel, and
at the same time a combination of hydraulic action and cor-
rasion undermines the ledge of the fall. The eddying spray
behind the fall itself is particularly effective in scouring out
the less resistant formations that underlie the ledge (Fig. 68).
Blocks of the ledge are then left unsupported and fall away at
intervals, thus causing a migration of the fall in an upstream
direction, and leaving a gorge in front.
(id
)
Attrition is the wear and tear suffered by the transported
materials themselves, whereby they are broken down, smoothed,
and rounded. The smaller fragments and the finer particles
liberated as by-products are then more easily carried away.
The solid part of the load carried by a river includes the
rock-waste supplied to it by rain-wash, surface creep and slump,
etc., and by tributaries and external agents such as glaciers
and the wind, together with that acquired by its own destruc-
tive work, as described above. The debris is transported in
various ways. The smaller particles are carried with the
stream in suspension, the tendency to settle being counter-
balanced by eddies. Larger particles, which settle at intervals
and are then swirled up again, skip along in a series of jumps.
Pebbles and boulders roll or slide along the bottom, according
to their shapes. Very large blocks may move along on a layer
of cobbles which act like ball-bearings.
The transporting power of a stream rises very rapidly as
the velocity increases. Experiments show that with debris of
mixed shapes and sizes the load that can be carried by running
water is proportional to something between the third and
fourth power of the velocity. But for fragments of a given
shape, the largest size that can be moved is proportional to the
sixth power of the velocity. Very large boulders which may
remain stationary in the stream bed for long periods can thus
be carried downstream by intermittent storm waters.
If the supply of debris exceeds the load that can be trans-
ported, or if the velocity is checked, part of the material is left
151
behind or deposited on the river bed, to be picked up later

when the stream is running more vigorously. Each time, the
largest ingredients of the load are the first to be dropped and
the finest are the first to be moved on again. In consequence,
a river begins to sort out its burden as soon as it receives it.
From source to mouth the deposited materials gradually
change in type from coarse to fine.
In addition to their solid load, rivers transport a great deal
of material in solution, most of which is contributed to them
by surface and underground drainage waters. The proportions
of dissolved and solid load vary enormously from place to place
and from time to time. Data referring to the work of erosion
and transport achieved by all the chief rivers of the world over
representative periods of years show that on an average about
8,000 million tons of rock-waste are removed from the lands
and transferred to the sea every year. Of this total about
30 per cent, is carried in solution. The drainage areas are at
present being worn down at an average rate of one foot in
about 9,000 years, but the average rates for individual regions
range from a foot in 400 years for the Irrawaddy basin to a
foot in 47,000 years for the low-lying basins draining into
Hudson Bay.
Lengthening and Deepening of Valleys
In the upper part of a typical stream gradients are steep,

the water runs swiftly and the narrow valley is either a gorge
or V-shaped, and the walls are often rocky. This is the torrent
or mountain tract. Lower down, in the middle part of the stream
— —
the valley tract slopes are gentler and the valley has become
much wider. Nearer the sea, in the plain tract, the valley
includes a broad flood-plain, which is liable to deposition
whenever the river overflows its banks. Seawards, the river
may flow into an estuary, or the plain tract may grow outwards
as a delta.
Not all rivers, however, have yet had time to develop a
full sequence of tracts. Some streams pass directly from the
valley tract into the sea ;
others, much younger in develop-
>52
<t. Abialhim, Ltd., Kcsuick
Bi Alknial Hats maiking tin- site of a former lake. Borrowdalc. looking
north from Claramara. Deruentuater and Skiddau in
the background'
GORGES
ment, may be entirely in the torrent stage. Such condi-

still
tions are also brought about when submergence of the land

takes place, and the lower reaches of the rivers already de-
veloped are “ drowned.”
In general, young valleys are lengthened and grow back-
wards into the land by headward erosion, due to rain-wash,
gullying, and the creep and slump of surface materials at
their heads. The torrent tract thus extends inland. At the
same time the valley is deepened by the active co-operation of
all the processes of river erosion. When the valley floor is cut
down so rapidly that there has been insufficient time for any
appreciable widening of the sides, a vertical chasm with pre-
cipitous walls results. The well-known Liitschine and Via Mala
gorges of Switzerland are familiar examples (Plate 36). The
cutting of gorges is favoured in areas well above sea level
where the rocks are highly resistant to weathering and the
widening processes (page 160) act slowly. Waterfalls often
have gorges in front of them, formed during the cutting back
of the rock face over which the water plunges. The more
spectacular gorges, like those of the Himalayas and Andes, are
cut by extremely active rivers which have continued to saw
downwards through the rocks during the actual uplift of the
mountain ranges.
Where widening and deepening proceed together, as
happens more commonly (and in any case as the widening
processes catch up), V-shaped valleys are developed, and the
V gradually opens out as time goes on. The torrent tract thus
evolves into the valley tract, and each gradually migrates
inland as the source continues to recede by headward erosion.
Grading of Rivers
Since a river which flows into the sea must have a gradient
towards the sea, the deepening of a valley is necessarily limited
by sea level. An imaginary extension of sea level under the
land is called the base-level of river erosion. The profile of a
river along its length from mouth to source is therefore a line
153
which is tangential to sea level and rises inland. In youth the

profile is more or less irregular, but as maturity is attained the
major irregularities are smoothed out. It is easy to see how the
initial irregularities aredestined to disappear. At each point
of the profile there must be a certain minimum gradient which
will give the stream just sufficient velocity to carry off the load
which it acquires. Wherever the gradient is steeper than the
minimum there necessary, the increased velocity so brought
about promotes erosion which wears the gradient down.
Wherever the gradient is less steep, the velocity is checked,
and the resulting deposition of part of the load builds the
gradient up.
When the profile is developed so that it everywhere pro-
vides the necessary minimum gradient, it is called a graded
Fig. 65
To show the relation between base-level and grade
{After R. S. Tan and 0. D. von Engeln)
profile or a profde of equilibrium.

Since this gradient varies with
the stream-flow, being least when
the stream-flow is greatest,
it follows that it is steepest towards the source. Under ideal
conditions it would theoretically have the shape of a hyper-
bolic curve, concave upwards (Fig. 65). The ideal curve is
never quite attained, however, because it implies a delicate
balance between gradient on the one hand, and stream flow,
transporting power, and load to be carried, on the other ;
a balance which is never maintained for long. Variations in
the supply of water and rock-waste from time to time, and
where tributaries come in, inevitably involve slight
especially
and temporary fluctuations. A river or any of its reaches in
which the profile of equilibrium is thus approximately estab-
lished is itself said to be graded or at grade.
Downward erosion does not cease when a river is graded,
though it may then become very slow. For a given stream-
!
54
ELIMINATION OF LAKES
flow, the necessary gradient decreases as the load decreases.

And as the whole relief of the drainage area continues to be
reduced, the load itself systematically decreases as time goes
on, while the stream-flow remains about the same. Conse-
quently the graded profile can be, and is, slowly flattened out.
Base-level, always being approached but never quite attained,
is its only limit.
The level of the main river at the point where a tributary

enters acts as a local base-level for that tributary. In the normal
development of a river system graded tributaries thus become
so adjusted to the main stream that they join it tangentially.
When tributaries fail to behave in this way the absence of
adjustment is a clear indication that the cycle of erosion has
been interrupted by changes of slope due, as a rule, to earth
Fig. 66
To elimination of a lake by sedimentation at the inlets and head-
illustrate the
ward erosion at the outlet. Successive positions of graded profiles are shown
before and after elimination
movements or glaciation. The alluvial fans and cones de-

scribedon page 201, and the waterfalls of recently glaciated
countries (page 221), are conspicuous features resulting from
such interruptions.
Various irregularities in a river channel may postpone the
general establishment of grade, though above and below these
features individual reaches of the river may be temporarily
graded. A lake, for example, acts as a local base-level for the
streams discharging into it. Lakes that occupy deep depres-
sions have a very long life, but shallow ones are, geologically
speaking, soon eliminated. A lake is a trap for sediments,
destined to be silted up by deltaic outgrowths from the inlets.
At the same time down-cutting of the outlet lowers the level,
drains the lake, and reduces its area (Fig. 66). Ultimately the
lake is replaced by a broad lacustrine flat through which the
155
river flows (Plate 31).Down-cutting through the sediments and

underlying rock-floor then proceeds until continuity of grade
is established between the upper and lower reaches of the river.
In the Lake District (page 183) lakes can be seen in every stage
of elimination, together with lacustrine flats in which young
valleys are already being developed (Plate 37).
A resistant formation encountered by a river also retards
the establishment of grade and acts as a temporary base-level
for the stream above until it is cut through by waterfalls and
rapids. The latter persist so long as the outcrop of obstructive
rock remains out of grade with the graded reaches in the softer
rocks exposed above and below.
Waterfalls
Where an outcrop of resistant rock is followed downstream
by a weaker formation, the latter is relatively quickly worn
down. At the junction, and subsequently above it, the river-
bed is steepened and the stream rushes down the slope as a
Successive stages in the recession and elimination of a waterfall

I Profile of stream 'drawn as graded) above an early position of the falls
II Present profile above falls
III Future profile after degeneration of the falls into rapids
IV Future profile (graded throughout! after elimination of falls and rapids
rapid. If the face of the resistant rock becomes vertical, the

stream plunges over the crest as a waterfall. The processes
which bring about recession and gorge development have
already been described. \V aterfalls eventually degenerate into
rapids as the profile begins to approach grade (Fig. 67). A
fall that descends in a series of leaps is sometimes referred to
as a cascade. An exceptional volume of water is implied by the

term cataract, which may be applied either to waterfalls or,
156
NIAGARA FALLS
more commonly, to steep rapids. Rapids are favoured

throughout the wearing down of an obstructive formation if
its dip is steep. And where the dip is downstream, even if it
be moderate, the development of a vertical face is precluded,
and rapids are formed instead of a waterfall.
Where a bed of strong rock, horizontal or gently inclined,
Lockport Dolomite
M >n
t>sV’
Rochester Shale
Clinton Limestone
<A %
and Shale
~1
Thorold ^
Sandstone]
Mian Sandstone
and Shale ijlsMs
f)
//!
\f//
Whirlpool Sandstone
Queenston Shale
Fig. 68
Section across the Niagara Falls showing the sequence of hard and soft formations
and illustrating the mechanism of recession
overlies weaker beds the former is the “ fall-maker,” and

scouring of the softer beds underneath leads to undermining
and recession. At High Force in Teesdale the tough and
well-jointed Whin Sill is the fall-maker. In the Yorkshire dales
falls are commonly developed over ledges of limestone under-
lain by shale.
The Niagara Falls are the classic example of this type
(Plate 38). As shown in Fig. 68 the river plunges 160-170
157
feet (according to the depth of the river below) over a thick

ledge of limestone. The American Falls (frontage, 1,060 feet)
are separated from the Canadian or Horseshoe Falls (frontage,
2,800 feet) by Goat Island. The mean flow, before the
diversion of much
of the water to hydro-electric plants, was
eighty-five times that of the Thames. Most of the water passes
over the Horseshoe Falls which, as their name implies, are
receding much more rapidly (3-4 feet a year) than the
American Falls (a few inches only). After the withdrawal of
the Labrador ice sheet from this region about 20,000 years ago
(page 243), the Niagara River followed a course towards Lake
Ontario, which led it over a pre-glacial escarpment. The fall
thus initiated has since receded 7 miles, leaving a gorge of
which the rim is about 200 feet above the river, and on the
average about 360 feet above the river floor. Ultimately the
Falls will cut back until Lake Erie is partly drained.
The Kaieteur Falls on the Potaro River in British Guiana
are also of this type, though in this case the ledge, over
which the river makes a sheer leap of 740 feet, consists of
hard conglomerate, and is underlain by softer sandstones
and shales.
The Yellowstone Falls (Upper, 109 feet and Lower, 308
;
feet) are cutting through an immensely thick mass of rhyolite,

parts of which have been altered and weakened, and at the same
time gorgeously coloured, by chemical changes due to thermal
waters. Hot springs still emerge along the floor of the canyon,
which in places is 2,500 feet deep. The fresh resistant layers
of rhyolite are the fall-makers. Rugged spurs and pinnacles
are left as canyon are worn back, all vividly
the walls of the
splashed with colours of every hue (Plate 39a).
Uplifted areas of plateau basalts of varying resistance
provide the structural background for some of the world’s
greatest falls. In the British Isles the most attractive examples
of this kind are those of Glenariff in Antrim. At the Victoria
Falls (Plate 39b) the Zambesi drops 360
feet from a nearly
level basaltic plateau into a gorge
60 miles long, through
which it rushes as a powerful torrent with surging rapids at
intervals. The extraordinarily acute swerves of the upper
158
GREAT WATERFALLS
part of the gorge well illustrate the dependence of form on
structure. Shatter zones which lie athwart the general course
of the river, and furnish easily removable masses of basalt,
have controlled the zigzag course picked out by the falls and
followed during their recession.
The Guayra Falls of the Parana River, where it becomes the
boundary between Brazil and Paraguay, have developed in

the Parana basalts. Here a narrow gorge has been cut deeply
back into the broad river floor, so that, besides the fall at the
head, there are twelve lateral falls on one side of the gorge
and five on the other. Taken as a whole this three-mile pano-
rama of falls is the greatest in the world. Although the height
is only 130 feet, the mean flow of water is more than twice that
of Niagara, rising to six times greater when the river is in

flood. Tributaries entering the Parana below Guayra also
descend in falls, the chief example being the Iguazu Falls on the
Brazil-Argentine boundary. These are twice as high as the
Guayra Falls, but, except in flood, the volume is much less
than that of Niagara.
Where rivers pass from uplifted areas of metamorphic and
massive igneous rocks to a plain of weakly resistant formations,
waterfalls are initiated and often gain height as they recede.
The rivers flowing from the hard rocks of the Appalachian
uplands have thus developed falls and rapids along the “ fall-
zone ” before they reach the softer sediments of the Atlantic
coastal plain. More spectacular examples of this type are the
Paulo Affonso Falls of the Rio Francisco in the Pre-Cambrian
crystalline rocks of N.E. Brazil, with a drop of 270 feet at the
head of a canyon 42 miles long the Aughrabies Falls (460 feet),
;
where the Orange River passes into a grim and desolate gorge
of naked granite and gneiss (Fig. 69) ;
the Grand Falls oj
Labrador, remarkable for a steeply slanting crest which gives
the River Hamilton a high velocity even before it begins its leap
of 300 feet and the Gersoppa Falls in the Western Ghats of
;
N.W. Mysore, which have a sheer drop of 830 feet. In the

monsoon floods the last of these has some claim to be considered
the greatest single fall in the world, for it then combines great
height with exceptional volume. In the dry season, however,
i59
[South African Air Force

Fig. 69
Aerial view of the Aughrabies Falls and Gorge, Orange River, South Africa
it dwindles to a trifling flow. There are many falls of greater

height, but these are all due to small streams falling over
precipices already provided for them (see Plate 53).
Widening of Valleys
Widening of valleys by the wearing back of the sides is

accomplished by a great variety of processes. These include
the scouring and steepening of the channel sides by the river
itself ; rain-wash and gullying ; soil-creep, slumping, land-
slides, and avalanches ; chemical weathering and leaching by
ground-waters removal of loose material by wind and the
; ;
general co-operation of incoming tributaries, which widen the

160
WIDENING OF VALLEYS
main valley where they enter it. The V-shaped cross-profile,
which is characteristic of valleys widened in homogeneous
rocks, opens out more quickly in soft rocks than in hard, and
this effect controls the form of the profile at all levels. Thus,
in a valley carved through a series of alternating hard and soft
beds the sides rise by slopes which are steep and precipitous
where the edge of a hard band outcrops terraced, across the
;
exposed top of such a band and of intermediate gradient

;
where the rocks are less resistant. The Grand Canyon of

—
Colorado clearly illustrates at a very youthful stage this —
dependence of profile on structure (Fig. 70). At the other
extreme, valleys in broad structural downwarps at a low level
Section across the south wall of the Grand Canyon of the Colorado at Grand
Canyon Station, Arizona. Scale (horizontal and vertical) 1 inch — 1 mile.
(After Darton) See also Plate 1 and Fig. 10,3
are wide and have all the superficial characteristics of old age
from the start.
Young streams are rarely straight for any distance, but
tend to follow a winding course (Fig. 71), determined in the
first instance by variations in the rocks encountered and in
the structures of those rocks. Several effects of fundamental

importance are brought about as water flows round a curve.
The main current of the stream (AB in Fig. 72) is deflected
towards the outer bank well beyond the beginning of the curve.
At the same time the centrifugal force acting on the water
concentrates it towards the outer bank where it heaps up. A
return current along the bottom (a, b, c, d, in Fig. 72). directed
towards the inner bank, is thus set up. The stream flows, in
fact, with a screw-like motion. These conditions obviously
involve maximum erosion on the outside of the curve and
161
[H.M. Geol. Survey

Fig. 71
Youthful valley with overlapping spurs. Crossdale Beck,

Ennerdale, Cumberland
minimum erosion or deposition on the inside (Fig. 73). The

resulting changes in the river channel and valley may be
summarised in order, as follows (Fig. 74) :
1. The channel is deepened on the outer side of each bend.

2. The outer bank is worn back and undercut by lateral
erosion, and the slopes above are consequently steepened,
locally into river cliffs.
3. As each bend is thus widened and deepened laterally,
the river shifts towards the undercut slope, and a tapering

spur, sometimes called the “ slip-off slope ” is left on the
162
RIVER BENDS
Fig. 72
To illustrate the flow of a river round a bend (in plan)
opposite side, often with a shingle- or sand-bank at its foot,

deposited there by the bottom current (Plate 40a ; see also
Plate 41a). The valley thus becomes highly asymmetrical
Fig. 73
To illustrate the flow of water round a bend (in section) and the resulting
lateral corrasion and deposition
in cross-profile. In plan, the interlocking spurs alternate with

the undercut slopes.
4. Each bend is enlarged downstream as well as laterally,
and thus tends to migrate downstream as a whole. As each of
the migrating bends of a sinuous stream reaches the spur next
163
Widening of a valley flow by lateral erosion and downward migration of bends.

The swinging stream eventually trims each overlapping spur into a bluff,
leaving a “slip-off” slope on the other side
below, the latter in turn is undercut and trimmed off, while

deposition now begins at the foot of the temporarily abandoned
undercut slope on the other side of the valley floor. Each time
a bend swings back to a shingle bank the river is at a lower
level,and any portion of the bank that is not then removed
remains as a terrace on the valley side. Such features are, ol
course, only short-lived, because as the valley is widened they
inevitably disappear.
5. By the repetition of this widening process, as bend after
bend migrates downstream, the spurs are all gradually cut
away, and a trough-like valley with a nearly flat floor bounded
by bluffs is developed (Plate 30). The shingle banks are now
enlarged into broad embayments between successive bends.
The beginnings of a flood-plain are thus established. During
the continued development of this stage the river becomes
graded and further deepening is then extremely slow.
Meanders
As the river continues toswing from side to side, it under-
cuts the bluffs wherever a bend impinges upon them, and so
the widening of the valley floor proceeds, while the slopes above
are slowly wasting away. The channel is now entirely in river
164
i .
E.X.A
(B) Meanders of the Rio Grande, separating; Mexico behind") from United
States territorj :in front'. Note ox-bow lake bottom left
t,
MEANDERS
deposits, bedrock being exposed only in and below the bluffs.
Each part of the material deposited on the growing flood-plain
is worked over in turn during the downward
sweep of the bends,
fresh additions from above constantly making good the losses
by erosion and transport. The bends, now free to develop in
any direction, except where they encounter the valley side, are
more quickly modified. The stream is likely to be sluggish
and easily turned aside by obstructions in the channel floor
and by the inflow from tributaries. Freely developing bends
are called meanders from their prevalence in the River Meander
in Asia Minor.
As meanders of short radius are enlarged more rapidly
Fig. 75
Successive stages in the development of meanders, showing
the formation of an
ox-bow lake by the “ cut-off ” of a loop
than bigger ones, the curves all tend to have radii of about the
same order. But as the curves develop by undercutting and
deposition (Fig. 75), the meanders swell into loops which
gradually approach until they coalesce. If a flood occurs
when only a narrow neck of land is left between adjoining
loops, the momentum of the increased flow is likely to carry
the stream across the neck and thus short-circuit its course.
On the side of the“ cut-off” a deserted channel is left, forming
an ox-bow lake which soon degenerates into a swamp as it is
silted up by later floods (Plate 41b). By making artificial
cut-offs (1927—37) the Mississippi River Commission has short-
ened a 331-mile stretch of the riverby 116 miles. The natural
short-circuiting process sets a limit to the growing radii of
( 3®e)
165 I2
meanders. Where the channel of the Mississippi is about half

a mile wide, the meanders commonly have a radius of 4 or 5
miles. Narrower streams develop meanders of proportionately
smaller radius (cf. Plate 85a).
As a on growth, a meandering river
result of this restriction
flows in a meander-belt (Fig. 76), which is usually about 15 to
18 times the width of the river. The meanders themselves
swing down the belt with a snake-like motion. Relics of old
Meander-belt of the river Forth. Horizontal shading indicates land

above 400 feet
ox-bows, indicating the positions of former meanders, can be

clearly seen from the air, although they may no longer be
obvious on the ground. This is because the vegetation (re-
flecting the differences in soil and drainage) of an infilled ox-
bow lake differs from that of the normal alluvium.
The valley reaches full maturity when its width is about
the same meander-belt appropriate to the width

as that of the
of the river. As the bordering bluffs of the valley continue to
be cut back where individual meanders impinge against them,
the valley floor slowly expands until, after an immensely long
period, it may attain a width several times that of the meander-
belt. The latter then itself swings to and fro across the wide
flood-plain, and old age has been reached. An interesting
indication of the swing of a meander-belt was provided by the
1 66
FLOOD PLAINS
discovery of the ruins of the ancient city of Ur of the Chaldees
— —
the home of Abraham in Mesopotamia. Five thousand
years ago the Euphrates was 5 miles to the west of Ur ; now
it is 5 miles to the east of the ruins.
Flood Plains
If a stream is heavily laden with coarse debris from nearby

mountains, its minimum gradient and rate of flow are neces-
sarily greater than for one carrying finer material. As floods
subside, bars and islands of shingle are left on the channel floor
and the stream is obliged to divide into a network of many
channels. Such a stream is said to be braided (Plate 40b).
Fig. 77
Schematic section across the flood plain of a stream bordered by natural levees
Thoroughly clogged rivers of this kind, much of the water of

which flows below the surface through the interstices of the
shingle, do not make meanders.
In the early stages of flood-plain development by streams
less heavily charged with coarse debris, shingle-banks are
deposited on the inner sides of bends, and these may gradually
expand to form a deposit of river gravels over a wide area
(Plate 40a). Later, however, when meanders are developing,
the stream carries only sand, silt, and mud, and these are
spread by floods as a veneer of alluvium over the coarser
deposits of an earlier stage. Each time the stream overflows
its banks the current is checked at the margin of the channel,
and the coarsest part of the load is dropped there. Thus, a low
embankment or levee is built up on each side (Fig. 77). Beyond
the levees the ground slopes down, and in consequence is
liable to be marshy. During floods, levees may grow across
the junctions of small tributaries. The latter are then obliged
to follow a meandering course of their own, often for many
167
miles, before they find a new entrance into the main river.
Depressions occupied on the way become swampy. The
characteristic features of a flood-plain thus include meanders,
ox-bow lakes and marshes, levees, bordering swamps, and a
complicated pattern of lateral streams.
Levees afford protection from ordinary floods, but the
river then begins to silt up its confined channel with material
that would otherwise have been spread as alluvium over the
plain. Its level is raised, and the levees grow up with it, so
that the danger from major floods becomes greater than before.
To obtain increased protection artificial levees are often built,
but these provide only temporary security, since they accentuate
the tendency of the river floor to rise. In the flood-plains of the
Po in Italy and of the Hwang Ho and Yangtze Kiang in China
the built-up levees are locally higher than the neighbouring
house-tops, and the rivers flow at a level well above that of the
adjoining land. Such conditions are obviously extremely
dangerous, as a severe flood may break through the levee and
bring disaster to the agricultural lands over an enormous area.
Along the Mississippi and its tributaries the flood danger is
a serious menace. Little more than a century ago floods were
easily controlled by levees about four feet high. The levees
have since had to be raised several times. By 1927 they were
three or four times as high, but nevertheless a great flood then
broke through and devastated 25,000 square miles. Stronger
levees up to 20 or 30 feet high have now been built, but it is
clear that this method of flood control is far from satisfactory
by itself. It can, however, be supplemented by reforestation
of upstream regions (to reduce the rate of run-off), by the
straightening and dredging of river channels, and by the
allocation of certain areas as storage reservoirs for flood water.
The many others, result from heavy
Mississippi floods, like
rainfall in the early spring,supplemented by the melting of the
winter’s snow. If abnormal snowfall is followed by a wide-
spread sudden thaw, so that all the tributaries rise simul-
taneously (instead of successively, as usually happens) a serious
flood is inevitable. Disastrous flooding of the Euphrates
occurred in 1929 as a result of the sudden melting of snow in
1 68
DELTAS
the Anatolean Mountains, a thousand miles from Baghdad.

The Thames floods of the same year were due to abnormal
rainfall supplemented by the effects of spring tides and strong
gales blowing upstream. In rivers like the Indus great floods
may also be caused by the release of vast quantities of water
ponded back by great landslides which sooner or later collapse.
Deltas
When a river reaches the sea much of its load may be

quickly deposited, partly because the current is checked and
partly because the salt water coagulates the fine particles.
The sediment thus settles more rapidly than in fresh water.
Waves and currents, however, may be sufficiently strong to
sweep away the material and so prevent the mouth from being
silted up. If, in addition, the land and sea floor are subsiding,
Fio. 78
Section through the sediments of a delta. T, topset beds ;
F, foreset beds ;
B, bottomset beds
or have recently done so, the valley will be partly submerged,

and the river will pass into the sea by way of a tidal estuary.
On the other hand, if tides and currents are weak, as in
enclosed seas like the Mediterranean and the Black Sea, de-
position takes place at the mouth of the river on the sides and
floor of the channel, and in front of it, so that a broad, outward-
sloping fan of sediment is gradually built up on the sea floor.
The front of each part of the fan grows seawards, just as a
railway embankment is built forward during its construction,
and the flood plain gradually extends seawards across its flat
top (Fig. 78). While this is going on, and at a very early
stage if the sea is shallow, the channel of the river becomes so
choked and obstructed that it can no longer carry all the river
water. The swollen river therefore breaks through or across
169
itsbanks, and so acquires two or more exits to the sea. This

process of current bifurcation is repeated again and again, until
a system of branching channels, called distributaries, is formed.
The resulting seaward-growing terrace of sediment, traversed
by is a delta.
distributaries,
The
Delta of the Nile was the first to be so named, because
of the resemblance of its shape to the Greek letter A. It is an
example of the arcuate type, with a rounded outer edge, modified
in this case by fringing sand-spits shaped by sea currents
Fig. 79
Map of the Hwang Ho and its delta, showing the distribution of loess (dotted)
and alluvium (horizontal shading) derived from loess ( After G. B. Cressey)
(Fig. 135 ). After traversing a thousand miles of desert, the

Nile has comparatively little water left when it reaches the
Delta, and much of the load is deposited near the apex.
Frontal growth is therefore slow. The Po delta extends more
rapidly. Adria, now 14 miles inland, was a seaport 1,800
years ago, the average rate of advance thus indicated being
about 40 feet a year. Ostia, the seaport of ancient Rome, is
now four miles from the mouth of the Tiber. The richly fertile
delta of the Hwang Ho in North China has grown across what
was broad bay of the i ellow Sea.
originally a A
large island,
now the Shantung Peninsula, has been half surrounded. Since
170
THE MISSISSIPPI DELTA
1852 the main branch of the river has emptied to the north of
Shantung, but before then it flowed across the southern section
of the delta and reached the sea about 300 miles from its
present mouth. In that year and again in 1887 there were
calamitous floods in which the loss of life from drowning and
famine amounted to many hundreds of thousands (Fig. 79).
Most deltas, like those mentioned above, are arcuate. Their
distributaries may be braided if the sediment deposited is
coarse. If the sediment is finer, meandering courses tend to
develop. The Mississippi, however, is exceptional in that
Fig. 80
The Mississippi delta : fifty years’ growth
it extends mouths seawards by way of deep channels,

its
locally called “ passes,” which are outstretched like fingers.

This part of its delta is the chief example of the bird’sfoot type
(Fig. 80). The sediment brought down by the Mississippi
contains an unusually high proportion of fine mud, and
deposition is therefore mainly on the levee-like sides of the
channels, thus confining them within impervious banks of
clay. One such mouth is being extended 250 feet into the
Gulf of Mexico every year.
The Mississippi delta is slowly subsiding. As deposition
has kept pace with the sinking, thick accumulations of deltaic
sediments have been built up. Borings prove a thickness up
to at least 2,000 feet, and the low values of gravity correspond- —
171
.
ing to the low density of the underlying sediment suggest —

that the actual thickness may be immensely greater. Similar
accumulations of sediment, made possible by subsidence, are
characteristic of most large deltas, including those of the Nile
and Ganges-Brahmaputra.
Lakes, in which currents are negligible, provide highly
favourable conditions for delta growth. When, in addition,
their waters are salt, the rate of deposition from incoming
rivers reaches its maximum. The Terek delta is at present
growing outwards into the Caspian at the rate of 1,000 feet a
year.

G. K. Gilbert
Report on the Geology of the Henry Mountains (
Utah) U.S. Geographical
and Geology Survey of the Rocky Mountain Region, Washing-
ton, 1877.
C. F. S. Sharpe
Landslides and Related Phenomena. Columbia University Press, 1938.
G. C. Forrester
The Falls of Niagara. Van Nostrand, New York, 1928.
E. C. Rashleigh
Among the Waterfalls of the World. Jarrolds, London, 1935.
M. A. C. Hinton
Rivers and Lakes. Sheldon Press, London, 1924.
See also the list of works following Chapter XI.
172
Chapter XI
DEVELOPMENT OF RIVER SYSTEMS AND

ASSOCIATED LANDFORMS
Tributaries and Drainage Patterns
A consequent stream is one whose original course is deter-

mined by the initial slopes of a new land surface. From a
volcanic cone or an uplifted dome the first streams flow off
radially. A long upwarp or geanticline provides a linear crest
— —
the primary watershed or divide with slopes on each side.
In many cases the uplifted area consists of a coastal plain
backed by an older land already drained by rivers. These
continue down the new surface as extended consequents. If there
is no “ old land,” the consequents begin some way below the
crest, at each point where the drainage from above just suffices
to initiate and maintain a stream. As the valley head widens
and increased drainage is secured, each such stream is pro-
gressively lengthened by headward erosion. If uplift continues,
the consequent streams are correspondingly lengthened by
seaward extension.
As the consequents dig in, the valley sides furnish secondary
slopes down which tributaries can flow. The tributaries
lengthen by headward erosion, which picks out the least re-
sistant parts of the rocks encountered, such as jointed or
fractured belts, or beds of clay or shale. Subject to a general
tendency to flow at right angles to the contours of the
consequent valley, the pattern formed by tributaries and
consequents thus depends largely on the nature and structure
of the rocks which are being dissected. The latter may be
homogeneous through a considerable depth, or they may
consist of a stratified series of alternating strong and weak
beds.
Where the rocks have no conspicuous grain and offer
nearly uniform resistance to erosion, the headward growth of
i73
DEVELOPMENT OF RIVER SYSTEMS AND ASSOCIATED LANDFORMS
[South African Air Force

Fig. 81
Aerial view of the dendritic drainage developed by insequent streams on massive

granite. Tributaries of the Orange River near the Aughrabies Falls (see Fig. 69)
a tributary (beyond the slopes of the consequent valley) is

governed primarily by the initial regional slope, with modifica-
tions controlled by haphazard irregularities of surface and
structure. Because of these accidental controls such streams
are said to be insequent. The regional slope, however, generally
determines the prevalent direction followed by an insequent
tributary ;
it commonly makes an acute angle with the up-
stream part of the consequent valley. As each insequent

stream develops its own valley, it receives in turn a second
generation of tributaries. The branching drainage pattern so
established is tree-like in plan, and is described as dendritic
(Fig. 81 ). If the rocks are well jointed, however, a more
rectangular pattern is likely to result.
Where, as commonly happens, the rocks consist of belts of
alternately weak and strong beds, dipping seawards, the con-
sequent valley is narrow and steep-sided where it cuts through
resistant beds (sandstones, limestones, lavas, or sills), and broad
174
RECESSION OF WATERSHEDS
and open where it crosses outcrops of weak beds (clay or shale).

A tributary beginning on weak rocks has a great initial
advantage. Headward erosion guides it back along the weak

bed, parallel to the strike. Such a tributary is called a sub-
sequent stream. The rectangular drainage pattern formed by
consequent streams parallel to the dip and subsequent streams
parallel to the strike is described as trellised (cf. Fig. 85). Later
tributaries add further detail to the trellised pattern.
Shifting of Divides and River Capture
The position of a divide remains permanent only if the rates

of erosion are the same on each side, a state of affairs that is
practically never achieved. It usually happens that the
opposing slopes are unequally inclined, and that erosion is
more effective on the steeper side. Consequently the divide
is gradually pushed back towards the side with the gentler
slope (Fig. 82). In primary divides this effect is most rapidly
Section to show the recession of a watershed from A to B as a result of the

effect of unequal slopes on erosion
produced by headward erosion of the consequent valley-

heads. As the latter work back through the crest, some from
one side and some from the other, the divide becomes zigzag
or sinuous, while the crest is notched and becomes increasingly
uneven (Fig. 83). As deepening proceeds, the dissection of the
ridge is steadily elaborated and the more resistant rocks
between the valley heads stand out as peaks. Where the
headward migration of one valley head encroaches on a valley
head at the other side, the notch in the crest develops into a
col or pass.
*75
Fig. 83
Development of a zig-zag watershed by headward erosion
The new drainage area thus acquired by a headward

growing consequent stream is generally of little importance.
Migration of the secondary divides between neighbouring
consequent valleys leads to far more revolutionary changes.
Diagram to show impending river capture. The subsequent stream S x is

cutting back at a relatively low level towards the consequent stream Cs
One of the original consequent rivers is likely to have a bigger

—
drainage area than its neighbours either because it flows
through an undulation in the initial surface or because it is an
extension of an earlier river in the “ old land ” behind the
coastal plain —and the valley of this major stream is therefore
176
RIVER CAPTURE
deepened and widened more quickly than the neighbouring

consequent valleys.If the lateral divides are pushed back
Trellised drainage (consequent and subsequent streams), showing the dissection

of a gently dipping series of hard and soft beds into escarpments and inner
lowlands
until they reach these minor streams, the latter and their
drainage areas are absorbed by the major river.
Capture of drainage on a still bigger scale becomes possible
when the major river acquires vigorous subsequent tributaries,
each working along a feebly resistant formation and each
Later development of the river systems of Fig. 85, illustrating river capture by
the headward growth of the more vigorous subsequent streams
pushing back the secondary divide at its head (Figs. 84 and 85 ).

Endowed with a relatively low local base-level, a deeply
entrenched subsequent, e.g. S x in Fig. 86, cuts back towards
a consequent C 2 which is still draining an area at a higher
,
level. Eventually C 2 is intercepted, its headwaters are diverted

into S 1 , and its lower course is beheaded. This process is
177
Fig. 87
River systems of North-east England, to illustrate river capture by the Ouse

(along the outcrop of the soft Triassic beds dotted) and by the North Tyne
—
(along the outcrop of the Scremerston Coal Group inclined shading)
called The rectangular bend e at the point of

river capture.
diversion known as the elbow of capture. The beheaded river,

is
now deprived of much of its drainage, is described as a misfit,

sinceits diminished size is no longer appropriate to the valley
through which it flows. Its new source is some way below the
elbow of capture and the deserted notch, Wg, at the head of its
178
EXAMPLES OF RIVER CAPTURE
Valley becomes a wind gap. Asubsequent stream S 2 which

,
originally entered the captured stream near or above the elbow,

now has its local base-level lowered to that of Sj. It is thus
enabled to deepen its valley and to extend backwards until it,
in turn, reaches and beheads the next consequent, C 3 A major

.
consequent river, with the aid of its subsequents, may therefore

acquire a very large drainage area at the expense of its neigh-
bours.
The rivers which flow into the estuary of the Humber
illustrate the development of an actual river system by the
process of capture outlined above. The uplift of the Pennines
provided the slopes down which a number of consequent
streams flowed into the North Sea. Of these only the Aire
still maintains an uninterrupted course. The Wharfe, Calder,
and Don were probably tributaries of the Aire from the start.
The Nidd, Ure, and Swale, however, have each been captured
in turn by the Ouse, a powerful subsequent stream which
worked back northwards along the soft strata of the Trias
(Fig. 87). On the eastern side of the Ouse it is difficult to
trace the former courses of the beheaded streams, because of
uplift of the Cleveland Hills and obliteration of the older
valleys by glacial deposits. A more diagrammatic example is
provided by the rivers of Northumberland. The three main
streams, a, b, and c, of the North Tyne system clearly correspond
to the Wansbeck, a', a tributary of the Wansbeck, b', and the
Blyth, c'. The headwaters of the forerunners of these were
captured by the North Tyne, a subsequent of the Tyne, as it
worked back along the soft beds of the Scremerston Coal series.
River capture by subsequents which developed along the
clay formation known as the Gault has greatly modified the
drainage of the Weald. Examples can easily be recognised in
a map of the region.
Escarpments and Related Features
The is widened and deepened

valley of a subsequent stream
between divides formed by the bands of resistant rock on either
side. As the weak bed is gradually worn away, the upper
179
surface of the underlying resistant bed is uncovered, and on
this side the valley slope therefore approximates to the dip.
On the other side the overlying resistant bed is exposed at its
base and, as a result of undercutting and the falling away of
joint blocks,it soon begins to stand out as a prominent ledge
which steepens into an escarpment, facing inland (Fig. 88). As

the subsequent stream, keeping pace with consequent, its
continues to deepen its valley in the weak rocks, its channel
gradually shifts in the direction of dip, that is, towards the
escarpment. The latter is thus steadily worn back, leaving a
gentle dip slope on the other side of the valley. The valley
itself, as it becomes wide and extensive, develops into an
interior lowland. Small tributaries, known as obsequent streams,
Stages in the development and recession of an escarpment

H= hard, resistant formations S = soft, easily eroded formations
descend the escarpment, others, called secondary consequents,

flow down the dip slope, and both sets add to the trellis pattern
of the drainage (cf. Fig. 85).
Fig. 48 illustrates the succession of escarpments and interior
lowlands between Gloucester and the London Basin. From
the Lias clays and marls of the Severn valley the escarpment of
the oolitic limestones of the Jurassic rises to the crest of the
Cotswolds. The Oxford Clay is responsible for the interior
lowland occupied by the Thames above and below Oxford.
A minor escarpment, that of the Corallian limestone, is then
followed by the interior lowland of the Kimmeridge Clay.
Beyond this the Chiltern Hills represent the escarpment of the
Chalk, the dip slopes of which lead down to the London basin
(Fig. 57). The Chalk escarpment curves round to the eastern
side of the YV ash, and continues to the north as the Lincolnshire
180
ESCARPMENTS AND MESAS
To show the relation of various erosional landforms to the structure and dip
of the strata from which they are carved
and Yorkshire Wolds. On the south side of the London basin

the Chalk again emerges as the great escarpment of the North
Downs which, less conspicuously, swings round the Weald
[Grant, V.S. Dept, of Interior

Fig. 90
Mesa showing marginal “ badland erosion where the protective cap has been
’’
removed. Zion National Park, Utah
to appear again on the far side as the South Downs. For illus-
trations of other escarpments see Plates 18 and 42a.
An escarpment and its dip slope together form a feature
for which there is no English name. The Spanish term cuesta
(386) 181 13
(pronounced questa ) has therefore been adopted. If the beds
dip at a high angle, the dip slope becomes as steep as the
escarpment and the feature corresponding to the cuesta is
simply a ridge or hogback (Plate 42b). At the other extreme,
in horizontal beds, the cuesta becomes a mesa (Spanish for
table), that is, a tableland capped by a resistant bed and
having steep sides all round. Table Mountain, behind Cape
Town, is a small but high mesa which has been developed
by erosion from a fault-block. By long continued wearing
back of the sides, a mesa dwindles into an isolated flat-topped
hill. In America such a hill is called a butte, from its resem-
blance to the butt or bole of a tree, and the term has been
widely adopted. In Western America buttes commonly occur
where the beds dipping off the mountain flanks flatten out
(Figs. 89 and 90). In South Africa, however, similar residual
landforms, many of which are capped by isolated relics
(outliers) of once continuous dolerite sills are called kopjes.
Superimposed Drainage
In many now exposed

regions ancient folded rocks are
which were formerly hidden beneath an unconformable cover
of later sedimentary formations. The rivers initiated on the
cover, with a drainage pattern appropriate to its structure,
ultimately cut their valleys into the underlying rocks, main-
taining their courses with little or no relation to the very
different structures inwhich they then find themselves. As the

cover is gradually removed by denudation the old rocks are
exposed over a steadily increasing area, the drainage pattern
of which has been superimposed on it as an inheritance from the
vanished cover.
The clearest example of superimposed drainage in Britain
is afforded by the rivers and lakes of the Lake District. As
and 92, the Lake District consists of an
illustrated in Figs. 91
oval-shaped area of Lower Palaeozoic rocks (folded during the
Caledonian orogenesis and having a general trend from E.N.E.
to W.S.W.) enclosed in a frame of Carboniferous Limestone
182
THE ENGLISH LAKE DISTRICT
Fig. 91
Geological sketch-map of the Lake District, showing the radial pattern of the
superimposed drainage
and New Red Sandstone, the beds of which everywhere dip

outwards. The younger formations originally covered the
older rocks. During Tertiary times the region was uplifted
into a slightly elongated dome, with its axis curving towards
the east from a culminating point which lay above the present
summit of Scafell. The consequent streams that flowed
radially down the slopes of the dome still persist in the older
rocks. The radial pattern of the valleys and mountainous
183
Fig. 92
Section across the Lake District
(1) Skiddaw Slates (2) Borrowdale Volcanic Series (3) Coniston Limestone
(4) Silurian (5) Carboniferous Limestone (6) New Red Sandstone
(Igneous intrusions omitted)
ridges centred near Scafell is particularly striking. As num-

bered on Fig. 91, the valleys are (1) the headwaters of Borrow-
dale and Derwentwater ; (2) Langdale and Windermere ;
(3) The Duddon Eskdale

; (5) Wasdale and Wastwater ;
(4) ;
(6) Ennerdale and (7) Buttermere and Crummock. To the

;
east, the streams leading into Coniston, Ullswater, and Hawes-

water flow oft the axis towards the north and another set of ;
streams, some feeding Windermere and others draining

directly into the Irish Sea, flow southwards. The lakes and
many of the other scenic features of the Lake District have
resulted from glaciation.
The Appalachian rivers of the United States are now inter-
preted as an example of superimposed drainage. Long after
the folding of its Palaeozoic rocks, the Appalachian belt was
worn down and buried beneath a cover of marine sediments.
Later the region was uplifted, consequent streams flowed down
the slopes of the cover into the Atlantic, and eventually cut
their valleys into the foundation rocks. The cover has now
vanished, and the Appalachians have been dissected into long
mountainous ridges, but the main rivers continue to cross the
latter through deep “ water gaps.” Examples of these great
notches cut through the apparent barriers are the Hudson
gorge in the Highlands of New York State the Delaware ;
\\ ater Gap (Fig. 93) farther south ; and the gap of the
Potomac at Harper’s Ferry, famous during the Civil War as
the strategic gateway into the broad interior valley of the
Shenandoah.
184
LANDSCAPE DEVELOPMENT
[t'.S. Geol. Survey

Fig. 93
Delaware water gap, cut by the Delaware River (Pennsylvania-New Jersey
boundary) through a tilted formation of resistant conglomerate
The Normal Cycle of Erosion
It has now been made abundantly clear that rivers, and

indeed alllandscape features, are continually changing.
From the time when the sculpturing of a newly uplifted land
area begins, the valleys and associated landforms pass through
a series of well characterised stages, referred to as youth,
maturity, and old age, until, ifsudden interruptions due to
earth movements do not intervene, the whole area is reduced
to a peneplain. The whole sequence of the changes involved

in this evolutionary development of landscapes is called the
normal cycle of erosion (Fig. 94), the term “normal” implying
that the development is controlled by river action and surface
erosion under humid conditions. The cycle concept perhaps —
—
the most fertile ever contributed to geomorphology was later
applied by W. M. Davis, its originator, to the distinctive land-
forms developed by erosion under glacial and arid conditions.
185
Fig. 04
Profiles across an initially uptilted block showing successive stages in the cycle
of erosion. 1-2, youth ;
3-4, maturity ; 5-6, old age
A river system may be still youthful in its upper tracts,

while the characteristics of maturity have already been reached
lower down and may, indeed, have passed into the monotonous
features of old age towards the sea. The completion of each
stage is first attained in the lower tract, and slowly creeps
inland. In the course of a full cycle rivers and landscapes
have a relatively short but vigorous youth, a much longer
period of mature development, and an extremely long old age
of decline and degeneration (Fig. 95). The actual time re-
quired for a cycle to run its course in any given area naturally
varies enormously with the height and structure of the uplifted
area, and with the size of the rivers and their drainage basins.
A million years or so may suffice to bring comparatively small
rivers, like those of Britain, well into the stage of old age, but
the great rivers of high Asia may still be far from completing
their prodigious task in ahundred million years.
The stage of youth ideally begins with the dissection of a
featureless plateau or an undulating folded region. It is
essentially the period during which the valley form is under-
going vigorous development, especially in depth and headward
growth. The early rivers flow swiftly and have irregular
gradients. Lakes, rapids and waterfalls, and gorges are highly
characteristic features. In recently folded regions the main
riversoccupy synclinal furrows. Tributary development pro-
ceeds rapidly during youth and river capture is common. The
streams compete for space until the victorious ones acquire
well defined valleys and drainage areas. Between the valleys
there are at first extensive tracts of the original surface, poorly
drained and often swampy. As the valleys widen, and the
divides are pushed back, the areas of these tracts gradually
1 86
NORMAL CYCLE OF EROSION
Fig. 95
Wearing down of land surface from youth (a) through maturity (b) to old age (c)
( After V. C. Finch and G. T. Trewartha)
diminish. The youth of the landscape merges into maturity

when the relief attains its maximum amplitude. The area is
then “ all slopes ” (stage 3 in Fig. 94), the last traces of the
initial surface disappearing as the summits of the divides begin
to be worn down. River systems reach the end of the youthful
stage when the main stream and its chief tributaries become
187
graded, that is, after lakes, waterfalls, and rapids have been
eliminated, and gorges have widened into V-shaped valleys.
Most lakes disappear during early youth, but exceptionally
deep ones, like the Great Lakes of North America, survive
much longer. It may happen, of course, that grading of a river
system is not fully established for some time after the initial
surface has vanished. The land surface will then have reached
maturity before the rivers have completed their youth. Gener-
ally, however, when rivers still obviously in their infancy
traverse a landscape which is long past its youth, the contrast
is due to uplift of an area already dissected by an earlier but
uncompleted cycle.
At the beginning of the stage of maturity the divides are at
their maximum height above the valley floors, and the valleys
have reached their maximum width, as measured from divide
to divide. Thereafter, the upland surfaces are slowly lowered,
but slow as this drawn-out process may be, it is faster than the
lowering of the graded valley floors. Subsidiary tributaries
still develop, and some rivers may continue to increase their
drainage basins by the capture of neighbouring streams.

As the topography gradually becomes more subdued, the
mountain peaks of the greater ranges and the hills of the sub-
sidiary ridges are rounded off into broadly sweeping curves,
the convex summits passing into concave and graded slopes
which lead down to the bordering bluffs and flats of the valley
floors. The latter are increasingly widened by lateral erosion ;
meanders are characteristic, and alluvial plains become ex-

tensive. In youth the flat surfaces of the landscape are those
of the initial surface (apart from the lacustrine flats of infilled
lakes), whereas in maturity the flat surfaces are the valley
floors. Near the sea the divides between the flood plains of
neighbouring rivers may be completely worn down to plains
of bedrock mantled with rock-waste. Here chemical weather-
ing becomes dominant and deep soils develop. The flood
plains gradually coalesce and extend inland at the expense of
the wasting divides. In this region old age has already begun.
There are no distinctive features in terms of which the
the close of maturity can be clearly defined, but old age is
1 88
PENEPLAINS
generally considered to have set in when the valley floors

reach a width several times that of the meander-belt. The
widening of the valley floors by lateral erosion, and the lowering
of the divides by chemical weathering and surface creep, con-
tinue as before, but ever more slowly as base-level is approached
over a wider and wider area. The process may become almost
infinitely slow, but so long as streams can carry a load, even
if it be mainly in solution, the reduction of the land towards
base-level proceeds indefinitely. The uplands of youth and
the maze of slopes of maturity are replaced in old age by wide-
spread lowlands, rising gently inland. The region has become
a peneplain. Over sedimentary rocks the relief is uniformly
faint. But in geologically complex orogenic belts the pene-
plain developed across their crystalline roots is rarely, if ever,
brought to such perfection. The more stubborn rocks, such
as the most resistant parts of granite stocks and batholiths,
continue to stand out as occasional residual hills. Such
erosional survivals are called monadnocks, after Mount Monad-
nock in New Hampshire, which is a typical example of its class.
In sedimentary regions penetrated by volcanic necks, the latter
may long persist as monadnocks.
For the perfect development of a peneplain it is essential
that the area should not be uplifted. Few regions of the present
lands have escaped recent earth movements, and present-day
peneplains of considerable extent are therefore rare. Western
Siberia and the country around Hudson Bay are noteworthy
examples of peneplains, but even these have had their surfaces
modified in detail by glaciation.
The Isostatic Response to Denudation
The reduction of a region to a peneplain involves the re-

moval of an immense load of material, the mass of which is
proportional to the height of the initial surface. While the
crust was being thus unloaded by denudation, slow isostatic
uplift must have been continuously in progress, thereby giving
the rivers more work to do and delaying its completion. This
189
effect has so farbeen tacitly ignored for the sake of simplicity
of treatment, but it should not be overlooked.
Let us suppose that a thickness of 1,000 feet of rock having
an average density of 2-6 has been removed from a region while
isostatic equilibrium is maintained and no independent earth
movement occurs. The mass lost is proportional to 2-6 X 1,000,
and must be made good by the inflow at depth of a thick-
this
ness h of material with a density of about 3-4. The condition
for the maintenance of isostasy is that 3-4A= 2,600 feet; whence
h= 765 feet. This influx of sima raises the plane AB (Fig. 96)
to A'B', and the new surface is only 235 feet below the original
level of the denuded block of country. For the development of
Fig. 96
Isostatic response to denudation : uplift of
plateau
Fig. 97
Isostatic response to denudation : uplift of
mountain peaks
a peneplain from an initial surface which stood 1,000 feet

higher, the thickness of rock to be removed is not 1,000 feet,
but about four times as much.
Such uplift must be considered a normal accompaniment
of a cycle of erosion. It involves the curious effect that during
late youth and early maturity the summits of peaks and divides
become elevated above the initial surface. This is illustrated by
Fig. 97. W
hen the cross-sectional areas of valleys and divides
are equal, half the mass of the denuded block has
been removed.
The plane CD will by then have been raised to C'D', i.e. by
765/2 feet, assuming the summits to be 1,000 feet
above the
valley floors. The bearing of this remarkable
result on the
high altitudes of the Himalayan peaks is referred
to on page 201.
190
SUMMIT LEVEL OF THE GRAMPIANS
'
"
^
Crown Copyright Reserved] [

H.M . Geol. Survey
Fio. 98
The sky-line of the Grampians (the Mamore Forest), looking S. from Ben Nevis.
An example of an uplifted peneplain, now deeply dissected
Uplifted Peneplains
The isostatic uplift that accompanies denudation is, of

course, not competent to uplift a peneplain once it is formed.
It has merely delayed peneplanation. The uplifts referred to
here and in later sections are those due to independent earth
movements.
Peneplains representing the practically completed cycles of
former periods, but since uplifted to form the initial surfaces
of later or present-day cycles, can be detected in many land-
scapes. In the Grampian Highlands an old peneplain, now
dissected into a landscape of late youth or early maturity
(though modified by glaciation), is easily recognised by the
even skyline corresponding to a widespread uniformity of
summit levels at about 2,000 feet (Fig. 98). The occasional
higher peaks which rise above this “ summit plane ” represent
191
the monadnocks of the old peneplain. Asimilarly dissected
uplifted peneplain is represented in Wales by a summit plane
at 1,500 to 2,000 feet. These old peneplains are distinguishable
from uplifted coastal plains by the fact that they truncate a
great variety of orogenic structures.
A marked tendency to rise at intervals has controlled the
geological history of the interior of Africa for many hundreds
of millions of years. Several cycles of erosion are recorded by
enormous thicknesses of continental sediments and by uplifted
peneplains now represented by plateaus surmounted by monad-
nocks of the inselberg type (page 274).
A peneplain approaching completion during the Miocene
wasuplifted, warped, and rift-faulted at the close of the Miocene,
and again at the end of the Tertiary, to form the great plateau
of Tanganyika and adjoining territories (page 431). It now
stands at a height of 3,000 to 6,000 feet, and inselbergs which
rise from its surface are the residual hills from an earlier pene-
plain uplifted during the Jurassic. The faulted borders and
some of the rift valley scarps are deeply notched by gorges
and canyons with waterfalls at their heads. The Kalambo
Falls, the most celebrated example, came into existence on
the eastern fault-scarp of the Tanganyika rift. They have now
cut back three miles from the lake, and there the placid stream
above the head of the gorge suddenly takes a single leap of
705 feet over the brink (Fig. 99).
Farther south the interior plateaus of the Basuto Highlands
and the High Veld behind Natal were formed by the uplift
of a Cretaceous peneplain during late Cretaceous and Tertiary
times. These movements were accompanied by depressions
of the coastal area. An enormous escarpment has since been
worn back into the plateau country, which reaches a height
of 12,000 feet in the basalt-capped Drakensberg of Natal
(Plate 87a). On the plateau side of the escarpment the head-
waters of the Vaal and Orange Rivers begin their long journey
to the Atlantic, while the shorter rivers of Natal begin as mere
trickles which descend the escarpment. Not far from its source
the Tugela River plunges over the brink of a precipitous
amphitheatre in the escarpment. Five clear leaps with inter-
192
AFRICAN WATERFALLS
[Capt. J. Broun
Fig. 99
The Kalambo Falls, east of Lake Tangamika, Northern Rhodesia
vening cascades make up a total drop of 2,810 feet. The

Tugela Falls, though of insignificant volume, are probably the
highest in the world. The Great Escarpment, as it is called,
can be traced all round South Africa. It everywhere faces
the marginal lowlands and marks the edge of the plateaus of
the interior.
i93
Interruptions in the Cycle of Erosion
At any normal sequence

stage in an uncompleted cycle the
of changes may be interrupted by earth movements of uplift
or subsidence (often accompanied by faulting), by volcanic
action, or by changes of climate leading to glaciation, in-
creased rainfall, or aridity.The distinctive landscape features
developed under glacial and desert conditions are described in
later chapters. It should be noticed, however, that glaciation
is complicated by associated effects due to a abstraction of
( )
water from the sea to form ice-sheets and its subsequent

restoration when the ice melts away, with corresponding
changes in sea level and consequently in base level and ( b ;
isostatic depression due to the loading of an area by an ice-

sheet, followed by uplift when the ice-sheet retreats and dis-
appears.
Volcanic activity may introduce local accidents, such as
the obstruction of a valley by a lava flow. Youthful features
are then temporarily restored while the river is regrading
itscourse through the obstacle. On a larger scale whole land-
scapes may be buried beneath a thick cover of plateau basalts,
in which case a new cycle then begins on the volcanic surface.
If a region is depressed by earth movements, its surface is
brought nearer to base level, the work to be done by erosion is
diminished, and the stages of the cycle then in progress are

passed through more quickly. When a depression is localised
across the course of a river a lake is formed (see page 432),
—
When coastal regions subside unless sedimentation keeps
pace, as in subsiding deltas —the sea occupies the lower reaches
of valleys and estuaries are formed.
Tributaries which entered
the valley before was drowned now flow direcdy into the
it
tidal waters of the estuary and become dismembered streams.

Rivers like the Thames and Humber are sufficiently powerful
to keep their channels open. The sluggish rivers of old age,
however, may be unable to prevent the growth of obstructive
bars and and the latter then become
spits across their estuaries,
silted up. The Broads of East Anglia occupy the site of a
!94
EFFECTS OF REJUVENATION
former estuary which has already been largely obliterated in

this way. The rivers now flow between levee-like embank-
ments of sediment deposited along their sides, and thus they
have become separated, except for local channels, from the
shallow waters of the intervening broads.
The remaining sections of this chapter are devoted to out-
standing features, such as river terraces, incised meanders, and
canyons, developed by rivers in response to uplifts which are
rapid compared with the slow secular uplift involved in the
maintenance of isostasy. As a river is raised further above its
base level, the work to be done by erosion is increased, and the
river is obliged to begin afresh the task of grading its course.
The river has been rejuvenated and, as a further consequence,
the landscape is correspondingly revived. The change begins
where the gradient is steepened, with the restoration of youth-
ful features such as rapids, and gradually works upstream.
The newly deepened part of the valley is sunk as a gorge or
narrow V in the wider V
or trough-shaped floor of the pre-
existing valley. The cross-profile shows a marked change of
slope where the earlier valley form is intersected by the new.
Above the point to which the new features have receded
commonly marked by a step or “ knick ” in the long-profile
the river and its valley retain their former characteristics.
The knick-point is most marked when the uplift has been even ;
it may not be developed if the elevation was brought about by
gentle tilting.
River Terraces
When a river that has already established a flood-plain is

rejuvenated, it cuts through its own deposits into the under-
lying rocks. The sides of the original alluvial plain are then
left as flat above the new level of the river. In the
terraces
course of time the new valley is widened and a second flood-
plain forms within the first one, of which only local remnants
may survive. By subsequent uplift and rejuvenation a second
pair of terraces may then be left on the valley side. The sides
of many of the lowland valleys of Britain (Plate 43A) and
!95
—
Western Europe and indeed in many other parts of the
—
world are bordered by a series of such river terraces, each
corresponding to a phase of valley widening and deposition,
following one of uplift (relative to sea level), rejuvenation, and
valley deepening. A typical terrace is a platform of bedrock
thickly veneered with a sheet of river-gravel and sand passing
upwards into finer alluvium.
As illustrated in Fig. 100, the Thames has three terraces :
(1) The High or Boyn Hill Terrace, named after a locality near
Maidenhead, where it is well preserved. The gravels, mostly
composed of flint, contain the fossil remains of extinct species
of elephant, hippopotamus, and rhinoceros. The climate
indicated was warm and genial. Man had already appeared,
as palaeolithic flint implements are also found. (2) The Middle
WESTMINSTER HDLBORN PENTONVILLE

RIVER Boyn Hill Terrace
Taplow THAMES Taplow Terrace
3
2001
'
Terrace Flood Plain •
Terrace
London Cloy
Fig. 100
Section across London to show the paired alluvial terraces of the River Thames
(After H.M. Geol. Survey)
or Taplow Terrace. Bones and skulls of lions and bears occur

in the lower deposits, but nearer the surface these are absent,
and remains of the hairy mammoth appear. The change
points to the oncoming of the most recent phase of the ice-age.
(3) The Lowor Flood-plain Terrace. Here the mammoth is
stillfound, accompanied by reindeer and elk. These fossils,
as well as those of Arctic plants, indicate a thoroughly frigid
climate. During continued uplift the Thames cut a gorge
which was ultimately filled with gravel when a later subsidence
brought the valley to near its present level. A boring into the
gorge-gravels at Charing Cross failed to find bottom at 80 feet
below the river. Highly polished neolithic implements occur
in the deposits overlying the gorge. Excavation of the London
Docks has revealed the presence of three beds of peat containing
relics of oak and beech. Above the oldest of these fossil forests
196
GORGE OF THE WEAR AT DURHAM
the first signs of the Bronze Age appear. And so we reach the
“ made ground ” of historical times, and the alluvial meadows
and marshes of the present day.
Incised Meanders and Gorges

If, at the time of rejuvenation, a stream was meandering
on a valley floor underlain by resistant bedrock, with only a
thin cover of easily eroded mantle deposits, if any, the deepen-
ing channel is etched into the underlying rocks, while the
Fic. 101
Map and section showing the incised meander of the Ri\er Wear at Durham
1390 )
197 14
original winding course is still preserved. In this way incised
“ hair-pin gorge ”
or entrenched meanders are produced. The
of the Wear at Durham is a familiar British example (Fig. 101).
A well-protected site within the loop was selected for the
cathedral, which is thus enclosed by the gorge on three sides.
The fourth and easily vulnerable side was safeguarded by
building a castle there.
The change of form of incised meanders, and the wearing
back of the confining walls are relatively slow processes con-
trolled by lateral undercutting of the river banks. Localised
Fig. 102
To illustrate the origin of Rainbow Bridge, Utah (see Plate 44b)
undercutting on both sides of the narrow neck of a constricted

loop sometimes leads to the formation of a natural bridge.
On each side of the constriction a cave is worn, especially if
the rocks at river level are weak. Eventually the two caves
meet, and the stream then flows through the perforation.
The stronger rocks of the roof remain for a time as an arch
spanning the stream, and the loop-shaped gorge at the side
is abandoned. In Utah, where recent uplift has made possible
the development of many deeply incised meanders (Plate 44a),
there are several examples of such arches. The most impres-
sive of these is Rainbow Bridge (Plate 44b and Fig. 102), a
198
GRAND CANYON OF THE COLORADO
graceful arch of sandstone which rises 309 feet over Bridge
Creek in a span of 278 feet.
Bridge Creek is a tributary of the Colorado River, and is
thus related to one of the world’s most awe-inspiring scenic
—
wonders the Grand Canyon of the Colorado (Plate 1).
Towards the end of Tertiary times the region which is now the
high plateau of northern Arizona and southern Utah had been
reduced to a land of old age topography traversed by a valley,
50 miles wide or more, margined by mesas and buttes on the

north and by lava plains and cones on the south. A late
Tertiary uplift of 6,000 to 8,000 feet then rejuvenated the old
river, and the cutting of the Grand Canyon began. From
the Rocky Mountains of Colorado in the east the riv er receives
sufficient water to carry it successfully through hundreds of
miles of desert country.
The Canyon has now reached a maximum depth of 6,250
feet, but downward erosion by the heavily laden river is still
actively in progress. A narrow inner gorge has been cut
199
through a thousand feet of crystalline rocks (Fig. 70). The
walls above, carved through a nearly flat Palaeozoic cover ol
strong sandstones and limestones alternating with weak shales,
rise by a succession of steps and slopes of varied colours which
add to the architectural grandeur of the scene. As a result
of this differential erosion under arid conditions the width of
the Canyon from rim to rim is now 5 to 15 miles. During the
recession of the walls bold spurs between the bends have been
carved into pyramids and isolated pillars (Fig. 103). The
plateau is trenched by several tributary canyons, but otherwise
the general surface is but little dissected. The present cycle
is still in the stage of early youth (see Plate 84b).
Antecedent Drainage of the Himalayas

During the uplift of a great mountain range it may happen
that a river which was already flowing across the site of the
future mountains continues to deepen its valley while the
uplift is in progress, so that it becomes permanently entrenched
in the rising landscape. A continuously rejuvenated river
which thus succeeds in maintaining the slope of its channel
from a source behind the mountains to the plains in front is
called an antecedent river, to express the fact that the river was
in existence before the mountains which have risen across its
course.
By far the most remarkable examples of antecedent rivers
are those which cross the Himalayas. The watershed is not
along or near the highest peaks, as might have been expected,
but well to the north in Tibet. From their upper courses on
the plateau the Indus, Bramaputra, and many of the head-
waters of the Ganges traverse the ranges by way of deep gorges
cut in the bottom of steep V-shaped valleys. As the Indus
passes through Gilgit in Kashmir, the river itself is only 3,000
feet above itsbut the precipitous walls by which it is
delta,
confined of nearly 20,000 feet. Like a gigantic
rise to heights
saw the river has cut through 17,000 feet of rock, keeping pace
with a like amount of uplift.
200
KILIMANJARO, TANGANYIKA TERRITORY
UPLIFT OF THE HIMALAYAS
The River Arun, a tributary of the Ganges, rises in Tibet

at a height of 22,000 feet, and passes through a stupendous
gorge between Everest (29,140 feet) and Kangchenjunga
(28,146 feet), the rocks of which were originally continuous.
It could not have done this with the mountains at their present
heights. The erosion has been persistently downwards rather
than headwards, and there are no important waterfalls.
Wager, who has made a special study of the Mt. Everest region,
has suggested that the great peaks of the Himalayas owe much
of their exceptional elevation to the additional effect of isostatic
uplift of the kind discussed on page 190. Instead of a 1,000-foot
block carved into hills and valleys, we have here a 16,000-foot
block, and the corresponding isostatic uplift would therefore
be more than 6,000 feet. Denudation of the Himalayas has
led to the apparently paradoxical result of raising the moun-
tains.
Alluvial Fans and Cones
Many youthful mountain ranges, block mountains, and

plateaus descend steeply to the neighbouring lowlands ;
generally, but not in all cases, because they are bounded by

eroded fault-scarps. Where a heavily laden stream, flowing
swiftly through a ravine or canyon, emerges at the base of
such a slope, its velocity is suddenly checked by the abrupt
change of gradient and a large part of its load of sediment is
therefore dropped. —
The obstructed stream as in delta
—
formation divides into branching distributaries, and the heap
of debris spreads out as an alluvial fan. If the circumstances are
—
such arid or semi-arid conditions are specially favourable
that most of the water sinks into the porous deposit, practically
the whole of the load is dropped and the structure rapidly gains
height and becomes an alluvial cone. There are, of course, all
gradations from steep-sided cones of coarse debris, through
fans of moderate inclination, to widespread sloping sheets
of fine alluvium. Where closely spaced streams discharge from
a mountainous area across a piedmont (a mountain-foot low-
land), their deposits coalesce to form a piedmont alluvial plain.
201
All these features are well displayed in Western America :
along the eastern base of the Rockies, where the Front Range
faces the Great Plains along the eastern edge of the Sierra
;
Nevada, where the eroded fault scarp slopes down to the

Great Basin ; and in many other similar situations on the
flanks of the block mountains of the Great Basin (Fig. 104).
They are also developed on a large scale at the foot of the
Fig. 104
Alluvial cones at the mouths of canyons in southern Utah
( After U.S. Ceol. Survey)
Himalayas and the Andes. A familiar Alpine

example of an
alluvial fan of gentle slope the one between Lake Thun
is
and Lake Brienz on which Interlaken is built. Here a stream

coming down from the Jungfrau has built a fan across what
was originally a single lake, so dividing it into two. In all
the localities mentioned above there are many examples of
river deflection due to the growth across the channel of alluvial
fans deposited by tributaries.
802
GEOMORPHOLOGY

C. A. Cotton
Landscape as developed by the Processes of Normal Erosion. Cambridge
University Press, 1941.
The following general works apply also to many of the later chapters :
W. M. Davis
Geographical Essays. Ginn, Boston, 1909.
S. W. Wooldridge and R. S. Morgan
The Physical Basis of Geography. Longmans, London, 1937.
A. K. Lobeck
Geomorphology : An Introduction to the Study of Landscapes. McGraw-
Hill, New York, 1939.
C. A. Cotton
Geomorphology : An Introduction to the Study of Landforms. Whitcombe
and Tombs, Christchurch, New Zealand, 1942.
A. D. von Engeln
Geomorphology : Systematic and Regional. Macmillan Co., New York,
1942.
L. C. King
South African Scenery. Oliver and Boyd, Edinburgh, 1942.
203
Chapter XII
GLACIERS AND GLACIATION
Snow Fields and the Maintenance of Glaciers
Glaciers are masses of ice which, under the influence of

gravity, flow out from the snow fields where they originate.
Permanent snow fields occur in every continent except Australia.
The level up to which the snow melts in summer, i.e. the lower
edge of a permanent snow field (if present), is called the snow
line. Its height varies with latitude from sea level in the polar
regions to 2,000 feet in S. Greenland and S. Chile, 5,000 feet
in S. Norway and S. Alaska, 9,000 feet in the Alps, 13,000 -
16,000 feet in the Himalayas, and 17,000- 18,000 feet on the
high equatorial peaks of Africa (Plate 45) and the Andes.
It is of interest to notice that the higher summits of the Scottish
Highlands, e.g. Ben Nevis, just fail to reach the level of the
snow it would be in Scotland.
line as
Low temperature alone is not sufficient to ensure the growth
of a snow field. Although northern Siberia includes one of
the coldest regions of the globe, and has permanently frozen
soils from Novaya Zemlya to Bering Strait, it is kept free from
perpetual snow because the scanty winter falls are quickly dis-
sipated in the spring. Snow fields are formed and maintained
where the winter snowfall is so heavy that summer melting
and evaporation fail to remove it all. Snow may also be swept
away by the wind, or lost from steep slopes by avalanching.
The most favourable situations are therefore gentle slopes and
hollows shaded from the sun and sheltered from the wind. A
balance of the snowfall is then left over to accumulate from
year to year, and the snow field grows in depth and surface
area until pressure on the ice which is formed in depth is
outward flowage as a glacier.
sufficient to start its
The snow that first gathers in the collecting
loose feathery
—
grounds gradually passes into a closely packed form the neve
204
VALLEY GLACIERS
—
of the Alpine snow banks retaining a white colour because
of the presence of entangled air. As the snow crystals are
buried and compacted, the air between them is squeezed out,
water from melting snow seeps in and freezes, and so the
deeper layers are transformed into compact, but still porous
ice. The clear blue bands often seen in the exposed flanks of
glaciers differ from the opaque white ice in being free from
bubbles of air. Glacier ice in bulk is a granular aggregate of
interlocking grains, each grain being an individual crystal of ice.
Glaciers originating in valley heads creep slowly down-
wards as tongue-like streams of ice, flowage being maintained
by the yearly replenishment of the neve fields. Ultimately
the glaciers dwindle away by melting and evaporation, their
fronts or snouts reaching a position —
which may be thousands of
feet below the snow line— where the forward movement of the
ice is just balanced by the wastage. In response to seasons of
heavy snowfall whereby the supply is increased, or of low
temperature whereby the wastage is reduced, the glacier
extends farther down the valley, and the snout then becomes
steep. Conversely, in response to a falling off in the rate of
supply or to an increased rate of wastage the snout tapers (as
seen in Plates 5b and 48) and recedes up the valley. During
the present century the fronts of many glaciers are known to
have retreated, though previously they had been slowly
advancing for many years.
The capacity of powerful valley glaciers to reach low
levels before they melt away is due not only to the immense
supplies of ice which are drained from the uplands, but also
to the fact that the area exposed to wastageis small compared
with the volume of the ice. Because the glacier is very viscous,
and therefore moves extremely slowly, it occupies its valley
to a very great depth. To drain a given area the cross section
of a glacier has to be enormously greater than that of the
corresponding river, and in accordance with this comparison
the streams that suffice to carry off the summer melt-water
from the snout of a glacier always appear small and insignificant
(Plate 48).
When glaciers overflow the land and terminate in sea-
205
water sufficiently deep to allow the ice to float, huge masses

break away from the front and become icebergs. Theoretically
about nine-tenths of an iceberg would be submerged if it were
made of pure ice. The actual proportion, however, is subject
to variation according to the proportion of entangled air and
the load of rock debris present in the ice. Some of the vast
tabular icebergs liberated from the front of the Antarctic ice
float with as much as one sixth of their total height above
the sea.
Types of Glaciers
Glaciers fall naturally into three main classes :
(a) Ice sheets and ice caps that overspread continental or

plateau regions of supply, where the snow line is low, creeping
with a slow massive movement towards the margins.
(b) Mountain or valley glaciers occupying the pre-existing
valleys of mountain ranges that riseabove the snow line.
(c) Piedmont glaciers, consisting of sheets of ice formed by
the coalescence of several valley glaciers which have spread
—
out below the snow line like lakes of ice — over a lowland area
of wastage. Certain gradational and subsidiary types also
occur, and are referred to below.
Greenland and Antarctica provide the only examples of
There is, however, over-
continental ice sheets that still exist.
whelming evidence that 25,000 to 30,000 years ago immense
ice-sheets of similar character covered half of North America
and most of north-western Europe. The Greenland ice-sheet,
about half a million square miles in extent, is largely enclosed
within a mountainous rim. Near the middle of the sheet
the ice has been shown by seismic methods (page 371) to be
over 8,000 feet thick, indicating that the elevation of the actual
rock floor is only about 1,600 feet. Towards the edge the
higher peaks and ridges of the mountains project through the
ice as nunataks. The ice itself overflows through passes in the
mountain zone and feeds the valley glaciers of the valleys
leading down to the coast.
206
THE ANTARCTIC ICE SHEET
Fig. 105
Map of Antarctica
The ice sheet of Antarctica, seven times as extensive as that

of Greenland, forms a great plateau rising to over 9,000 feet
(Fig. 105). Except in a few localities of fringing mountains,
where there are marginal glaciers and a coastal strip, the ice
sheet overruns the coast and spreads over the ocean as vast
floating masses of shelf ice. The best known of these is the
Great Ross Barrier, which terminates in sheer cliffs of floating
207
[E.N.A.
Fig. 106
A view of the Malaspina Glacier with snow-clad mountains of Alaska in the

background
ice rising 100 to 160 feet above the Ross Sea. The Barrier ice
isworn away by submarine thawing, marine erosion, and the
breaking off of gigantic tabular icebergs. Wastage is slowly
overtaking supply, for the high walls of the Barrier have
receded many miles to the south since Ross first discovered
them in 1841.
Smaller ice-sheets, distinguished as plateau glaciers or ice caps,
cover large areas in Iceland and Spitsbergen, from which they
emerge through marginal depressions as blunt lobes or large
valley glaciers. The tips of the underlying mountains project
from certain less continuous caps of highland ice. Where the
supply of ice is rather less, these pass into a network of con-
nected glacier systems, the ice of each valley system over-
flowing the cols into neighbouring valleys and smothering all
the lower divides. Such gradational types, well represented in
Spitsbergen, lead to the more familiar valley glaciers.
Trunk glaciers and the tributary valley glaciers feeding
208
(B) Aerial v
GLACIERS AND THEIR MOVEMENT
into them occupy the upper parts or the whole of the valley
system of a single drainage area (Plate 46). Smaller valley
glaciers, with few tributaries or none, are confined to single
valleys. —
Valley glaciers apart from those draining ice-
sheets and caps —characteristically originate in deep arm-
chair-shaped hollows, called conies or cirques, situated at the
valley heads (see Plate 47). Small isolated glaciers occupying
hanging valleys (page 221) or subsidiary corries perched high
on the side of a deeper valley are referred to as hanging glaciers
(tongue-shaped) or corrie glaciers (horseshoe-shaped).
Where a glacier passes from a restricted channel to a more
open lowland, it fans out into an expanded foot, and where
several neighbouring glaciers so emerge a piedmont glacier
results. The outstanding example of the latter type is the great
Malaspina Glacier of Alaska (Fig. 106). Maintained by the
confluence of the glaciers from Mt. St. Elias and the neigh-
bourhood of the Canadian frontier, it has an area of about
1,500 square miles, and locally reaches the sea. As a result of
surface melting, much of the outer margin is thickly covered
with morainic debris and soil which here and there support
dense forests of pine.
The Movement of Glaciers

Were it not for the fact that ice in bulk can flow, the world
would now present a very different appearance. Practically all
the water of the oceans would be locked up in gigantic circum-
polar ice fields of enormous thickness. The lands of the tropical
belts would be deserts of sand and rock, and the ocean floors
vast plains of salt. Life would survive only around the margins
of the ice fields and in rare oases fed by juvenile water.
The most rapidly moving glaciers are those of Greenland,
some of which advance as much as 60 feet a day in the summer.
In general, however, a few feet a day is a more characteristic
rate. The Mer de Glace (Plate 46a) barely exceeds two feet
a day, and the Beardmore Glacier of Antarctica, the greatest
in the world, moves at less than three feet a day.
209
By observing the changes in position of lines of stakes driven

into the ice it is found that the middle of a glacier moves more
rapidly than the sides, and that there is a similar decrease in
velocity near the floor. The rate of flow increases with the
steepness of the slope, with the thickness and temperature of
the and with constriction of the valley sides. Movement is
ice,
retarded by the presence in the ice of a heavy load of debris
and by friction against the rocky channel. These facts suggest
that the flowage of glaciers —
a remarkable phenomenon that
calls for —
a brief explanation is controlled mainly by stress
differences and temperature.
The liberation of molecules of water from ice by a rise of
temperature is a familiar process. It is also well known that
uniform pressure lowers the melting point of ice, and thus
stimulates its transformation into water. More important,
however, is the fact that non-uniform pressure or stress is
many times more effective in liberating molecules of water
from rigid grains of ice. The mechanism of skating provides
the key to the problem of glacier flow. A skater really glides
in a narrow groove of water formed momentarily under the
intense stress applied to the ice by the thin blades of his skates.
As he water immediately freezes again.
passes, the
The interlocking crystal grains within a sloping mass of
ice are subjected to stresses which vary from point to point,
and wherever the strain is most severe, mobile molecules and
microscopic films of water are liberated. These act as a lubri-
cant and facilitate minute movements of the grains among
themselves. As one grain is pressed against another water
diffuses along intergranular boundaries into places of lower
stress, and there freezes on to the grains with which it is in
contact. Slipping along fracture planes in individual grains
also contributes to themovement. Since the pressure gradient
Ts downthe valley, there is a constant migration of material
from grain to grain in that direction. The whole mass thus
flows by a process of re-crystallization which is closely akin to
that involved in the metamorphism of ordinary rocks. The
agents of metamorphism are high temperature, stress differ-
ences, and migrating fluids. In the interior of a glacier the
210
CREVASSES
temperature is already very “ high ” in the sense that the

ice is near its melting point stress is provided by the weight of
;
overlying and upstream ice and the migrating fluid is water.

;
A glacier has an outer crust of ice, averaging perhaps about

200 which the stress differences are in-
feet in thickness, in
promote flowage. This rigid crust is carried along,
sufficient to
and friction against the sides and floor is also overcome, by
the movement of the deeper parts. Where the movement is
retarded or stopped by a load of debris, as often happens
near the snout of a glacier, the stagnant ice is over-ridden by
clean, mobile ice along well-marked thrust planes.
Surface Features of Glaciers
Crevasses. —
Within limits controlled by the processes of ice
flowage a glacier can accommodate itself to its channel. Where
the ice passes through a constriction in a valley it thickens, and
the rigid crust is thrown by lateral compression into wave-like
pressure-ridges. On the other hand, where the valley opens out,
or where a glacier passes over a declivity or round a bend, or
fans out into an expanded foot, the ice is stretched and cracked
into a series of deep gaping crevasses. These may be hidden by
snow bridges, and the treacherous surface then becomes very
dangerous to cross.
Transverse crevasses develop across a glacier wherever there
is a marked steepening of the slope of its floor (Fig 108).
Longitudinal roughly parallel to the direction of
crevasses,
flowage, are formed wherever ice is obliged to spread out.
Marginal crevasses (Plate 47), pointing upstream from the
sides of the glacier, develop as shown in Fig. 107. Because of
the higher velocity towards the middle, a line AB is later ex-
tended to A' B', and the resulting tension cracks the ice at
right angles to A' B'. When two or more sets of crevasses
intersect, the surface of the glacier is torn into a broken mass
of jagged ice pinnacles known as seracs. Near the top of the
neve field of a corrie a very wide and deep crevasse, called the
bergschrund, opens in summer where the head of the glacier
21 i
Fig. 107
Diagram to show the development of marginal crevasses as a result
of differential ice flowage
pulls away from snow adhering to the precipitous

the ice and
walls (Fig. 108 ). It happens that several such fissures
often
are formed instead of an especially large one (Fig. 109 ).
Moraines . —
Rock fragments liberated from the steep slopes
above a glacier, mainly by frost shattering, tumble down on
the ice and are carried away. Thus the sides of a glacier
Fig. 108
Schematic section through a corrie occupied by the head of a glacier, showing
the bergschrund near the top and transverse crevasses above the threshold
212
[
h\ ,Y. Ashcroft
Ice cave and source of the Rhone, snout of the Rhone Glacier, near Gletsch,
Sw lt/erland
PT VTE 49
TYPES OF MORAINES
become streaked with long ribbons of debris described as

lateral moraines. When two glaciers from adjacent valleys
coalesce, the inner moraines of each unite and form a medial
moraine on the surface of the united glacier (Plate 46 b ) A trunk.
glacier fed by many tributaries may thus come to be ridged

with a series of medial moraines composed of materials from
different parts of the area of supply, thus providing samples
of rocks that might otherwise be unobtainable.
Sooner or later part of the debris is engulfed by or washed
Fig. 109
Sketch of bergschrund crevasses in an ice-filled come at the head of a glacier

(From an aerial photograph of the Gelmerhorner, Switzerland)
into crevasses. Material that is enclosed within the ice is re-

ferred to as englacial moraine. A certain proportion reaches the
and there, together with the material plucked
sole of the glacier,
or scraped from the rocky floor, it constitutes subglacial moraine.
If the lower part of the ice becomes so heavily charged with
debris that it cannot transport it all, the excess is deposited as
ground moraine, which is then over-ridden by the more active
ice above. All the varied debris, ranging from angular blocks
and boulders to the finest ground-down rock flour, that finally
arrives at the terminus of the glacier is dumped down when
(396) 15
213
the ice melts. If the ice front remains stationary for several
years an arcuate ridge is built up, called a terminal or end
moraine. If, however, the snout is retreating summer after
summer, no piling up of a ridge is possible. The load liberated
from the receding front then forms an irregular sheet which
rests on the ground moraine already deposited.
Melting and Drainage. —
Thin isolated slabs of rock or patches
of debris on the surface may be sufficiently heated by the sun
to melt the underlying ice. Larger blocks, however, act as a
protection from the sun’s rays, and as the surrounding ice
melts away they are left as glacier tables perched on a column
of ice. Even morainic ridges may stand out for a time on thick
walls of ice.
In sunny weather small pools and rills diversify the surface,
gathering into streams which mostly fall into crevasses. By
a combination of melting and pot-hole action (aided by sand
and boulders) deep cauldrons called glacier mills or moulins are
worn through the fissured ice, and the water may escape to the
snout through a tunnel. There, with melt-water draining down
the tapering end, it begins to flow down the valley as a milky
stream laden with fine particles (Plate 48).
Glacial Erosion
As we have seen, a glacier soon acquires a load of morainic

material. Moreover, loose debris on the floor and sides is
quickly dislodged and engulfed by actively advancing ice.
Blocks from protuberances of jointed bedrocks are sheared off
and withdrawn from the down-stream and unsupported side
by a quarrying process referred to as plucking. The ice works
its way into joints, bedding planes, and other fractures and
closes round projecting masses with a firm grip, so that block

after block is torn out of position and carried away. The
ra gg e d surface left behind readily susceptible to further
is
plucking,and the process continues until the obstruction is

removed or the glacier wanes.
Thus, even pure ice, which by itself would be a very in-
effective tool for eroding massive rocks, is sooner or later trans-
214
RATE OF GLACIAL EROSION
formed into a gigantic flexible file with embedded fragments
of rock for teeth. the scraping and scratching of rock
Abrasion is
surfaces by debris frozen into the sole of a glacier or ice sheet.

The larger fragments cut into and groove the floor and sides,
and are themselves worn and striated. The finer materials
flat
and polishing the rock surfaces

act like sandpaper, smoothing
and producing more powdered rock, or rock flour, in the process.
The is extremely variable.
rate of glacial erosion Theoreti-
cally, the rate of abrasion approximately proportional to the
is
cube of the velocity of the ice against its channel. Thus a

powerful glacier in northern Greenland may be 30,000 times
more effective than the sluggish glaciers of the Alps. A conti-
nental ice sheet moves so slowly that it cannot be expected to
do much more than remove the soil and smooth-off the minor
irregularities of the buried landscape. In such a case the
broader features of the pre-glacial relief are, on the whole,
protected from denudation, though the surface is modified
in detail into a characteristically hummocky form of knobs
and hollows which reflect the varying resistances offered by
the rocks to abrasion. But when the outflowing ice, or a
valley glacier in a mountain district, is concentrated in a
steeply descending valley, the erosive power reaches its maxi-
mum, and the pre-glacial relief is strongly accentuated.
Beyond the region of steep gradients and rapid movement the
rate of erosion graduallyfalls off and gives place to deposition
becomes overloaded and reaches the zone of wastage.

as the ice
The three realms of supply, movement, and wastage are clearly
seen in Plate 46b.
The geological work accomplished by ice, including erosion
and deposition and the resulting effects of these processes on
the surface, is collectively known as glaciation. The sculpturing
of the surface beneath existing glaciers can
be studied directly
only in a limited way, by exploring ice caves and descending
crevasses. Much more can be learned by taking advantage of
the fact that many glaciers have receded up their valleys in
recent years. Throughout the historical period there have
been periodic fluctuations in the volume and extent of glaciers,
but without any indication of cumulative waxing or waning.
215
[C. E. Wegmann
Fig. 110
Ice-moulded surface of roche mnutonnee type, Sitterskar, Soderskar Archipelago,
south coast of Finland
Thousands of years ago, however, the glaciers were enormously

greater than their shrunken descendants of to-day. One of the
earliest observers to suspect that glaciation was formerly far
more extensive was de Saussure, the first scientific explorer
of the Alps. In 1760 he noticed that miles below the snouts
of the Alpine glaciers the rock surfaces were scratched and
—
smoothed in striking contrast with the frost-splintered peaks
—
above and strewn with morainic material exactly like that
still being carried and deposited by the ice. He rightly con-
cluded that the glaciers had formerly extended many miles
beyond their then limits. But the “ glacial theory ” met
with scant approval until after 1840, when the great naturalist
Louis Agassiz awakened more general interest in the subject
by the publication of his classic studies on the glaciation of the
Alps. In later years Agassiz recognised that the similarly
striated rock surfaces and morainic deposits of Scotland were
also due to the former passage of ice. With a wealth of irre-
216
ROCHES MO UTONNEES
sistible evidence he convinced the scientific world that such

features could be accounted for in no other way. It is now_
familiar knowledge that the landscapes of vast areas of Europe
and North America bear the unmistakable hall-marks of
glaciation. Thus it happens that in many countries the char-
acteristic effects of ice erosion and deposition, modified but
little by subsequent weathering and river action, can be seen
and studied close at hand.

Among the evidences of erosion by continental and valley
glaciers, striated surfaces (Plate 49a) and ice-moulded hum-
mocks of the more resistant bedrocks (Fig. 110) are of wide-
spread occurrence. The residual hummocks vary widely in
size, and have a characteristically stream-lined form which is
Section across a typical roche moutonnie, showing the effect of ice abrasion where
the rock is sparsely jointed, and of plucking where jointing is well developed
related to the direction of ice movement. The side up which

the ice advanced rises as a smoothly abraded slope, while the
lee side falls more steeply, sometimes as an abraded slope, but
often by a and ledges obviously due to
step-like series of crags
the plucking out of joint blocks (Fig. Ill and Plate 49b).
Seen from a distance the more isolated examples resemble
sheep lying down, or wigs placed “ face ” downwards. They
are, therefore, described as roches moutonnees, a term first used
by de Saussure in 1804 in reference to the sheep-skin wigs,
styled moutonnees which were then in vogue.
,
Highly resistant obstructions, such as old volcanic plugs,

that lay in the path of the ice, like protruding knots in a plank
of wood, are responsible for an erosional feature known as
crag and tail (Fig. 112). The crag boldly faces the direction
217
from which the ice came, while the tail (bedrock with or
without a covering of boulder clay) is a gentle slope on the
sheltered side, where the softer sediments were protected by
the obstruction from the full rigour of ice erosion. A classic
example is provided by the Castle Rock of Edinburgh, from
the eastern side of which the High Street follows the sloping
crest of the tail. The massive basalt plug diverted the ice-
flow, and deep channels, now occupied by Princes Street
Direction of
Volcanic Plug Upper Lower

Old Red Sandstone Carboniferous
of Basalt \
i
Cohnton Castle
Fault Fault
Fig. 112
Crag and Tail, Edinburgh
Gardens and the Grassmarket, were excavated in the sedi-

ments on each side of the crag and tail feature.
In mountainous and upland coastal regions with well
developed valley systems the topographic modifications super-
imposed on the landscape by glacial erosion include U-shaped
valleys with truncated spurs and hanging tributary valleys ;
corries or cirques surmounted by sharp-edged ridges and
pyramidal peaks and reck basins and fjords. Waterfalls
;
descending the precipitous valley sides, and lakes occupying

the overdeepened hollows of the valley floors, add variety to
an assemblage of features that can be easily distinguished from
those of unglaciated regions.
Corries and Associated Features
Ithas been observed that chance hollows occupied by per-

sistent snowbanks are steadily cut back and deepened by (a)
disintegration of the marginal and underlying rocks by frost
218
[i- . X. Ashcroft
U-shaped glaciated \allev. \’al Giuf tributary of the Upper Rhine) viewed
i
from the slopes near the chalets of Milrz, Switzerland, Granite of the Aar
Massif m the middle distance
DEVELOPMENT OF CORRIES
and thaw, and (b) removal of the shattered debris by falling,
avalanching, and transport by melt-water. By this process of
snow-patch erosion or nivation the slopes above are undercut
and the surrounding walls are kept steep as they recede
(Fig. 113). The larger hollow's grow more rapidly than the
smaller ones, especially near and above the snow line, until
the mountain slopes and valley sides are festooned with deep
snowfields, the largest of all being at the valley heads. Eventu-
ally these nourish small glaciers which carry away the debris
and begin more active excavation of the floor. Headward
erosion of the walls continues, not only by frost sapping at the
exposed edges of the snowfield, but also by a process of sub-
glacial disintegration which comes into play whenever the
[F. Nansen
Fig. 113
Conies developed by snow rotting on the cliffs of Spitsbergen
bergschrund allows surface water to reach the rocks behind

or beneath the ice. Draining into cracks and joints, the water
freezes and breaks up the rocks until they are gripped by the
ice, and carried away as ground moraine. Thus by the co-
operation of several processes the great amphitheatres known
as corries or cirques are eventually hollowed out (Figs. 108
and 116).
During the stage of most intense glaciation the floor of a
growing corrie is subjected to especially vigorous scouring in
consequence of the great thickness and high pressure of the ice
and snow above it. A shallow rock basin may then be ex-
cavated by the outflowing ice, to become the site of a mountain
tarn or lake after the ice has disappeared (Plate 54b). Such
lakes may also be held back by arcuate ridges of morainic
2ig
material left stranded by the waning ice during its final re-
cession.
Two adjoining corries may approach and intersect until
only a sharp-edged dividing wall remains between them. The
resulting precipitous ridge is known as an arete. When the
ice has gone the steep rocky slopes fall a ready prey to frost
action, and soon become aproned with screes. Many an
upland region has been eaten into by corrie erosion from
several sides at once, and so reduced to a series of aretes
radiating like a starfish from a central summit. Snowdon and
Helvellyn are good examples. At a later stage the aretes
themselves are worn down, and the central mass, where the
heads of three or more corries come together, remains isolated
as a conspicuous pyramidal peak. In this way the horns of the
Alps have been formed, the world-famous Matterhorn being
the type example of its class (Plate 50).
Modifications of Valleys by Glacial Erosion
By the passage of a vigorous glacier through a pre-existing

river valley the mantle of rock-waste is removed, the over-
lapping spurs are trimmed off and ground into facets, and the
flooris worn down. The valley is thus widened and deepened,
and is eventually remodelled into a U-shaped trough with a
broad floor and steep sides and a notable freedom from bends
of small radius (Plate 51). Flat floors are not uncommon,
however, where the bottom of the trough has been levelled up
by subsequent deposition of alluvium. Whole valley systems
may have been completely overwhelmed by ice-sheets, but in
less severely glaciated regions, where have not been
valleys
entirely filled by ice, the upper slopes may remain as high-level
benches which meet the ice-steepened walls in a prominent
shoulder (Plate 51). The cross profile is like a U sunk in a V.
Tributary valleys have their lower ends cut clean away
as the spurs between them are ground back and truncated
(Fig. 114). The floor of a trunk glacier, moreover, is deepened
more effectively than that of a weak lateral feeder. Thus, after
220
GLACIATED VALLEYS
Fig. 114
Block diagram illustrating some of the characteristic landscape features of glaciated
valleys and mountains U-shaped valley
: truncated spurs ; hanging tributary
;
valleys ;
corries, aretes, and horns. The summit
of the hill on the right and the
bench across the lower right-hand comer are depicted as they would be if they
had remained unglaciated. (Modifie.l after W. M. Davis)
a period of prolonged glaciation the side valleys are left hang-

ing high up on the flanks of the main trough. The streams
from such hanging valleys plunge over their discordant lips in
cascades or waterfalls, some of which are amongst the highest
in the world. The Yosemite Valley in the Sierra Nevada of
California is renowned for its impressive examples of these
and other spectacular features due to glacial erosion (Plates
221
52 and 53). A remarkably similar glacial trough the finest —

of its kind in Europe —
is the Lauterbrunnen valley, with its
celebrated Falls, between Interlaken and the Jungfrau.

Glacially excavated floors are deepened very unevenly,
the effect at each point depending on the thickness and velocity
of the ice and the nature and structure of the bedrocks. Poorly
consolidated strata are scoured out more rapidly than resistant
rocks, and tracts of well-jointed rocks are selectively quarried
away by plucking. Thus, where the ice encounters a sequence
of rocks of varied resistance the floor is excavated into a series
of successive steps, often with abrupt descents from one tread
Longitudinal profile along the Yosemite Valley (see Fig. 184 for locality).
A typical “ glacial stairway ” developed by selective ice erosion
Length of section = 36 miles
to the next (Fig. 115). A comparable effect due to varying

ice thickness is seen where a trunk glacier has been fed from
a cluster of corrie glaciers around the valley head. Below the
confluence the suddenly increased erosive power is witnessed
by a correspondingly sudden drop from the corrie mouths to
the floor of the glacial trough (Fig. 116). “Trough-end”
walls of this kind are magnificently developed at the heads of
the Lauterbrunnen and Zermatt valleys.
Thus it happens that the long profile of a glaciated valley
may resemble a giant stairway. The treads may even be
hollowed into basins with a barrier of resistant rock in front.
Such rock basins are now occupied by lakes of various lengths
and depths, or by tracts of alluvium representing the sites of
222
PLATE
Valley
I'.ills
Ilanmmj
Bridalvcil
Rocks,
and
Cathedral
115)
Fig.
(see
Survey)
Park
California
_ National
Dome
Nevada,
Half
U.S.
v
Sierra
Amlerwii,
//.
Valley,
(A*.
loseiime
Capital!
111
of
spur
l'ruucateU
Fig. 116
Block diagram to illustrate the “ trough-end ” rock step at the head of a
glaciated valley fed by several confluent corries. ( Modified after W. M. Davies)
that is, up the slope at the lower end. However,

in the appro-
priate circumstances, this objection not well founded. Ice
is
can move upslope provided that it has a means of exit, and is

moving from a place of high pressure to a place of low pressure.
The condition for flow is that the surface of the ice should have
a downward gradient sufficiently steep to maintain the re-
223
quisitehead of pressure. The following are examples of lakes

which occupy conspicuously overdeepened basins and troughs.
Lake Maximum depth Height of surface Maximum depth oj
(in feet) aboie sea level below sea level
Lake District
Windermere * 219 128 91
Wastwater * 258 200 58
Scotland
Loch Coruisk, 125 25 100
(Plate 54a]
Loch Lomond * 653 20 633
Loch Ness 754 53 701
Loch Morar 1,017 30 987
Swiss-Italian Alp
L. Maggiore 1,220 636 584
L. Como 1,345 650 695
* Morainic deposits on the terminal bedrock barrier add slightly to the depth
of these lakes. See Fig. 118 for an example of such a double barrier.
Fjords are greatly overdeepened glacial troughs that reach

the coast below sea level, so that, instead of forming elongated
lake basins, they have become long arms of the sea stretching
inland between steep rocky walls. The terminal rock barrier
(with or without a cover of moraine) occurs near the seaward
entrance, and is usually submerged at a shallow depth. This
is the threshold of the fjord. Along the west coast of Scotland
gradations from the exposed barriers of the freshwater lochs
to the submerged thresholds of the fjords (sea lochs) are well
illustrated. The terminal rim of Loch Morar is within a very
and is only 30 feet high. The thres-
short distance of the sea,
holds of Lochs Etive and Leven are lower, but are uncovered
at low tide. Twenty-three of the remaining sea lochs have
permanently submerged thresholds near their entrances.
Fjords have been developed during the intense glaciation
of dissected coastal plateaus and mountains of appropriate
structure in countries such as Scotland, Norway, Greenland,
Labrador, British Columbia, Alaska, Patagonia, and New
Zealand. In plan (Fig. 117) they everywhere have a char-
acteristic rectilinear pattern which is clearly determined by
the distribution of belts of structural weakness. The latter
may be synclines of relatively weak sediments or schists
enclosed by massive crystalline rocks (as in the Sogne and Har-
224
FJORDS
danger Fjords of Norway), but more commonly they are

fractured belts with closely spaced joints locally accompanied
by faults and dykes. The pre-glacial rivers carved their valleys
along these lines of least resistance. The valleys in turn
confined the ice and guided its flow, and because the structure
facilitated plucking, the valley floors were steadily deepened,
often to an extraordinary degree. In some of the fjords of
Norway and Patagonia the sea is over 4,000 feet deep. Neigh-
bouring fjords, however, vary enormously in depth, in accord-
ance with the varying resistance of the excavated rocks. The
distribution of fjords is thus conditioned by ( a ) appropriate
tectonic structures in upland regions near the sea, ( b ) pre-
existing valleys which followed these structures, and (c) heavy
glaciation by seaward-moving ice of sufficient thickness and
surface slope to ensure that the main valleys were overdeepened
up to or beyond the coast.
Glacial Deposits
As glaciers and reach the zone of wastage beyond

ice sheets
the region of active erosion they become overloaded and begin
to drop their burden of debris. During the subsequent dis-
appearance of the ice, in response to an amelioration of climate,
the zone of deposition retreats with the receding ice front until
the whole of the load has been deposited. The glacial deposits
thus left stranded on the landscape and the glaciofluvial sands
and gravels transported and deposited by the associated melt-
waters have long been grouped together under the general
term drift. At one time the vast spreads of drift that indicate
the former extent of the ice across Europe and North America
were thought to be flood deposits, and many attempts were
made to assign them to the deluge of Noah. Eventually,
however, it was recognized that the commonest type of drift,
the haphazard assemblage of material known as boulder clay
or till, could not possibly have been deposited by water.
Boulder clay has obviously been dumped down anyhow
in a completely unsorted and unstratified condition (Plate 55a).
226
GLACIAL ERRATICS
Its constituents range from the finest rock flour to stones of all
sizes up to boulders that are occasionally of immense bulk.
It usually consists of a varied assortment of stones embedded
in a tenacious matrix of sand, clay, and rock flour. Most of
the stones, like those of screes, are irregular fragments showing
little or no sign of wear or tear, but a few can generally be
found which have been rubbed down and scratched and

grooved, clearly by scraping along the rocky floor over which
they were dragged by the ice.
A characteristic feature of glaciated regions is the occurrence
of scattered boulders of rocks that are foreign to the place where
they have been dropped. These ice-transported blocks,
carried far from their parent outcrops, are called erratics.
The largest ones commonly rest on abraded surfaces where the

normal drift is thinly scattered or confined to hollows. Some
have been stranded in exposed and precarious positions.
Such perched blocks (Plate 55b) are striking monuments to the
former passage of ice, and as such they were amongst the first
evidences of glaciation to be recognized. In 1815 Playfair
pointed out that “ a glacier . which conveys the rocks on its
. .
surface ... is the only agent we now see capable of transporting

them to such a distance.” The long trails of erratic blocks of
easily recognized rocks affordan unfailing guide to the direction
of movement of the ice that carried them. Boulders of the
well known Shap granite, for example, can be traced from
their original home in Westmorland across the Pennines by
way of the Stainmore Pass into the Vale of York. Ailsa Craig
in the Firth of Clyde is an upstanding mass of finely speckled
granitic rock which can be identified with certainty. Erratics
of this rock, as isolated blocks and as stones embedded in
boulder clay, are found in Antrim, Galloway, and the Isle of
Man, and on both sides of the Irish Sea as far as Wicklow and
South Wales, and show that Ailsa Craig lay in the track of a
great southward-travelling glacier. Familiar Norwegian rocks
from the Oslo district, such as rhomb-porphyry and larvikite
(a shimmering blue syenite much used for shop fronts) occur
as erratics along the Durham and Yorkshire coasts, and prove
that the Scandinavian ice sheet crossed the North Sea and at
227
times overran the British shores. At other times these rocks

reached northern Germany, showing that the directions of
ice dispersal were not always the same.
At or near the maximum extension of a glacier or ice sheet
the front may have remained stationary or nearly so (forward
movement being just balanced by wastage) sufficiently long
for a ridge-like terminal moraine to be heaped up. Similar, but
later, terminal moraines, sometimes distinguished as recessional
[Planet News Ltd.

Fig. 118
Lake on the Dontouz-Orun Pass, Caucasus Mountains, occupying an ice-eroded

rock basin with a terminal moraine on the threshold
moraines, mark the sites of halting stages during the shrinkage

of the ice, and indicate that forward flow was maintained so
that a steady supply of debris was brought up to the ice front.
In certain lowland regions the ice appears to have fanned out
and become stagnant. Such “ dead ” ice simply melts away
from the top and sides, and also from the edges of crevasses,
and liberates its debris without forming terminal moraines.
The terminal and chief recessional moraines formed at
successive stands of the European ice sheet are shown in Fig.
228
PLATE 34
If M (jcoI. Survey
(A) Loch Coruisk, Sk\e. A double rock basin exca\ated bv ice. Cuillin
Hills behind
rG P. Ahiahum, Ltd Kcsuick

. ,
(Rj Snowdon and Glaslyn. vieued from Glib Goch, North Wales
JI.M. (n’ol. Survey
(B Glacial erratic of Silurian grit resting on a plinth of Carboniferous
Limestone, Norber. near Austwick, Yorkshire
MORAINIC DAMS
127. Similar features, traversing the country in broad loops to

the south of the Great Lakes, mark the various pauses in the
recession of the North American ice. The terminal moraines
left by mountain glaciers cross their valleys as crescent-shaped
ridges (concave upstream) which in some cases continue along
the valley sides as less conspicuous lateral moraines (Fig. 118).
Many of the lakes of glaciated valleys are held up by morainic
dams, and there were formerly innumerable smaller ones that
Fic. 119
Map of the lateral and terminal moraines bordering the rock basin of Lake Garda
at the foot of the Italian Alps
have since been silted up or drained by a river gorge cut through

the barrier. Lake Constance and Lake Garda (Fig. 119) are
notable examples of lakes owing their existence to terminal
moraines across the outer valleys of the Alps. Most of the
larger lakes of the Lake District and many of the Scottish
lochs have morainic dams.
In the tracts between the moraines the spreads of boulder
clay naturally vary widely in character and thickness from
place to place. In certain regions where the boulder clay is
thickly plastered over a floor of low relief it has been moulded
( 396 ) 229 16
by the into swarms of whale-backed mounds called

ice
drumlins. Being distributed more or less en echelon the mounds
”
give rise to what has been aptly described as “ basket of eggs
topography (Fig. 120). In the intervening depressions drainage
is poor and confused, and is responsible for such features as
ponds, marshes, and water-logged meadows. One of the most

densely packed drumlin belts stretches across northern Ireland
from Co. Down to Donegal Bay, and contains tens of thousands
of these stream-lined mounds (Fig. 121).
Drumlins are commonly a quarter to half a mile long, but
there is every gradation from low swells to enormous examples
a mile or two in length and 100 to 200 feet high. Most of them
are elongated in the direction of ice movement, and the end
Fig. 120
“ Basket of eggs ” topography. A typical drumlin landscape moulded by ice
which moved from right to left
facing upstream is characteristically blunt and steep compared
with the tapering downstream end. The profile along the

length is thus the reverse of that of a roche moutonnee. The
latter being a product of ice erosion, it is reasonable to suppose
that drumlins would have had a similar form if they had
been moulded by over-riding ice from boulder clay already
deposited. Since they have not this form, it becomes probable
that drumlins were fashioned from ground moraine that was
being freshly deposited by ice in which the capacity to erode,
transport, and deposit was distributed in a curiously rhythmic
pattern.
It is generally accepted that drumlins were formed under
deep ice at a distance of several miles from the front towards
which it was advancing. There must have been variations
in the ice flow, possibly owing to such causes as (a) underlying
230
Fig. 121
Map showing part of a drumlin tract in Co. Down

(After J. K. Charlesworth)
231
)
obstructions —
some drumlins have cores of solid rock with
boulder clay banked against them (
b longitudinal crevasses
;
in the icewhich localized places of deposition and (c) varia- ;
tions in the load carried, the clearer and more vigorous ice
flowing round the more heavily charged and sluggish ice.
From the nature of the case, however, drumlins have never
been seen in course of formation, and the exact mechanisms
involved are still far from being clearly understood.
Glaciofluvial Deposits
The drainage from the long front of an ice sheet escapes

by way of an immense number of more or less temporary and
constantly shifting streams. These carry off a great deal of
sediment and, as the velocity is checked, low alluvial fans or
Fig. 122
The characteristic assemblage of features seen on a recently glaciated area of
low relief
deltas are deposited, according as the ice terminates on land

or in standing water. On
land the fans spread out and coalesce
into gently sloping outwash plains (Fig. 122) of irregularly
ranging from coarse gravels near the source
stratified drift,
to sand further out, and finally to clay. Valley floors are
choked with spreads of similar deposits ; mainly coarse, how-
232
OUTWASH PLAINS
ever, because the finer materials are rapidly washed down-

stream.
Beyond the terminal moraines of ice sheets, where the
supply of debris is abundant, outwash plains may extend for
many miles. Vast areas of the North American prairies have
been smoothly veneered with sediment in this way. Between
successive moraines the outwash drifts rest on previously
deposited boulder clay. Masses of stagnant ice, left stranded
between deep crevasses as the main front melted back, are
often surrounded and even buried by drift, and as they melt
away they leave the surface pitted with depressions known as
kettle holes. These are extremely irregular in shape and distri-
bution, and many of them still contain lakes or ponds (Fig. 123).
Such drift-covered regions are further diversified by mounds
( kames
), long
winding ridges ( eskers Fig. 124), and relatively
,
short and straight ridges ( crevasse infillings). All of these are

built of crudely bedded gravel and sand, showing that they
are features for which glacial streams were responsible.
233
Fig. 124
Esker, Tolvajarvi, Finland. The lake occupies the depressions in an irregular
surface of glaciofluvial sands and gravels
Karnes are isolated or clustered mounds, each of which

represents a steep-faced localized alluvial cone or delta built
up by a stream emerging at a high level from an ice wall, or
perhaps between two confining walls at the head of an embay-
ment in the ice front. As the ice withdrew, the unsupported
back or sides of the accumulation slumped down, leaving a
mound with slopes corresponding to the angle of repose of
gravel or sand. If such a stream, instead of being short and
temporary, were long and persistent, then the deposit con-
tinuously formed at its mouth would grow backwards as the
ice retreated, thus extending into a winding ridge that would
reproduce the course of the stream. Some eskers may have
originated in this way. But such a stream would also deposit
sediment while flowing through its tunnel in the ice, thus
gradually raising its floor. Most eskers are therefore regarded
as the infillings of the tunnels of unusually long sub-glacial
streams. In some cases, where later outwash drift has lapped
against them, it is obvious that they originated within the ice
234
ESKERS AND MARGINAL LAKES
before it receded. Eskers characteristically disregard the

underlying topography, which they cross like long railway
embankments, this form being assumed as a result of the
inevitable slumping of the original sides. Their courses,
though winding, are generally aligned more or less at right
angles to the receding ice front. In glaciated lands riddled
with lakes and marshes, like Finland and Sweden, eskers
provide natural causeways across many districts where road
and railway construction would otherwise be difficult. Ridges
that differ from eskers in being short and straight are inter-
preted as the infillings of wide crevasses. They probably
formed within large sheets of stagnant ice which melted away
from the edges of the crevasses as well as from their sides.
Ice-dammed Marginal Lakes
A glacier occupying a main valley may obstruct the mouth

of a tributary valley and so impound the drainage and form
a lake. The Maijelen See, held up in this way by the Aletsch
Glacier (between the Jungfrau and the Upper Rhone) is a
small-scale example, and there are many others, large and
small, in Norway, Iceland (Plate 56a), and Greenland. Where
the ice barrier is sufficiently high and massive the lake rises to
the col or pass at the head of its valley and escapes through
an overflow channel into the valley on the other side. During
the degeneration of an ice sheet into valley glaciers, the
higher ridges of a divide between two neighbouring valleys
may be uncovered, while the ice still extends across the
divide at a lower level. Melt-water then accumulates along
the margin of the ice against the flanks of the hills, and if it
overflows from one side of the ridge to the other, a channel
is cut in the ridge itself, which thus becomes notched.
Notching by marginal overflow channels may be repeated
again and again at successively lower levels while the ice
is retreating.
During the recession of the great Pleistocene ice sheets
enormous numbers of ice-dammed lakes came into existence
235
(cf. some of them being immensely larger than any

Fig. 126),
of the lakes of the present day. They formed wherever a
drainage basin sloped down towards an obstructing ice dam.
Some of these marginal lakes overflowed at successively lower
levels before they eventually disappeared, each stand of the
lake being determined by the height of the lowest outlet
available at any given stage during the wasting away of the
barrier. The features and deposits left behind by such lakes
(thereby providing evidence of their former existence) include :
(a) overflow channels at the outlets, often eroded into con-
Fig. 125
Map of the Parallel Roads of Glen Roy
spicuous valleys and gorges (now dry) situated at the heads

of the valleys from which the lake waters escaped, or across
the ridges and spurs between neighbouring valleys ; (
b ) shore-
line deposits and terraces, formed by the action of waves and
currents ;
(c) deltas deposited by streams flowing into the
lake ;
and (d) lake-floor sediments, including the varved sedi-
ments to be described later.
The most celebrated shore terraces of this kind are the
Parallel Roads of Glen Roy (Plate 56b). These are beaches
about 40 to 50 feet wide which follow the contours at the
levels shown on the map (Fig. 125). Ice extending from Ben
236
MARGINAL LAKES
Nevis across the valleys to the north blocked the entrances
to Glen Roy and Glen Gloy. The highest lake (1,165 feet),
that of Glen Gloy, discharged across the watershed at A into
the Glen Roy lake (1,149 feet) which overflowed into the
river Spey at B. Later, an outlet into a tributary of Glen
Spean was uncovered at C (1,065 feet), and the lake rapidly
drained to the level so determined. This stage, during which
the waters escaped over the eastward-sloping surface of the
Glen Spean ice, lasted until further withdrawal of the ice
allowed the lake to extend along Glen Spean, whence it over-
flowed through D (857 feet), the outlet at its head. In due
course the lobe of ice that blocked the lower part of the valley
Fig. 126
Diagram toshow the isostatic depression of a land surface loaded by a continental
ice sheet, and the consequent development of marginal lakes during the recession
of the ice. Vertical scale and slopes greatly exaggerated
dwindled away sufficiently to allow the lake to drain towards

the sea (Loch Linnhe) and so, finally, to disappear.
During the recession of the continental ice sheets of Europe
and North America conditions were highly favourable to the
development of widespread marginal lakes (Fig. 126). At
the time of maximum extension of the ice the less mountainous
parts of the underlying floor were depressed into a shallow
bowl by the isostatic effect of the load of ice. The thickness
of the ice was of the order 6,000 feet, tapering off towards the
margins. The corresponding subsidence of the crust, where
the load was greatest, would therefore be about 1,600 feet ;
sufficient, that is, to depress vast areas of the rock surface well
below sea level. During the retreat of the ice the crust was
gradually unloaded and isostatic recovery worked in from the
margins, though with a considerable lag. Consequently, for
thousands of years there were large tracts, abandoned by the
237
ice, that sloped towards and beneath the receding ice front.
Many of these became giant lakes, while others were invaded
by the sea. The recovery already achieved since the
isostatic
disappearance of the ice clearly demonstrated by the occur-
is
rence of raised beaches at various heights and by the tilted

attitude of many lake terraces. Moreover, the fact that the
shores of Hudson Bay and the Gulf of Bothnia are steadily
rising even now shows that the process of restoring isostatic
equilibrium is still going on.
Actual depressions within the
areas that were formerly inundated are, of course,
still occupied
by lakes or by the sea. The Great Lakes of North America

and the Baltic Sea are outstanding examples.
From the sediments that accumulated on the floors of the
marginal lakes in Europe the history of the ice recession can
be deciphered with great accuracy almost all the way from
the terminal moraines of Germany to the Scandinavian
glaciers of to-day (Fig. 127). Each spring and summer, as the
ice thawed, the lake in front received a supply of sand, silt,
and clay from the streams that flowed into it. The coarser
material settled down at once, but the finer particles remained
in suspension much longer and were not completely deposited
until much later in the year. But during the late autumn and
winter the glacial streams were frozen and the lake, itself
frozen over, received no further sediment. The still suspended
mud slowly sank to the bottom, forming a thin layer of dark
clay, easily distinguishable from the thicker layer of sandy silt
beneath it. The following year the sediment liberated from
the ice was again sorted out and deposited in two well marked
seasonal layers, sharply separated from the underlying pair.
Each such pair is called a varve, and the sediments characterized
by this annual banding are described as varved clays (Plate 57b).
As this process continued year after year the area of deposit
moved northwards with the receding ice, and the varves thus
became superimposed after the fashion of wedge-shaped tiles
on a roof.
In 1885 Baron De Geer began to count the varves, starting
in Scania at the southern end of Sweden. The thickness of
the deposit at any one locality rarely exceeds 30 feet, but this
238
DATES RECORDED BY VARVES
Fig. 127
Map to illustrate successive stages in the recession of the last European ice sheet,
with dates (in years before 1900 a.d.) established by counting varves
may contain several hundreds of varves, each representing a

single year. Those near the top can then be matched against
the lower varves at a neighbouring locality to the north,
where the sequence includes the varves of the immediately

succeeding years. Thus, by tracing the overlapping varves
through sections and borings at more than 1,500 localities,
and so carrying on the counting bit by bit, De Geer and his
students succeeded, after thirty years of laborious work, in
establishing an absolute chronology of the retreat stages of the
last of the European ice sheets. The Ice Age is conventionally
regarded as having ended about 8,700 years ago, when the
ice sheet reached Ragunda and separated into two isolated
caps (Fig. 127). Thirteen thousand seven hundred years have
elapsed since the ice stood along the site of the Scanian moraine.
From the Brandenburg moraine (marking the culmination of
the ice) to the Scanian moraine the count is necessarily incom-
plete, because the sea now interrupts the sequence, but the
gaps are few and the time of recession can be estimated without
much error at rather more than 11,000 years. It is therefore
approximately 25,000 years since the last European ice sheet
began to disappear. The corresponding estimate in North
America for the time elapsed since the Labrador ice sheet
began its withdrawal from the Long Island terminal moraine
is about 30,000 years.
Lakes : A General Summary

It will already have been gathered that lakes are amongst
the most characteristic features of the landscape of glaciated
regions. Finland is renowned for its innumerable lakes,
35,000 of which have been mapped. Very appropriately, the
—
Finns call their country Suomi the Land of Lakes. Many
parts of Ontario and the neighbouring Provinces and States
are riddled with a comparable network of lakes and waterways.
The extraordinary abundance at the present time of lakes of
glacial origin —they are far more numerous than all other
types put together —
is a result of two circumstances (a) im- :
mense numbers came into existence on the irregular surface

left behind by the retreating ice and ( b they originated so
;
recently that only some of the shallower ones have since been
eliminated.
240
LAKES OF GLACIATED REGIONS
Given a supply of water in excess of the amounts lost by

evaporation or by seepage through the floor and sides, a lake
continues to exist so long as the floor of its basin remains below
the lowest part of the rim. Lakes are therefore conveniently
classified according to the modes of origin of their basins.
Those of glacial origin may occupy :
(a) ice-eroded rock basins in valleys or corries with or

without morainic fringes (pages 219 and 224)
(b) valleys obstructed by morainic barriers (pages 228-29)
c)
(i depressions due to irregular deposition of glacial drift
(page 230)
d kettle holes left by the melting of buried or partly
( )
buried masses of stagnant ice (page 233)

(e) valleys obstructed by ice barriers (page 235)
Among lakes with a more varied history, involving both

glacial erosion and deposition, with modifications in many
cases due to earth movements, the Great Lakes of North
America are the most remarkable. Lake Superior, the largest
freshwater lake in the world, is 1180 feet deep, and the bottom
descends to 580 feet below sea level. The others have depths
ranging from 210 to 870 feet. These immense basins were
primarily gouged out by ice along the sites of broad pre-glacial
valleys. Morainic barriers arranged in loops around their
southern margins further increased their capacity. Isostatic
changes of level, possibly combined with additional crustal
warping due to independent earth movements, brought about
other changes in their outlines and outlets. As the ice receded,
marginal lakes accumulated in the depressions between the ice
wall on the north and the moraines on the south. In successive
stages, illustrated by Figs. 128 to 131, the overflow was, first
Hudson, and finally into the
into the Mississippi, then into the
St. Lawrence. The upper lakes are somewhat delicately
balanced just where lakes might have been least expected
on the watershed between the rivers flowing south to the
Gulf of Mexico and north to Hudson Bay. A gentle tilt of
their basins would suffice to send their waters southward
241
Figs. 128-131. — Successive stages inthe development of the Great Lakes of
North America
Fig. 128
An early stage (about 25,000 years ago), showing marginal lakes with an outlet
near Chicago (C) into the Mississippi. For other lettering see Fig. 131
Fig. 129
The marginal ancestors of Lakes Superior and Michigan drain into the Mississippi.
The eastern lakes drain into the Hudson River
instead of eastward. Isostatic up-tilting towards the north is
still and the rate of movement is such that in

in progress,
about 1,500 years Lake Michigan will again drain to the
Mississippi, unless some counterbalancing process (such as
human interference) intervenes.
To the north-west another series of lakes originated in
242
DEVELOPMENT OF THE GREAT LAKES
Fig. 130
The upper lakes, swollen into the ancestralLake Algonquin, drain, together with
Lake Erie, into a seaway that occupied the St. Lawrence valley and extended
over the site of Lake Ontario. An occasional overflow from Lake Michigan
spills into the Mississippi
Fig. 131
The approach their present-day outlines, the upper lakes draining into the
lakes
dwindling St. Lawrence seaway through the valley of what is now the Ottowa
River. E, Lake Erie F, Finger Lakes of New York State ; H, Lake Huron ;
;
M, Lake Michigan ; S, Lake Superior
much the same way, beginning their history as marginal lakes

which coalesced into a gigantic sheet of water referred to as
Lake Agassiz. The flat-lying sediments of the floor of this
vanished lake form the rich wheatlands of North Dakota and
Manitoba. At first Lake Agassiz drained into the Mississippi,
243
;
but lower outlets were uncovered to the north later on, and
the vast lake was reduced to remnants, including Lake Winni-
peg, Lake Manitoba, and the Lake of the Woods. In Europe
Lakes Ladoga and Onega had a similar origin and history.
Lake basins owing their origin to other geological pro-
cesses are described in the appropriate chapters, but for con-
venience the following summary is added here.
Lake Basins due to Earth Movements

Tectonic depressions are responsible for the largest of the
world’s lakes (Caspian Sea), the deepest (L. Tanganiyka), the
lowest (Dead Sea), and amongst those of notable size the
highest (L. Titicaca, Bolivian plateau), as well as for many
shallow lakes, both large (L. Victoria) and small (local sink-
ings of the ground accompanying earthquakes). In terms of
origin the chief types are due to :
Crustal Warping (L. Victoria, page 431 ;

Lough Neagh) and
the backtilting of valley systems (L. Kyoga, page 432).
Differential Faulting, especially in the African Rift Valleys
(page 432) and in the Great Basin of the Western States (page
423).
Tear Faults across a pre-existing valley, whereby it may be
obstructed by a hill range (L. Joux, Jura Mountains

Fahlensee, Santis Alps).
Lake Basins due to Volcanic Activity
Craters and Calderas of extinct or dormant volcanoes (Crater

Lake, Oregon, Plate 94).
Lava Flows forming barriers across valleys (Sea of Galilee ;
L. Kivu).
Lake Basins due to Deposition of Sediments
Obstruction of valleys and river channels may be brought

about by :
Landslides (page 148) and occasionally by avalanches and

screes.
River Deposits : Delta growth from relatively vigorous side
streams. The sealing off at both ends of abandoned meander
2 44
PLATE 50
L. IlaiL'kcs
(A. Ice-dammcd lake on side of the Hoffcll Glacier, Hornafjord, south-
east Iceland
( rcncn Copyright Reserved. [II.M heot. Siinw

I he Paiallel Roads of Glen Roy, Inverness-shire
PL \TE 57
[//..V. (jt'ol. bmvev

(A) Moraines and moraine-dammed lochans. Valley of Ail t a'Choire
Mhor, Ross-shire
THE PLEISTOCENE ICE AGE
loops (ox-bow lakes, page 165). Levee building in general

(flood plain lakes and swamps and delta lagoons, page 168 and
Fig. 79). Norfolk Broads (pages 194-95).
Glacial Deposits (see above).
Wind Deposits : Coastal sand dunes enclosing lagoons and
marshes (as in the Landes of south-west France).
Marine Deposits : Closed bars and barrier beaches enclosing
coastal lagoons (page 299).
Lake Basins due to Denudation
It should not be overlooked that the depressions occupied

by an immense number of barrier lakes are river valleys that
have been locally obstructed. The following processes are
directly responsible for the excavation of basins :
Solvent Action of Ground-water Swallow holes, of which the

outlets have been clogged by residual clays. Surface subsi-
dences due to underground solution of limestone (page 134) or
of rock salt (the meres of Cheshire).
Solvent Action of Rivers : Expansion and deepenings of river
beds by surface solution of limestone (some of the Alpine lakes
Lough Derg, an expansion of the River Shannon).
Glacial Erosion (see above).
Wind Deflation : Hollows excavated in arid regions to a
depth where an adequate supply of ground-water is tapped.
The Pleistocene Ice Age
The maximum extent of the Pleistocene ice sheets in the

Northern Hemisphere is shown in Fig. 132. The total area
overwhelmed by ice approached eight million square miles.
Half of this was in North America, where the ice radiated from
three main centres Labrador, the Keewatin district border-
:
ing Hudson Bay, and the Cordilleran ranges of the west. The
European ice continuation beyond the Urals covered
and its
nearly three million square miles. A

subsidiary ice cap radiated
from the Alps, and the Himalayas and other ranges of Asia
were similarly glaciated. It is remarkable that some of the
(.396)
245 17
Map showing the maximum extent of the Pleistocene ice in the Northern
Hemisphere. (After E. Antevs)
Arctic islands and alsomuch of Siberia [cf. page 204 ) appear to

have escaped general glaciation. In recent years, however,
Soviet geologists have been finding increasing evidence that
the ice spread over wider areas than those shown on Fig. 132 .
In the Southern Hemisphere the Kosciusko plateau of New

South Wales and a considerable part of Tasmania were glaci-
ated. New Zealand, where glaciers still persist, was largely
shrouded in ice, and so were extensive tracts of Patagonia and
southern Chile. The Antarctic sheet, like that of Greenland,
was thicker and more extensive than it is to-day. In Central
Africa moraines occur more than 5,000 feet below the ice that
still remains on Ruwenzori and the higher volcanic peaks (see
Plate 45 ). The climatic changes evidently led to a general

lowering of the snow line, and were clearly world-wide in their
effects.
Detailed studies of the drift deposits have shown that
246
THE FOUR ALPINE GLACIATIONS
glaciation was not confined to the effects of a single great
advance and retreat of the ice. In many places around the
Alps four successive sets of boulder clays, moraines, and outwash
gravels are preserved, showing that there was a fourfold re-
petition of the glacial cycle. The fourth and latest cycle, of
which naturally most is known, appears to have involved one
or two minor advances and retreats before the final withdrawal
to present conditions. The intervals between the major glacial
— —
phases known as interglacial stages are represented by ancient
soils, and locally by lake and river deposits or screes, sand-
wiched between older and younger beds of glacial or glacio-

fluvial drift.
Some of the interglacial beds contain plant remains, and
so provide valuable information as to the climatic conditions
at the times when they were deposited. A consolidated scree,
formed during the third interglacial stage, and still preserved
as a thick wedge of breccia near Innsbruck, contains fossil
leaves and other relics of trees that no longer grow in the Alps
but only in warmer regions. At that time the average tem-
peratures in the Alpine region must have been higher than now,
and quite possibly it was then too warm for glaciers to be
nourished at all. From similar lines of evidence elsewhere it
appears that the first interglacial stage was also relatively
warm, but that the second was cooler, with a climate com-
parable to that of the present.
The four major Alpine glaciations and the intervening
known by the names listed in the follow-
interglacial stages are
ing table, the names being those of places where the glacial
deposits so distinguished are well exposed. Attempts have
been made to estimate the durations of the interglacial stages
by comparing the effects of weathering and erosion on the
underlying deposits in each case. The depth reached by
weathering during Riss-Wiirm time is found to be about three
times that achieved on similar but later deposits exposed
during post-Wiirm time. Since the latter is about 25,000
years, it follows that the Riss-Wiirm interval cannot have been
less than 75,000 years. Indeed, it may well have been con-
siderably longer, because the second foot of weathering other—
247
Comparable
Major GLACIAL and Relative durations Corresponding estimates
climatic Interglacial estimated from depth periods
from Iowa,
cycles Stages (Alps) of weathering in years
U.S.A.
Post- Wurm 1 representing 25,000 25,000

< t WURM ? 30,000
, f Hiss- Wurm 3 75,000 120,000
3 0 7
\ RISS
0 r Mindel-Riss 12 300,000 300,000
“
1
MINDEL 7 ?
GiinzrMindel 3 75,000 200,000
*< GUNZ ? ?
!
things being equal — takes considerably longer than the first,
and so on. The figures in the table are therefore to be regarded

as minima. Estimates from America based on the same com-
parative method are given in the last column. Allowing for
the very considerable margin of uncertainty, it can hardly be
less than a million years since the first great ice sheet began its
advance— and it may well be more.

What of the future? Reference to Fig. 133 shows that
during the latest chapter of the earth’s history there have
already been four major climatic cycles. Clearly we can have
no assurance that these remarkable oscillations have come to
an end. The persistence of the Greenland and Antarctic
ice sheets shows that, strictly speaking, the Ice Age is still with
us. But even though the former severity has relaxed, it is
important to realise that present conditions are themselves,
geologically speaking, very unusual. Ice ages have been quite
exceptional during the earth’s history, and it is to be expected
that sooner or later there will be a return to the milder and
more normal climates that have characterised most of the
geological past. Ice sheets will then be banished from the
globe.
Meanwhile — apart from the remote possibility that present
conditions will persist indefinitely —the future holds out the
prospect of one of two alternatives. It may be that we are
now living in the early years of a cool interglacial stage, com-
parable with that of the Mindel-Riss, to be followed by yet
248
THE PROBLEM OF ICE AGES
> 75,000 >300,000 years > 75,000 26,000 '
M-R
Cool Interglacial
Gun! M i n d e I
Time Present Day'
Fig. 133
Diagrammatic representation of the climate cycles in the Alps during the Plei-
stocene, showing the four major ice advances and the intervening interglacial
stages
another glaciation which, this time, would relentlessly oblite-

rate many On the
of the most cherished homes of mankind.
other hand, the present may be merely a pause during the slow
restoration of a more genial climate under which the existing
ice-sheets will melt away. If this happens, the liberated melt-
waters w’ill raise the level of the oceans by about 100 feet.
London and New York and all other ports and lowland cities
throughout the world will then be submerged, and obliged to

migrate to higher sites. Either alternative is sufficiently alarm-
ing to justify the most careful study of present-day climatic
tendencies, and of the possible causes of ice ages.
Suggested Causes of Ice Ages
If we knew how to explain the strange vicissitudes of climate

that have occurred in the past, we might be able to forecast
future developments more
But so far this highly
confidently.
complex problem remains unsolved, although many plausible
hypotheses have been proposed. Only a brief outline of the
more important suggestions can be given here.
Changes in the Bodily Movements of the Earth, e.g. in the eccen-
tricityof the earth’s orbit, and in the inclination of the earth’s

axis to the ecliptic. These are real periodic changes, and
since they must be accompanied by correspondingly periodic
climatic variations, they must have a genuine bearing on the
climatic problem. If, however, the variations reached ex-
249
tremes sufficient to account for the glaciations of the Pleisto-

cene, there should have been hundreds of similar ice ages since
the Cambrian period. Actually, there has been only one other
comparable case during the last five hundred million years—at
the close of the Carboniferous period (see page 499). It may
therefore be concluded that, by themselves, these changes are
inadequate.
Changes in the Position of the Poles relative to the Continents or

vice versa. — It is generally agreed that movements of the poles
on a scale sufficient to bring about important climatic changes
are highly improbable. The
earth behaves like a gyroscope
and only very slight changes in the position of the axis are
dynamically possible. It is much less improbable that the
outer crust may have moved over the interior, or that the
continents may have changed their positions relatively to one
another. Continental drift (discussed in Chapter XXI) might
account for the distribution of certain glaciated areas, particularly
in the late Carboniferous, when India and Central Africa were
amongst the glaciated lands but neither continental drift
;
nor polar wandering begins to explain how these areas came

to be glaciated at all. Polar ice caps are exceptional features
in the earth’s history, and a region does not automatically
receive a shroud of ice merely because it happens to lie over
or near one of the poles.
Changes in Topography. — Mountain growth and continental

uplifts have occurred during geological time, and
at intervals
if such elevated regions rose above the snow line glaciation
would obviously be favoured. Moreover, changes in topog-
raphy, and in the distribution of land and sea, affect the cir-
culations of moisture-laden winds and of oceanic currents,
both of which affect in turn the height of the snow line and
climatic conditions generally. Here, then, we have an im-
portant contributory factor, but certainly not a sufficient one
in itself, for there is no correlation between the times of maxi-
mum elevation and those of widespread glaciation. Con-
versely, the late Carboniferous ice sheets developed over lands
that were for the most part low-lying.
2=^0
.
UNCONFIRMED HYPOTHESES
Changes in the Composition of the Atmosphere . —The carbon
dioxide of the air absorbs a small amount of the heat reflected
from the lands. It has therefore been suggested that increase
in the proportion of atmospheric carbon dioxide would lead
to a rise of temperature, and decrease to a fall of temperature.
The effect is very slight, and is almost wholly offset by com-
plementary effects due to water vapour. Appeal has also been
made to the fact that after paroxysmal volcanic eruptions ( e.g
Krakataoin 1883 page 470 ) the dust blown into the air reflects
,
back some of the sun’s rays, and so brings about a temporary

drop of temperature. There is, however, no correlation
between times of prolonged volcanic activity and times of
glaciation.
Changes in the Sun’s Radiation. —Small-scale variations in the

the amount of heat received from the sun are known to occur,
especially in connection with the sun-spot cycle. This fact
has stimulated what is perhaps the most promising of all the
ideas so far proposed : the hypothesis that there may be large-
scale fluctuations in the sun’s radiation, recurring at very long
intervals. When we consider that the earth has orogenic
cycles of 200 million years or so, reflecting the behaviour of
its it seems not unreasonable to suppose that the sun
interior,
may be subject to long-period disturbances arising from
also
within it. The difficulty about this hypothesis is not that it is
inadequate, but that it cannot be tested and therefore cannot
be either proved or disproved. It may very well be true, but
for the present it remains no more than an attractive specu-
lation.

A. E. H. Tutton
The Natural History of Ice and Snow. Kegan Paul, Trench, Trubner
and Co., London, 1927.
W. H. Hobbs
Characteristics of Existing Glaciers. Macmillan Co., New York, 1911.
251
C. A. Cotton
Climatic Accidents in Landscape- Making. White ombe and Tombs,
Christchurch, New Zealand, 1942.
F. E. Mathes
Geologic History of the Yosemite Valley. United States Geological
Survey, Professional Paper 160, Washington, 1930.
E. G. Woods
The Baltic Region. Methuen, London, 1932.
A. P. Coleman
Ice Ages : Recent and Ancient. Macmillan, London, 1926.
W. B. Wright
The Quaternary Ice Age. Macmillan, London, 1936.
C. E. P. Brooks
Climate through the Ages. Benn, London, 1926.
R. A. Daly
The Changing World of the Ice Age. Yale University Press, New Haven,
Conn., 1934.
G. C. Simpson
World Climate during the Quaternary Period. Quarterly Journal of the
Royal Meteorological Society, Vol. LX., pp. 425-78, 1934.
252
Chapter XIII
WIND ACTION AND DESERT LANDSCAPES

The Geological Work of Wind
As an agent of transport, and therefore of erosion and de-
position, the work of the wind is familiar wherever loose surface
materials are unprotected by a covering of vegetation. The
raising of clouds of dust from ploughed fields after a spell of
dry weather and the drift of wind-swept sand along a dry
beach are known to everyone. It is with this direct action on
the land surface that we shall be concerned in this chapter.
But it should not be overlooked that the wind also distributes
moisture over the face of the earth, and is therefore one
of the primary factors responsible for weather and weather-
ing and for the maintenance of rivers and glaciers. More-
over, by transferring part of its energy to the surface water
of the sea, the wind is responsible for waves and their erosive
work.
In humid regions, except along the seashore, wind erosion
is limited by the prevalent cover of grass and trees and by the
binding action of moisture in the soil. But the trials of desert

warfare have made it hardly necessary to stress the fact that
in arid regions the effects of the wind are unrestrained. The
“ scorching sand-laden breath of the desert ” wages its own
war of nerves. Dust storms darken the sky, transform the air
into a suffocating blast, and carry enormous quantities of
material over great distances (Fig. 134). Vessels passing
through the Red Sea often receive a baptism of fine sand from
the desert winds and, beyond the other end of the Sahara,
;
dunes have accumulated in the Canary Islands from sand

blown across the intervening sea. Red dust sometimes falls
in Italy, and even in Germany, carried north by great
storms from the Sahara. The “ blood rains ” of Italy are
due to this same red dust washed down by rain from the
hazy atmosphere.
253
Dust storm approaching Port Sudan
By itself the wind can remove only dry incoherent deposits.

This process of lowering the land surface is called deflation (L.
deflare, to blow away). Armed with the sand grains thus
acquired, the wind near the ground becomes a powerful
scouring or abrading agent. The resulting erosion is described
as wind abrasion. By innumerable impacts the grains themselves
are gradually worn down and rounded. This third aspect of
wind erosion, the wear and tear of the “ tools,” is distinguished
as attrition.
The winnowing action of the wind effectively sorts out the
transported particles according to their sizes. This is well
illustrated in the desert wherever mixed deposits of gravel,
sand, and mud are worked over by the wind. Such materials
are continually liberated by weathering and abrasion from
bedrock surfaces, but far larger supplies are furnished by
sheets of poorly assorted alluvium spread over the desert floor
by the occasional torrents that flush out the wadis. Particles
of silt and dust are whirled high into the air and transported
254
WORK OF THE WIND
far from their source, accumulate beyond the desert as

to
deposits of loess (page 267). Sand grains are swept along near
the surface, travelling by leaps and bounds, until the wind drops
or some obstacle is encountered. The dunes and other accumu-
lations of wind-blown sand thus come to be composed of clean
and uniform grains, the finer particles having been sifted out
and the larger fragments left behind. It follows that pebbles
and gravel are steadily concentrated on the wind-swept sur-
faces of the original mantle of rock-waste.
As a result of wind erosion, transport, and deposition three
distinctive types of desert surface are produced :
(1) The rocky desert (the hammada of the Sahara), with a sur-
face of bedrock kept dusted by deflation and smoothed by
abrasion (Plate 61 a) ;
(2) the stony desert, with a surface of gravel (the reg of the
Algerian Sahara), or of pebbles (the serir of Libya and
Egypt) and
;
(3) the sandy desert (the erg of the Sahara).
Complementary to these is the loess of the bordering steppes,

deposited from the dust-laden winds that blow from the desert.
Wind Erosion
The most serious effects of wind deflation —from the human
point of view — are experienced in semi-arid regions like the
Great Plains west of the Mississippi, where in recent years vast
quantities of soil have been blown and washed away from
thousands of formerly productive wheat-growing farms. Orig-
inally an unbroken cover of grass stabilised the ground, but
long-continued ploughing and over-exploitation finally de-
stroyed the binding power of the soil and exposed it as a loose
powder to the driving force of the wind. This national menace
became critical during a period of severe droughts, culminating
in 1934-35, when great dust storms originating in the “ Dust
Bowl ” of Kansas swept over the States towards the Atlantic.
255
Rainwash and the creeping disease of badland erosion ex-
tended the devastation. Widespread measures of reclamation
and protection have been undertaken to minimise the growing
wastage and desolation.
A characteristic result of deflation, especially over regions
where unconsolidated clays and friable shales are exposed ( e.g
in the North African, Kalahari, and Mongolian deserts), is the
production of wide plains and basin-like depressions. The
excavation of hollows is limited only by the fact that even
in deserts underground water may be present. Once the
desert floor has been lowered to the level of the ground-
water, the wind can no longer pick up the moistened particles.
The base level for wind action is that of the water table,
which may be far below sea level. The “ pans ” of South
Africa and the Kalahari and the more impressive depressions
and oases of Egypt and Libya have all been excavated by
ablation.
Westward from Cairo to Jarabub there is a remarkable
series of basins with their floors well below sea level (Fig. 135),
reaching — 420 feet in the salt marshes of the immense Qattara
depression. Some of the smaller basins tap a copious supply of
ground-water at depths of —50 to — 100 feet, and have become
fertile oases. To the north the surface rises by abrupt escarp-
ments to terraced tablelands formed of hard sandstones and
limestones which formerly extended across the softer rocks of
the depressions. To the south, following the direction of the
prevailing wind, long stretches of sand dunes represent part
of the removed materials. The other well-known oases of
—
Egypt Baharia, Farafra, Dakhla, and Kharga are above —
sea level, but they have originated in the same way. All of
them have steep escarpments of resistant rock to the north,
and Baharia is entirely enclosed within rocky walls. The
floors consist of the same soft strata as those found at the base
of the escarpments. The depressions are not crustal sags, nor
have they been cut by water, for the intermittent floods due
to rare cloudbursts tend to fill them up with debris. The wind
has been the sole excavator.
The effects of wind abrasion are unmistakably expressed
256
WIND EROSION
Fio. 135
Map showing depressions, sand wastes, and lines ol sand dunes in the Egyptian
Desert
in the forms and surfaces of the desert bedrocks. Just as an

sand blast is used to clean and polish building stones
artificial
and to etch glass, so the natural sand blast of the wind attacks
destructively everything that lies across its path. Cars driven
against wind-blown sand may have their wind-screens frosted
and their paint scoured off. The action on exposed rocks is
highly selective. Like a delicate etching tool, the sand blast
picks out every detail of the structure. Hard pebbles, nodules,
and fossils are left protruding from their softer matrix until
they fall out. Variably cemented rocks are fritted and honey-
combed like fantastic carvings. Where there are thin alter-
nations of hard and soft strata the soft bands are scoured away
more rapidly than the hard, which thus come to stand out in
strong relief, like fluted shelves and cornices with deep grooves
between. Where the wind blows steadily in one direction
257
over strata of this kind, especially if the beds are tilted rather
than flat, the softer materials are excavated into long passage-
ways (parallel to the dominant wind direction) separating
deeply undercut, overhanging ridges. Such fantastically
carved “ cockscomb ” ridges are common in certain parts of
the Asiatic deserts, where they are called yar clangs.
Undercutting is everywhere a marked feature of wind
abrasion, owing to the fact that the process is most effective

just above the surface, where the sand is most abundant.
Telegraph poles in sandy stretches of desert have to be protected
by piles of stones against the cutting action of the sand grains
hurled against them. Along the base of escarpments alcoves
and small caverns may be hollowed out. As always, the effect
of undercutting on slowly weathered formations is to maintain
steep slopes. Joints are readily attacked and opened up and
these commonly determine the outlines of rock towers and
pinnacles, left isolated like detached bastions in front of
the receding wall of an escarpment (Plate 64). Upstanding
rocks attacked at the base by winds from varied directions
pass through a stage in which they resemble mushrooms,
especially when a resistant cap-rock overlies a weaker
formation.
Where the bedrock of the desert floor
is exposed to blown
sand it maybe smoothed or pitted or furrowed, according to

its structure. Compact limestones become polished, massive
granites are smoothed or pitted, and gneisses and schists are
ribbed and fluted parallel to their foliation. Where pebbles
have become sufficiently concentrated by removal of finer
material they become closely packed, and in time their upper
surfaces are ground flat. In this way mosaic-like tracts of
desert-pavement are developed. Isolated pebbles or rock frag-
ments strewn on the desert surface are bevelled on the wind-
ward side until a smooth face is cut. If the direction of the
wind changes seasonally, or if the pebble is undermined and
turned over, two or more facets may be cut, each pair meeting
in a sharp edge. Such wind-faceted pebbles, which often
resemble Brazil nuts, except that their surfaces are polished,
are known as dreikanter or ventifacts (Fig. 136).
258
WIND-WORN PEBBLES AND SAND GRAINS
Fio. 136
Pebbles faceted by sand blast (dreikanter or ventifacts). Drawn by M. V. Binosi
from specimens in the Cairo Geological Museum. (From IP. //. Hume, “Geology
of Egypt," Vol. I., by permission)
As a result of continual attrition due to the friction of

rolling and impact the sand grains themselves are gradually
worn down and rounded. The prolonged action of wind is
far more effective in rounding sand grains than that of running
water, because of (a) the greater velocity of the wind (b) the ;
greater distances traversed by the grains as they bound and

roll and collide with each other backwards and forwards across
wide stretches of desert and (c) the absence of a protective
;
sheath of water. Some of the millet seed sands of the desert are
almost perfect spheres with a mat surface like that of ground
glass. It is also noteworthy that visible flakes of mica, such as
are commonly seen in water-deposited sands and sandstones,
are very rare in desert sands and dunes. The easy cleavage
of mica facilitates constant fraying during the wear and tear
of wind action. Mica is thus reduced to an impalpable powder
that is winnowed away from the heavier sand grains. These
259
contrasts between water-laid and aeolian sands are of great
value in deciding whether ancient sandstones have been
formed in deserts or under water. The Penrith Sandstone of
the Eden valley is a well-known example of a Permian desert
sand. Its rounded grains, the absence of mica, and the cross-
bedding of the formation all testify to the desert conditions of
the time.
Coastal Dunes and Sandhills
Along low-lying stretches of sandy coasts and lake shores,

where the prevailing winds are onshore, drifting sand is blown
landwards and piled up into dunes which form a natural
bulwark of sandhills. Any mound or ridge of sand with a crest
or definite summit is called a dune. Deposition begins wherever
the force of the wind is broken by obstructive irregularities of
the surface, including grasses and trees. In humid regions the
Structure of Stationery Dune Structure of Migrating Dune
Fig. 137
Sections to illustrate the growth, migration, and structures of sand dunes. A
stationary dune, A, grows in height with a forward and upward advance of the
crest. When the sand supply and wind velocity involve migration of the dune,
the crest advances to positions such as b, c, . . . g, and H
conditions governing growth and removal are very complex.
The wind varies in strength and direction. Vegetation and
moisture tend to fix the sand, but fixation is often incomplete.
During severe gales old dunes may be breached and scooped
out into deep “ blow outs.” The resulting confused assemblage
of hummocks and hollows gives coastal sandhills a character-
istically chaotic relief.
An ideal dune has a long windward slope rising to a crest

and a much steeper leeward slope (Fig. 137). The latter is
determined by the fact that sand blown over the crest falls
into a wind shadow, and comes to rest at its natural angle of
—
repose about 30° to 35° for dry sand. The windward slope
260
PLATE 5S
('renin Cnpyuyht Kcsenvd] 7/..V. Goa. Smvcy

! A A stage in the forward drift [from right to left’ of the Culbin Sands,
Moray
[H.M. Gcol. Survey

(B Maviston Sandhills (•south-west continuation of the Culbin Sands),
partially arrested by bent. Moray
Jl.M. Gcol. Survey
W . > |
Remains of a plantation formerly bulled by band which lias now

drifted farther on
MAVIS! OX SANDHILLS, MORAY

MIGRATION OF SAND DUNES
isoften beautifully rippled (Plates 5a and 60a). In situations

where dunes are not effectively arrested by vegetation, or kept
within bounds by winds from opposing quarters, they slowly
migrate in the direction of the prevailing wind. When the
wind is not fully loaded with newly acquired sand it sweeps
up more from the windward slope, and drops it over the crest,
where it streams down the “ slip-face.” By subtraction of
sand from one side and addition to the other the dune travels
forward.
As one belt of dunes is driven inland away from the beach,
another arises in its place “ so that a series of huge sandy
billows, as it were, is continually on the move from the sea-
margin towards the interior ” (Plate 58a). Very wide belts

of sandhills have spread inland in this way along the coasts
of Holland and North Germany and in the Landes of Gascony
adjoining the Bay of Biscay. The Culbin sandhills, near the
mouth of the Findhorn on the southern shore of the Moray
Firth, furnish a classic example of the destruction of cultivated
lands and habitations by advancing sand. Prior to 1694 the
sandhills had already reached the fringe of the Culbin estate.
In that year a great storm started a phase of accelerated en-
croachment which finally led to the complete obliteration of
houses, farms, and orchards, and even to the burial of fir
plantations (Plate 59).
In many
threatened regions measures have been taken to
restrain the advance of the dunes. Tough binding grasses
such as bent or marram are excellent for this purpose (Plate
58b). The harsh tufts check the wind, trap oncoming supplies
of sand, and continue to grow outwards as the entangled sand
accumulates, leaving behind them an intricate network of
long roots. Such protected dunes become levelled up and
turfed over. Subsequent growth of the dunes then tends to be
seawards. Where the problem is less easily solved, as in the
Landes, the dunes are anchored more securely by starting
plantations of conifers on the landward side, and gradually
extending them across the sand already partially fixed by
grasses.
Desert Dunes and Sand Sheets

About a fifth is desert, and on an average
of the land surface
about a fifth mantled with sand. A high
of the desert areas is
proportion of the desert floor is an erosion surface of bedrock,

locally strewn with coarse rock-waste (Plate 61a). Regions
of shale and limestone provide little or no sand, but wdiere
sandstone is being disintegrated or mixed alluvium is being
deflated, the wind picks up the loose grains and concentrates
them into vast sand wastes (Plate 60) and long chains of
dunes.
Complications due to vegetation and moisture arise only
around oases or in the transition zones where the desert merges
Fig. 138
A typical barchan (drawn from a photograph)
into steppe or savanna country. In the heart of the desert the

wind has But nevertheless, as a masterly study by
free play.
Brigadier Bagnold has made abundantly clear, the factors
controlling the form of the sand accumulations are far from
simple. They include the nature, extent, and rate of erosion
of the source of supply the sizes of the sand grains and
;
associated fragments the varying strength and direction of

;
the wind ; and the roughness or smoothness of the surface

(e.g. the presence or absence of pebbles) across which the sand
is drifted and deposited. Of the resulting sand forms four main

types can be distinguished :
(a) Sand drifts caused by protruding rocks or cliffs. These

call for no further mention.
(
b) or barchans (a Turkestan
Crescentic dunes name w'hich has
been generally adopted), w'hich occur as isolated units
(Fig. 13S), either sporadically or in long chain-like
262
BARCHANS AND SEIF DUNES
swarms, or as colonies, more or less linked together

laterally, which advance across the desert like gigantic
but irregular ripples.
(c) Linear ridges or longitudinal dunes (known as seifs in the
Sahara), which commonly occur in parallel ranges of
immense length, each diversified by peak after peak
“ in regular succession like the teeth of a monstrous
saw.”
d) Sand sheets of wide extent, which
(i may be flat or un-
dulating (Plate 60 and Fig. 141).
Dunes arise wherever a sand-laden wind piles up sand on

the slopes of a random patch. The mound grows in height
until a “ slip-face ” is established by avalanching on the
sheltered leeward side. As the dune migrates, the extremities,
offering less resistance to the wind than the summit region,
advance more rapidly, until they extend into
wings of such a
length that their total obstructive power becomes equal to
that of the middle of the dune. The resulting crescentic form
then persists with only minor modifications of shape and size,
so long as the wind blows from the same quarter. The width
of a barchan is commonly about a dozen times the height,
which ranges up to a maximum of 100 feet or so. Winds
blowing continually from nearly the same direction are
essential for the growth and stability of barchans. Under such
conditions, and given a sufficient supply of sand, elongated
swarms of barchans march slowly forward, like a stream of
vehicles in aone-way street (Fig. 139). The rate of progress
varies up to 20 feet a year for high dunes, and up to 50 feet
a year for small ones.
Where the prevailing wind is occasionally interrupted by
strong cross winds which drive in sand from the sides, the
conditions are like those of a one-way street which becomes
densely crowded and choked by the inflow of traffic from every
cross-road. Instead of a chain of barchans, a long seif dune is
developed, a high, continuous, but serrated ridge parallel to
the direction of the prevailing wind, and in some examples
running dead straight for a hundred miles or more. In South
263
Fio. 139
Plan of a procession of barchans in the Libyan Desert
Iran such ridges occasionally reach a height of 700 feet. In

Egypt 300 feet is common. S.S.E. of the Qattara depression
there is a long tract of many parallel seifs with corridors of
bare desert floor between (Fig. 135). One of the most re-
markable features of desert dunes is their apparent power of
Fig. 140
Diagram to illustrate the shepherding effect of wind on sand ridges. The
wind is strongest between the ridges, and is retarded by friction against them.
Eddies are therefore set up as shown
collecting all the sand in their neighbourhood. The explana-

tion appears to be that the wind
exerts a shepherding effect.
If the surface between the dunes is fairly smooth, the drag on
wind there is less than it is along the edge of the dune. Eddies
are thus set up which blow towards the dune, and so keep the
intervening surface swept clean (Fig. 140).
264
3 **.
(A'i In the sand-sea of the Algerian Sahara

IN THE ALGERIAN SAHARA
Fig. 141
A view of an Egyptian “ sand sea.” Drawn by M. V. Binosi.

characteristic
{From W. H. Hume, ‘‘ Geology of Egypt,” Vol. by permission)
In the great Egyptian Sand Sea (Fig. 141), however,

another factor operates. Here the dunes, which include
seifs in some parts, and colonies of irregular barchans in
others, rise above a platform of coarse sand. This has

accumulated around their bases because the grains are too
large to accompany the finer sand which is carried up
the slopes of the dunes. Wide, featureless sand sheets
accumulate where the wind disperses chance mounds of
sand, instead of developing them into dunes. Here the
effect of sprinkled pebbles comes in. A pebbly surface
increases the drag on the wind to such an extent that
the velocity near the ground is less than it is over a
patch of clean sand. The resulting eddies therefore blow
from the patch towards the pebbly surface until the sand is
again evenly distributed between the pebbles. Widespread
sand sheets are also characteristically developed on the borders
of deserts, where a scanty vegetation diversifies the surface
265
[6. C. Gilbert, V.S. Geol. Survey
Fio. 142
Desert scrub, with basalt-capped mesa behind, N.E. of The Solitario, Texas
(Fig. 142 and Plate 63b). Grass and scrub break the force
of the wind and inblown sand is evenly distributed.
One of the most curious phenomena associated with dry
dune sands is the mysterious booming which suddenly inter-
rupts the silence in certain desert localities. Bagnold writes,
“ Native tales have woven it into fantasy sometimes it is the
;
song of sirens who lure travellers to a waterless doom some- ;
times it is said to come upwards from bells still tolling under-

ground in a sand-engulfed monastery or maybe it is merely
;
the anger of the jinn But the legends, as collected by the late
!
Lord Curzon, are hardly more astonishing than the thing itself.
I have heard it in south-western Egypt 300 miles from the
nearest habitation. On two occasions it happened on a still
night, suddenly — a vibrant booming so loud that I had to
shout to be heard by my companion. Soon other sources set
going by the disturbance joined their music to the first.
This weird chorus went on for more than five minutes con-
tinuously before silence returned.” “ Roaring ” sands have
266
LOESS
been described from a place in the Kalahari desert, where a

long tongue of white sand is advancing across a red sandy
plain. When the white sand is disturbed by walking or sliding
down its steep face, the sand sets up a loud roaring, recalling
the starting of an aeroplane engine. Despite much investiga-
tion and experiment no satisfactory explanation of the sounds
is yet forthcoming. In some way an avalanching layer of
sand sets all the grains vibrating together. An analogy is
suggested by the sound given out when one corrugated book
cover is drawn over another.
Loess
We must next consider what happens to the vast quantities

of dust which have been winnowed from the desert sands and
exported by the wind. From the deserts of Asia the wind
carries the dust to the south and south-east over the grassy
regions of China, where it is deposited and accumulated as a
thick blanket of loess (Fig. 79). From the North African
deserts much of the dust reaches the Atlantic, either directly
or by way of the Congo basin, where it is retained for a time
in the soil. Of the rest, some is trapped by the Mediterranean
and Red Seas, while the balance is carried towards the steppes
of Russia and Turkestan. Another very important supply of
fine and dust was formerly provided by the rock flour of
silt
glacial deposits. During and after the retreat of each of the

successive ice sheets, the finer material was sifted out by the
wind and deposited over the steppe-like country outside the
glaciated regions (Fig. 143). Thus it happens that a long belt
of loess, mainly composed of glacial material in the west and
of desert material in the east, stretches from France to China.
The term loess comes from a town of that name in Alsace.
Beginning as straggling belts of hills in France and Germany,
the deposit becomes thicker and more extensive as it is traced
across Russia and Turkestan, until in Shansi and the adjoining
provinces of China it reaches its maximum development.
Loess is an accumulation of wind-borne dust and silt,
washed down from the air by rain, and retained by the pro-
267
Fig. 143
Map showing the distribution in Europe of loess and loamy soil formed
from loess (dotted), and its marginal relation to the last European ice sheet.
(After S. von Bubnoff)
Each spring the grass

tective grip of the grasses of the steppe.
would grow a higher on the material collected during the
little
previous year, leaving behind a ramifying system of withered

roots. Over immense areas many hundreds of feet have
accumulated, whole landscapes having been buried, except
where the higher peaks project above the blanket of loess.
The material itself is yellow or light buff, very fine grained,
and devoid of stratification. Although it is very friable and
porous, the successive generations of grass roots, now repre-
sented by narrow tubes partly occupied by calcium carbonate,
make it sufficiently coherent to stand up in vertical walls which
do not crumble unless they are disturbed. The passage of
traffic along country roads loosens the material, clouds of dust
are removed by the wind, and the roads are worn down into
steep-walled gullies and miniature canyons.
In the loess provinces of China rain and small streams
carve the surface into a maze of ravines and badland topo-
graphy (Plate 63a). The larger rivers flow in broad and fertile
alluvial plains bordered by vertical bluffs. Here and in the
lowland and deltaic plains to the east most of the alluvium is
268
DESERT WEATHERING
simply loess redistributed by water. In the loess uplands

cultivation of the slopes is by terracing. The
made possible
steep-sided cliffs andwhether natural or artificial, are
walls,
often riddled with cave dwellings, many of which have their
chimneys opening into the fields above. This mode of habi-
tation has occasionally led to great disasters. In 1556, for
example, widespread landslides and floods were started by a
catastrophic earthquake and nearly a million peasants lost
their lives.
In the semi-arid regions of the western States and in the
Mississippi valley there are thick deposits of adobe which cor-
respond in all essentials to the loess of Europe and Asia. Here
again the material has been sifted by the wind from the rock
waste of deserts and glaciated areas, and blown far outside
the regions of supply, to find lodgment on surfaces, like the
prairies, protected by vegetation. Similar deposits also occur
in the pampas of South America.
Weathering and Stream Work in the Desert
The Sahara and the other sub-tropical deserts lie between

the equatorial and temperate zones along belts of high atmos-
pheric pressure where the prevalent descending winds are dry.
The deserts of higher latitudes occur in regions where the rain-
fall is low either because of the great distance from the oceans
e.g the Gobi of Central Asia) or because of the rain barrier

(i
interposed by high mountains near the coast (e.g. the deserts

of N. and S. America). In all these arid regions the rainfall
is rare and sporadic, both temperature and wind intensity are
subject to violent daily and seasonal fluctuations, and vegeta-

tion extremely scanty or entirely lacking.
is
Under these conditions mechanical weathering is dominant,

involving the splitting, exfoliation, and crumbling of rocks by
alternations of scorching heat and icy cold (Plate 23b). Never-
theless chemical weathering, though extremely slow, plays a
far from negligible part. By decomposition and solution, rocks
that would otherwise successfully resist the stresses set up by
269
temperature changes are gradually weakened until they can
be shattered. By evaporation minute quantities of dissolved
matter are brought to the surface. The loose salts are blown
away, but oxides of iron, accompanied by traces of manganese
and other similar oxides, form a red, brown, or black film
which is firmly retained. The surfaces of long-exposed
“
pebbles thus acquire a characteristic coat of desert
rocks and
varnish.”
Although most desert localities remain entirely unvisited
by rain for years on end, no part can be regarded as per-
manently free from rain. Over a representative period of
years the rainfall is limited to an average of a few inches per
annum, 10 inches being the limit, reached only on the desert
margins. In the adjoining semi-arid regions the annual rain-
fall rises to 10-20 inches, but even here long periods of drought
are usual.
Evaporation far exceeds the rainfall in the desert. Per-
manent streams cannot originate under such conditions, al-
though well nourished rivers, like the Nile, with adequate
sources in humid regions, may cross the desert without entirely
dwindling away. Outflowing streams are otherwise short and
intermittent, and confined to coastal districts where, moreover,
the rainfall is less scanty. The desert drainage is almost
wholly internal, and directed towards the lowest parts of the
many depressions which, owing to earth movements and wind
erosion, characterise the desert surface. Evaporation prevents
the growth of lakes from which, as in humid regions, the over-
flow could find an exit. The poetical generalisation that “ the
weariest river winds somewhere safe to sea ” does not apply to
arid regions.
The gorges and steep-sided wadis that dissect the uplands
(Plate 62), the alluvial spreads that floor the depressions,and
the salt deposits and other dried-up remains of ephemeral and
vanished lakes all testify to the fact that running water as well
as wind is concerned with the evolution of the desert landscape.
Most of these features are an inheritance from Pleistocene
times, when there were climatic oscillations involving periods
of greater rainfall (Plate 63b). In the Sahara and Syria
270
CLOUDBURSTS IN THE DESERT
occurrences of buried habitations, stone implements, and the

fossil remains of trees of modern types all point to the com-
paratively recent prevalence of more genial conditions. Evi-

dence for the more active flow of streams is found in the extinct
“ waterfalls ” of certain wadis.
Despite the extreme desiccation of to-day, spasmodic
stream action still continues. The chief characteristics of the
rare desert rainstorms are their erratic distribution and brief
—
duration, and apart from occasional light showers their —
intense violence. Houses of dried salt mud are washed away
when a sudden “ cloudburst ” descends on an oasis, and
travellershave been drowned in the floods that race down the
dry wadis with little or no warning. Such torrents, swiftly
generated in a distant ipland rainstorm, carry a heavy load
of mixed debris, prepared for them by years of weathering and
wind erosion. Much of the load is dropped to form alluvial
fans and “ deltaic ” deposits in front of the mountains or escarp-
ments, as the carrying power is reduced by loss of gradient or
by seepage. Choked by their own deposits the streams sub-
divide into innumerable channels and spread out laterally
across the plain, which thus becomes covered with a veneer of
fine sediment. If the rain falls over the gentle slopes of a
depression a shallow sheet-flood may carry a mud flow towards
the centre. If the water reaches the centre before it is lost by
further seepage and evaporation a temporary lake is formed.
There the dissolved material is concentrated by evaporation
and finally deposited to form salt muds or glistening white
sheets of rock salt, gypsum, and other salts. In western America
the resulting alluvial plains are called playas and the more
saline tracts are distinguished as salinas (Plate 85b). As only
the finer debris is exported by wind, and salt-encrusted
surfaces offer considerable resistance to ablation, the lower-
ing of the surface by wind erosion is a very slow process
which does not everywhere succeed in keeping pace with
deposition.
If the alluvium increases in thicknessit extends up the
valleys and encroaches on the wind-swept and flood-washed

rock slopes that lead up to the spurs and cliffs of mountains
271
and escarpments. So long as more material is brought in by
streams than is removed by wind the depressions continue to
be filled. It is doubtful whether this can happen in the heart
of the desert, but it certainly takes place in semi-arid regions
like the Great Basin (page 423), and it must have occurred
in
the present deserts when the rainfall was less scanty than now.
In any case, the low swells between adjoining basins are sooner
or later worn down, and the spreads of alluvium gradually
extend from one depression to another, thus becoming united
into still vaster plains.
A striking feature of desert and semi-arid landscapes is the
sudden change of slope which is maintained as the edges of
the uplands are worn back by erosion. The slopes of mountains
and escarpments and the walls of the wadis are kept steep not
only by wind abrasion at the base, but also because stream
erosion, when it does occur, is mainly lateral. Near the heads
of gorges and wadis the rare floods flush out the loose material
and cut into the bedrock, but lower down the main channel is
blocked by debris, and the outer distributaries, diverted to one
side or the other, undercut the walls (Plate 62). Another effect
which is important is that rain falling on a plateau of
locally
pervious rocks soaks through them and feeds intermittent
springs at the foot of the bordering cliffs, which are thus again
sapped away at their base. Meanwhile, the recession of the
upper slopes is extremely slow, because disintegration is very
slight compared with the effects of weathering in humid
regions. Thus the typical wadis and desert gorges come to have
broad flat floors and steep sides, rising in terraced steps to a
vertical lip. Near their mouths they open out into alluvial
fans which slope down to the plains. Between neighbouring
wadis the plateaus rise by similarly abrupt cliffs and terraces,
corresponding to the varied resistances of the truncated strata.
In the absence of intervening soft beds, as in the tableland of
the Gilf Kibir to the south of the Egyptian oases, the ascent
may be by a single bold and precipitous escarpment. Between
the cliffs and the feather edge of the alluvium there is generally
a gently sloping rock surface for which the term pediment has
been proposed.
272
PLATE 62
1
'
/ > "fc
By permission of the Egyptian Government' [O. II. Little
(A) Gorge of the Wadi Barud
By permission oj the Egyptian Government' [H\ F. Hume

: Bj Wadi Gasab in the Plateau of Ma'aza
WADIS IX THE EGYPTIAN DESERT

PLATE <)!i
DEVELOPMENT OF DESERT LANDSCAPES
The Cycle of Erosion in Arid Regions
There is still considerable doubt regarding the relative

importance of the effects of wind and water in the desert. As
we have seen, there have been marked fluctuations of climate
in the not very distant past, and certain desert features, such
as the larger wadis and the extensive deposits of former lakes,
could hardly have developed under the parching conditions
of the present day. Such features are therefore shared by the
landscapes of both desert and semi-arid regions. The following
outline, though brief, will serve to indicate how the processes
known to operate in the desert could ultimately reduce any
given land surface to a desert peneplain.
The arid cycle may be supposed to begin on a region of
uplands and depressions, determined, let us say, by earth
movements. During the stage of youth each basin is separate
from its neighbours. In the uplands short intermittent streams
cut gullies and ravines which develop into wadis and canyons.
The rock-waste not carried off to the sea, but is spread over
is
the depressions, turning them into sandy plains passing inwards

into salt-cemented playas. By erosion the slopes of the uplands
gradually recede, leaving narrow rock pediments in front, and
the upper surfaces are slowly worn down. By accumulation of
alluvium the basins are filled up. Consequently the general
effect is towards levelling, although, as already pointed out,
the contrast of cliffand plain never ceases to remain sharp.
Moreover, offsetting the reduction of relief, new hollows are
excavated by the wind, and dunes begin to grow and to migrate
from the sources of sand supply.
The stage of maturity may be said to be reached when the
lower basins begin to tap the higher ones by valleys which have
extended headwards. The deposits of a higher basin may
develop into typical badland topography as they are eroded
away. Neighbouring basins thus gradually coalesce into more
extensive ones which for the most part are almost perfectly
flat plains, with tracts of mosaic pebble-pavement and vast
sand wastes and chains of dunes. Meanwhile the uplands are

dwindling away, and their remnants, flanked by outlying
273
buttes, rock towers, and pinnacles, rise abruptly from broaden-
ing pediments which slope down to the plains (Plate 64).
As the uplands are lowered the rainfall decreases still
further, and less and less new material is contributed to the
plains by running water. Wind action then becomes pre-
dominant. The general lowering of the surface by deflation
and abrasion proceeds with little or no interruption by sheet
floods and mud flows. This stage marks the beginning of old
age. The shrinking uplands are worn back, leaving increasingly
widespread pediments, diversified by clustered or isolated
residual knobs and peaks, the most spectacular examples being
the inselbergs developed from resistant masses of granitic and
gneissic rocks. Until the cycle is interrupted by earth move-
ments or a change of climate, wind erosion continues to lower
the surface towards the base level fixed by the water table.
And so the region is reduced to a peneplain of rock, desert
pavement, and sand, the only breaks in the monotonous land-
— —
scape apart from barchans and seif dunes being provided
by residual domes and occasional steep-sided relics of the more
persistent inselbergs.
The most remarkable examples of the old-age inselberg type
of landscape occur in certain regions of Africa ( e.g Mozambique
.
and Tanganyika territory) where the present climate is one of

well marked wet and dry seasons. Precipitous, smoothly
rounded peaks, isolated or in detached groups, rise boldly and
abruptly above the forest, like islands above a sea of vegeta-
tion (Fig. 144). For this reason they have been called inselberge
(island mounts). Other examples, only a little less spectacular,
occur in the semi-arid regions of South-W est Africa, Northern
Nigeria (Plate 65a), and Western Australia, and in the Arabian
desert, everywhere carved out of the more resistant granites
and gneisses.
The surrounding plain or plateau is a more or less uplifted
plain or plateau of the pediment type, with only a thin covering
of soil, sand, gravel, or laterite, according to the nature of the
present-day climate. An inselberg is clearly a special type of
monadnock (page 189), but it differs from the latter, as de-
veloped in the humid cycle, in its astonishingly steep sides and
274
INSELBERG LANDSCAPE
_
r
Fig. 144
Sketches of the inselberg landscape of Mozambique by E. J. Wayland :
(a) Koldwi, north of the Mtupa Pass, Ribawe

(
b) Lebi, Mbulla Range (vegetation omitted)
in the —
abrupt transition by way of a sharply concave bend—
to the surface of thepediment. This contrast suggests that the
inselberg landscape of eastern Africa may not have developed
under present-day conditions, but is more likely to be an
inheritance from a time when the climate was arid or semi-
arid. Certainly the landscape is a very ancient one. In some
coastal districts the pediment continues beneath a cover of
Tertiary or Cretaceous sediments. There has therefore been
ample time for a wide range of climatic conditions. Wherever
the inselberg landscapeis developed, a thick cover of younger
formations must long ago have been removed from the old
Pre-Cambrian crystalline rocks of which the inselbergs and the
surrounding pediments are composed.
275

W. F. Hume
Geology of Egypt, Vol. I. Government Press, Cairo, 1925.
E. F. Gautier (
translated by D. F. Mayhew)
Sahara, the Great Desert. Columbia University Press, New York,
1935 -
G. PlCKWELL
Deserts. Whittlesey House (McGraw-Hill), New York, 1939.
R. A. Bagnold
The Physics of Blown Sand and Desert Dunes. Methuen, London, 1941.
C. A. Cotton
Climatic Accidents in Landscape- Making. Whitcombe and Tombs,
Christchurch, New Zealand, 1942.
276
aiS
;
Chapter XIV
COASTAL SCENERY AND THE WORK OF

THE SEA
Shore Lines
Nearly all coast lines have been initiated by relative move-
ments between land and sea. A rise of sea level (see page 417),
or a depression of the land, leads to the submergence of a
landscape already moulded by sub-aerial agents. The drown-
ing of a region of hills and valleys gives an indented coast line
of bays, estuaries, gulfs, fjords, and straits, separated by head-
lands, peninsulas, and off-lying islands. Very broad bays,
like the Great Australian Bight, result from the submergence
of plains. Coasts that have originated in these ways are called
coasts of submergence. Conversely, a fall of sea level or an eleva-
tion of the land and adjoining continental shelf leads to retreat of
the sea and emergence of part of the sea floor. As the sea floor
is essentially a realm of sedimentation its surface is generally
smooth, and the resulting coasts of emergence have correspondingly

simple, broadly flowing outlines. Other varieties of coast line
include those determined by (a) volcanic activity (b) faulting
;
(c) glaciation ;
and (d) the growth of coral reefs and atolls.
The general outlines of a newly formed coast are soon
modified by marine erosion and deposition, with development
of a wide variety of shore features and coastal scenery. By the
incessant pounding of waves, which break up the rocks and
wear back the cliffs, the sea cuts its way into the land like a
horizontal saw. The liberated rock fragments are rounded by
innumerable impacts and continual grinding as the line of
breakers is carried backwards and forwards over the foreshore
by the ebb and flow of the tide. The worn-down material is
supplied to currents which dispose of it, together with all the
products of land-waste brought in by rivers and glaciers and
the wind. Much of the sediment is carried into deeper water
before it comes to rest out on the sea floor, but some is drifted
396 )
(
277 19
COASTAL SCENERY AND THE WORK OF THE SEA
along the shore by waves and currents, to form shoals and

beaches, and to build up bars and spits wherever the trans-
porting power is checked. Sheltered inlets thus barred off from
the sea by long embankments of sand and shingle become silted
up. In this and other ways ( e.g by the seaward growth of a
delta) new land is added to the fringe of the old in partial
compensation for the losses suffered elsewhere.
The waters of the seas and oceans readily respond by move-
ment to thebrushing of the wind over the surface to varia-
;
tions of temperature and salinity ; and to the gravitational

attraction of the moon and sun (combined with the earth’s
rotation). The work of erosion, transport, and deposition
carried out by the sea depends on the varied and often highly
complex interplay of the waves, currents, and tides that result
from these movements.
It may be noted in passing that lakes, especially the larger
ones, behave in much the same way as enclosed seas. In
consequence, the shore features of lakes and seas have much
in common. A lake formed by obstruction (e.g. the lava-
dammed Lake Kivu, p. 436) drowns the surrounding land and
so acquires a shore line of the submergence type. A lake which
has shrunk from its former extent in response to climatic or
other changes (e.g. the Great Salt Lake of Utah, p. 424) is
margined by flats and terraces of sediment (Plate 63b), and so
acquires a shore line of the emergence type. Tides are negligible
in lakes, but seasonal variations of rainfall may cause the water
to advance and recede over a tract of shore which is alternately
covered and exposed, though less frequently than the tide-
swept foreshore of a coast. Waves and currents operate
exactly as in land-locked seas of similar extent and depth,
and are responsible for erosion and deposition on a corre-
sponding scale. There are, of course, important biological
contrasts. The swamps into which shallow
lakes degenerate,
with their luxuriant growth of aquatic vegetation and accumu-
lations of peat, are very different from the mangrove swamps
and tidal marshes that locally border the sea. On the other
hand, lake shores have nothing to be compared with the coral
reefs of tropical seas.
278
GENERATION OF TIDES
Tides and Currents
The tide is the periodic rise and fall of the sea which, on
an average, occurs every 12 hours 26 minutes. Tides are
essentially due to the passage around the earth, as it rotates,
of two antipodal bulges of water produced by the differential
attraction of the moon and sun. It is easy to understand that
the water facing the moon should bulge up a little, but it is
less obvious why there should be a similar bulge in the opposite
direction on the other side of the earth. The basis of the
explanation is that the water centred at A
(Fig. 145) is attracted
towards the moon more than the earth, centred at E, while the
EM-60r * oc
'M
BM=6lr oc
Fig. 145
Diagram to illustrate the generation of tides
earth, in turn, is attracted more than the water centred at B.

The water at the far side is thus left behind to almost the same
extent as the water on the near side is pulled forward. From
places such as C and D the water is drawn away and low tide
results. As the earth rotates each meridian comes in turn
beneath the positions of high and low tide nearly twice a day ;
not exactly twice, because allowance must be made for the

forward movement of the moon.
The effect of the sun is similar to that of the moon, but
considerably less powerful. When the earth, moon, and sun
along the same straight line, the tide-raising forces of sun
fall
and moon help each other, and tides of maximum range,

279
known as spring tides , result. When the sun and moon are at
right angles relative to the earth, the moon produces high tides
where the sun produces low tides. The tides are then less high
and low than usual and are called neap tides.
In the open ocean the difference in level between high and
low tide is only a few feet. In shallow seas, however, and
especially where the tide is concentrated between converging
shores, ranges of 20 to 30 feet are common and tidal currents
are generated. A current of 2 miles an hour accompanies
the inflowing tide as it advances up the English Channel.
In the Bristol Channel the spring tides reach a height of 42
feet, and give rise to a current of 10 miles an hour. In extreme
cases, such as the latter, where the tidal stream is crowded into
the narrow end of a shallowing funnel, the water advances
with a wave-like front of roaring surf which is known as a bore.
Near the shore inflowing tidal currents are often sufficiently
powerful to move shingle and so to scour the bottom and
transport sediment inshore or alongshore. The complementary
outflowing currents of the ebb tide are less effective as eroding
and transporting agents, because they start in shallow water
and advance into deeper water. Pebbles and sand are left
behind, and only the finer material is drawn back. In estuaries,
where the outward flow of river water is added to the ebb
current, transport is dominantly seaward. But since the fresh
river water, carrying a load of silt and mud, tends to slide out
over the heavier salt water which has crept in along the bottom,
it is the upper suspended load that is mainly swept out to sea,
while the coarser debris is stranded and tends to accumulate

as sand bars.
Currents of a convectional type arise as a result of variations
in the density of sea water, brought about to a very slight
extent by heating in tropical regions and cooling in polar
regions, but far more effectively by changes of salinity. Such
changes depend on {a) inflow of rivers, rainfall, and the melting
of ice, all of which freshen the water and ( b ) evaporation,
;
which increases the salinity. The dissolved salts in sea water

have the following average composition, corresponding to a
salinity of 35 :
280
[II. M. Gcol. Survey
(B) Recession of cliffs of glacial sands and gravels, 1930. South of
Lowestoft. Suffolk
COAST EROSION
SALINITY CURRENTS
Parts per 1 ,000 Salinity in Particular Regions
Sodium chloride NaCI 27-213 North Red Sea 41
Magnesium chloride MgCl 2 3-8(17 Eastern Mediterran ean 39
Magnesium sulphate MgSO, 1-658
Calcium sulphate CaSO, 1-260 North Sea 34
Potassium sulphate k so 4
2
0-803 Near Greenland 31-33
Calcium carbonate CaC0 3 0-123 Black Sea 18
Magnesium bromide MgBr2 0-070 Baltic Sea 2-8
Evaporation over the Mediterranean lowers its surface and
increases the salinity and density. Surface currents therefore

flow into the Mediterranean through the Dardanelles from
the Sea of Marmora and the Black Sea (where the evaporation
is more than neutralized by the inflow of rivers), and through
the Straits of Gibraltar from the Atlantic (Fig. 146). In each
Fig. 146
Salinity currents of the Mediterranean and Black Seas. Water of low salinity
(thin arrows) flows into the Mediterranean, and water of high salinity (thick
arrows) flows out
case undercurrents of higher salinity flow outwards from the

Mediterranean along the bottom. Shore deposits are affected
by the surface currents, while in deeper water the floor is
scoured by the bottom current. A similar interchange of
water takes place between the highly saline Red Sea and the
Indian Ocean, and also between the comparatively fresh
Baltic and the North Sea. Since the less saline and lighter
water always tends spread over the more saline and heavier
to
water, the surface current flows in all cases towards the region
of higher salinity, while the bottom current is in the opposite
direction.
The prevailing winds and the configuration of the conti-
nents determine the superficial but gigantic eddies which
main currents of the oceans. These have note-
constitute the
worthy climatic and biological effects, but are otherwise of
281
little In the vicinity of land, however,

geological importance.
the currents due to wind drift contribute to the general
move
ments responsible for shore-line processes.
Waves
Apart from and accidental disturbances of the
tidal effects
sea associated with earthquakes and volcanic eruptions, waves
are entirely due to the sweeping of winds over the surface of
the water. The surface is thrown into undulations which
move forward and gradually increase in height and speed.
The height of a wave is the vertical distance from trough to
crest (Ca in Fig. 148). The horizontal distance from crest
-Wave Length
Crest
Direction of wave advance
Fig. 147
Profile of a wave of oscillation from crest to crest, showing the directions of
movement of water particles at various points
to crest —
or from trough to trough— is called the wave length.
The height ultimately attained by a wind-driven wave, where
it is not restricted by shallowing water, depends on the strength,
duration, and fetch of the wind, the fetch being the length of the
open stretch of water across which the wind is blowing. When
the loss of energy involved in the propagation of the waves
through the water is just balanced by the amount of energy
supplied from the wind, the height reaches its maximum. At
this limit the numerically, roughly half the
height in feet is,
speed of the wind in miles per hour. In the open ocean heights
of 5 to 15 feet are common, increasing to 40 or 50 feet in severe
storms. The corresponding wave
lengths range from 200 to
700 and the speeds from 20 to 60 miles per hour. Waves
feet,
may travel into regions of calm weather far beyond the fetch
of wind in which they were generated, thus giving rise to a
groundswell. The “ free ” waves of a groundswell gradually
282
WAVE MOTION
flatten out relative to the “ forced ” waves of the high seas
from which they spread, and wave lengths of 1,000 feet or
more are then not uncommon.
It important to realize that in the open sea apart from
is —
wind drift —
it is only the wave form that moves forward, not
the water itself. Each particle of water moves round a circular

orbit during the passage of each complete wave, the diameter
being equal to the height of the wave (see Figs. 147—148). This
is demonstrated by the behaviour of a floating cork under
which a train of waves is passing. Each time the cork rises

and falls it also sways to and fro, without advancing appreciably
from its mean position. Such waves are called waves of oscilla-
Diagram showing the orbit of a water particle during the passage of a wave of
oscillation.A, B, C, D, E mark successive positions of the crest ; a, b, c, d, t
are the corresponding positions of the particle. AE = wave length ; Ca = wave
height
tion. If the wind is however, each water particle

strong,
advances a little farther than
recedes.
it Similarly, in shallow
water, where friction against the bottom begins to be felt, each
particle recedes a little less than it advances. In both cases the
orbit, instead of being a closed circle, then resembles an
ellipse which is not quite closed and the water therefore slowly
drifts forward in the direction of wave advance.
When gusts of wind brush over a field of corn the stalks
repeatedly bend forward and recover, and waves visibly
spread across the surface. Here it is obvious that the wave
motion is not confined to the surface, since it is shared by the
stalks right down to the ground. In the same way the energy
contributed by the wind to a body of water is transmitted
downwards as well as along the surface. Owing to friction
the diameters of the orbits rapidly diminish in depth until at
a depth of the same order as the wave length they become
283
, .
negligible. The depth at which sediment on the

greatest
sea floor can just be stirredby the oscillating water is called the
wave base. In shallow seas and on the continental shelf, where
the depth of water is less than this (i.e. less than about 600 feet),
pebbles, sand grains, and mud particles can thus be kept moving
until currents roll or waft them to depths where they remain
at rest, undisturbed by wave action. Off Land’s End stones
up to a pound in weight are sometimes washed into lobster
crawls at depths of 100 to 200 feet. Mud particles may even
be kept on the move until they are quietly dropped over the
edge of the continental shelf.
Agitation of the sea floor involves friction against the bottom,
Fig. 149
Wave refraction. Diagram to the swing of oblique waves towards
illustrate
parallelism with the shore. While the crest at a advances to a', the crest at b
(in shallower water) advances a shorter distance to b' and so on. The crest
lines abed thus become curved as shown
and in consequence, as the

water shallows, the waves in front
are retarded and the wave length decreases.
For this reason,
as a wave approaches a shore line obliquely its crest line swings
towards parallelism with the shore, as shown in Fig. 149.
The effect of wave retardation off an indented coast is illus-
trated by Fig. 150. The waves advance more rapidly through
the deeper water opposite a bay than through the shallower
water opposite a headland. Thus the crest of a wave at a
moves to a', while the moves only to b' The waves
crest at b
thus become curved or again towards parallelism with
refracted,
the shore line. Consequently, when the shore is reached by
a wave such as abode, all the energy from the long stretch ac
is concentrated on the headland AC (and that from de on DE).
284
WAVES AND BREAKERS
In contrast, the very much smaller amount of energy from the
short stretch cd is spread around the shores of the bay from
C to D. Thus, while headlands are being vigorously attacked
by powerful waves, the bays are not unduly disturbed and
their waters provide safe anchorage for vessels sheltering from
a storm. In the same way waves entering a harbour between
the piers spread out and merely ruffle the water inside.
When a wave reaches the foreshore and enters water of
which the mean depth is about the same as the height of the
Refraction of waves approaching bays and headlands, showing the concentration

ofwave energy against the headlands and its diffusion around the shores of the bays
wave, it becomes a breaker. The amount of water in front

being insufficient to complete the wave form, while the orbital
motion still continues, the crest is left unsupported and plunges
over. The surface water then surges forward as a turbulent
sheet of surf. The wave of oscillation has passed into a wave oj
translation. It is the bodily advance of the water in this kind
of wave that makes surf-riding possible. The final uprush or
swash on a shelving beach is followed by a backwash, as the
water returns down the slope.
Surf waves pile up the water against the coast, and onshore
285
)
currents, especially the drift maintained by winds blowing

strongly towards the shore, have the same
This tendency
effect.
to raise the sea level counterbalanced by a compensating
is
current, called the undertow, which flows along the bottom

away from the shore. The undertow is a pulsating current, be-
coming stronger under the wave troughs, where it is reinforced
by the orbital motion of the water, and correspondingly weaker
under the crests. It also tends to be concentrated along hidden
depressions on the sea floor. The undertow is often a source
of serious peril to bathers.
Marine Erosion
The sea operates as an agent of erosion in four different
ways :
(a) by the hydraulic action of the water itself, involving the
picking up of loose material by currents and waves, and the

shattering of rocks as the waves crash, like giant water-hammers,
against the cliffs (Plate 66a) ;
(b) by corrosion, when waves, armed with rock fragments,

hurl them against the cliffs and, co-operating with currents,
drag them to and fro across the rocks of the foreshore
;
(c) by attrition, as the fragments or “ tools ”

are themselves
worn down by impact and friction and ;
(
d by corrosion, i.e. solvent and chemical action, which in
the case of sea water is of limited importance, except on lime-
stone coasts.
The destructive impact of breakers against obstructions is

often far greater than generally realized.
is The pressure
exerted by Atlantic waves averages over 600 lb. per square
foot during the summer, and over 2,000 during
the winter,
while in great storms it may exceed even 6,000. Thus, not
only cliffs, but also sea walls, breakwaters, and exposed light-
houses are subjected to shocks of enormous intensity. Cracks
and crevices are quickly opened up and extended. Water is
forcibly driven into every opening, tightly
r
compressing the air

confined within the rocks. As each w ave recedes the coni'
r
286
CLIFF SCENERY
pressed air suddenly expands with explosive force, and large

blocks as well as small thus become loosened and ultimately
blown out by pressure from the back. The combined activity
of bombardment and blasting is most effective as a quarrying
process on rocks that are already divided into blocks by
jointing and bedding, or otherwise fractured, e.g. along faults
and crush-zones.
Cliffs originate by the undercutting action of waves against
the slopes of the coastal land. By collapse of the rocks over-
hanging the notch which is excavated at the base of the cliffs,
the latter gradually recede and present a steep face towards
the advancing sea (Plate 68b). But where the cliffs are pro-
tected for a time by fallen debris, and especially if they are
composed of poorly consolidated rocks, the upper slopes may
be worn back by weathering, rainwash, and slumping. At
any given place the actual form of the cliff depends on the
nature and structure of the rocks there exposed, and on the
relative rates of marine erosion and subaerial denudation.
Since there are few stretches of coast along which the rocks
are equally resistant, coasts are generally worn back with
pronounced irregularities. Caves are excavated in belts of
weakness and especially where the rocks are strongly jointed.
By subsequent roof collapse and removal of the debris long
narrow inlets develop. In Scotland and the Faroes such a
tidal inlet (Plate 67a) is called a geo (“ g ” hard—Norse, gya a,
creek). —
A cave at the landward end of a geo or indeed any sea
—
cave may communicate with the surface by way of a vertical
shaft which may be some distance from the edge of the cliff.
A natural chimney of this kind (Plate 67b) is known as a blow
hole or gloup (a throat). The opening is formed by the falling
in of joint blocks loosened by the hydraulic action of wave-
compressed air already described. The name blow hole refers to
the fact that during storms spray is forcibly blown into the air
each time a foaming breaker surges through the cave beneath.
When two caves on opposite sides of a headland unite, a
natural arch results, and may persist for a time (Plate 68a).
Later the arch falls in, and the seaward portion of the headland
then remains as an isolated stack (Plate 68b). Well known
287
II. M. heol. Survey
Fit;. i.)2
The Old Man of Hoy, Orkneys. A stack of Old Red Sandstone, 4-50 feet high
rising from a platform of lava on which the sandstone lies unconformably
SHORE PLATFORMS
examples of stacks are the Chalk pinnacles at the western

extremity of the Isle of Wight known as The Needles (Fig. 151),
and the impressive towers of Old Red Sandstone in the Orkneys,
one of which, 450 feet high, is called the Old Man of Hoy
(Fig. 152).
As the worn back a wave-cut platform is left in
cliffs are
upper part of which is visible as the rocky
front (Fig. 153), the
foreshore exposed at low tide (Plate 69). There may be
patches of sand and pebbles in depressions, and beach-like
fringes strewn with fallen debris along the foot of the cliffs,
but allsuch material is continually being broken up by the

waves and used by them for further erosive work, until finally
it is ground down to sizes that can be carried away by currents.
Cliff
'
Natch ^ High Tide Level

w
• fd up -cut Platform ' Low Tide Level
Fro. 153
Diagrammatic section showing a stage in the development of sea cliff, wave-cut
platform, and wave-built shore-face terrace
While going on the platform itself is being abraded

all this is
as debris swept to and fro across its surface. Since the
is
outer parts have been subjected to scouring longer than the

inner, a gentle seaward slope is developed. In massive and
resistant rocks this is an extremely slow process. Consequently,
as the cliffs recede and the platform becomes very wide, the
waves have to cross a broad expanse of shallow water, so that
when they reach the cliffs most of their energy has already
been dissipated. Thus the rate of coast erosion is automatically
reduced. In high latitudes, however, the cliffs may still
continue to be worn back by frost and thaw, provided that the
waves are able to remove what would otherwise become a
protective apron of scree. The wave-cut platform off the rocky
coast of west and north-west Norway there known as the —
strandflat —
has reached an exceptional width by this co-
289
s.w. N.EL
Fedje Fjord Fens Fjord,

F Lindas
Igneous and Metamorphic Rocks Peninsula
Fig. 154
Section across the strandflat north of Bergen, Norway. Length of section 32 miles.
(After F. .Xamen)
operation of processes, locally up to as much as 37 miles. It

now stands slightly above the present sea level (because of
recent isostatic uplift) and innumerable stacks and skerries rise
above its surface (Fig. 154).
On the other hand, if the sea is encroaching on a coast of
Fig. 153
Map shotting the coast erosion and lost villagesof Holderness since Roman times
{After T. Sheppard)
290
MARINE EROSION AND DEPOSITION
poorly consolidated rocks, the platform in front is much more
quickly abraded and normal coast erosion proceeds vigorously
(Plate 66b). In some localities the inroads of the sea reach
alarming proportions. The most serious loss of land in Britain
is suffered along the Yorkshire coast south of Flamborough
Head, where the waves have the easy task of demolishing glacial
deposits of sand, gravel, and boulder clay. Since Roman
times this 35-mile stretch of coast has been worn back 2^ or
3 miles, and many villages and ancient landmarks have been
swept away (Fig. 155). During the last hundred years the
average rate of cliff recession has been 5 or 6 feet per year.
The rate is not uniform, however, for severe storms and localized
cliff falls do more damage in a short time than is otherwise
accomplished in several average years.
The Profile of Equilibrium

In appropriate circumstances some of the sediment in
transit across thewave-cut platform accumulates in the deeper
water beyond, to form a shoreface terrace which grows forward
like a broad embankment with its upper surface in smooth
continuity with the platform. The combined shore and off-
shore surface in this case, as in others, is a product of the joint
action of erosion and deposition, each of which varies con-
siderably from time to time and from place to place. The
supply of sediment, for example, is irregular both in rate and
distribution, since contributions are received from rivers and
currents as well as from cliff wastage and platform abrasion,
all fluctuating sources of income. The processes concerned
in the removal of sediment —
also widely variable —
are them-
selves largely controlled by the slope of the shore and its sea-
ward continuation — that is to say, by the profile of the surface
taken at right angles to the shore. A relatively steep slope
favours removal of sediment from the landward side, so that
the slope becomes less steep. Conversely, a relatively gentle
slope favours beach deposition on the landward side, so that
the slope becomes steeper. The surface is therefore being
291
continually modified, and in such a way that at each point it

tends to acquire just the right slope to ensure that incoming
supplies of sediment can be carried away as fast as they are
received. When the profile is so adjusted that this state of
balance is achieved, it is called a profile of equilibrium.
For every set of conditions there is an appropriate profile
of equilibrium and, as the conditions vary, so the actual profile
is modified to keep pace with them, generally by fluctuations
about an average which is fairly stable.

Along a shore of submergence the slope of the initial
surface may
be either steeper or gentler than that of the ideal
profile of equilibrium. Suppose AB in Fig. 156 represents a
steep initial slope. In transforming this into the profile of
equilibrium CD, the waves cut a cliff-backed platform and
Fig. 156 Fig. 157

The development of a profile of The development of a profile of
equilibrium CD from a more steeply equilibrium cd from a more gently
sloping initial surface, AB sloping initial surface, ab
wear it down by abrasion, and the resulting sediment is de-

posited as a shoreface terrace, as already illustrated in Fig. 153.
If conditions then change so that the profile already in existence
(ab in Fig. 157) steep than the new profile of equilibrium
is less
cd, then the platform is built up by deposition of sediment.
Thus it commonly happens that a wave-cut platform is partially

or wholly covered with a veneer of beach material.
Next, suppose ab represents an initial slope that is relatively
gentle — as when a broad valley becomes a bay by submergence.
In transforming this into the profile of equilibrium cd, waves
and currents build up a beach around the shores of the bay.
Initial surfaces that may be almost flat are provided by the
drowning of extensive alluvial plains, and on a still more
widespread scale by uplift of the sea floor. In these cases, as
described on page 304, barrier beaches, called offshore bars,
begin to develop far from the ill-defined mainland, and loose
292
BEACHES
sediment from the sea floor is driven landwards. As the waves

lose energy the sediment is deposited long before the shore is
reached, and thus the submarine foundations of an offshore bar
are laid. The further development of these constructions is
described on pages 304-6).
Transport and Deposition Transverse to the Shore

Leaving the longshoredrift of sediment to be considered
later, let us now review
the actual processes that control the
seaward and landward migration of beach and terrace material.
In bays where the initial slope is gentle, the landward
currents due to translatory waves and inflowing tides are at
first more
effective on balance than the weak seaward currents
due to undertow and outflowing tides (cf. page 286). A beach
is therefore built up to the equilibrium profile. The most
obvious landward movement is seen when the uprush of a
breaker sweeps a mixed assemblage of sediment up the beach.
Some of the coarser material, often including cobbles and
pebbles, is left stranded at the top, while the rest is dragged
back by the backwash. Storm waves at specially high tides
thus build up a coarse storm beach well above the reach of
ordinary waves and tides. Once the beach profile is adjusted
to average conditions any additional supplies of sediment are
removed, either by seaward currents (which now have the
advantage of gravity owing to the steepened slope) or, when the
sand is dry, by wind.
If the prevailing winds drive the beach sand landwards
to form a belt of sand dunes, the profile is restored by com-
pensating additions from the sea floor. Such landward
migration is convincingly demonstrated by the existence of
beaches and dunes that are largely composed of ground-up
shells that could only have come from the sea floor. Excellent
examples occur at St. Ives and Perranporth along the north-
west coast of Cornwall.
The undertow, by outflowing tides, carries sand
assisted
and finer sediment seawards, and very large quantities may
be scoured away from the beach when onshore gales raise the
(396; 293 20
head of water against the coast, and so maintain a powerful

undertow. During such storms beaches become thinned and
considerable patches may be entirely removed for a time. But
gradually the beaches are built up again, restored by waves
and currents, including shoreward undercurrents set up by
offshore winds which lower the head of water against the coast.
Beach growth is thus particularly favoured by gales blowing
offshore. It should be noticed that the hydraulic currents
along the sea floor always transport material in the direction
opposite to that of the wind.
Interference with the natural profile by offshore dredging
may increase the undertow to such an extent that the beach
is withdrawn until the artificial depression is filled up. In
1896, for example, dredging was started off the coast north of
Start Point to furnish shingle for use in the harbour works
at Plymouth. As a result the beach at Hallsands began to dis-
appear, cliff erosion became an ever-increasing menace, and
in 1917 the village itself was washed away.
Off headlands the sea floor often shelves comparatively
steeply to begin with and the seaward currents are the more
effective ones from the start, a strong undertow being favoured
both by wave concentration and gravity. Figs. 153 and 156
illustrate the results. The boulders and large pebbles that
remain for a time under the cliffs are steadily reduced in size
until they are sufficiently ground down to be removed. A
shingle beach is simply a transitional stage through which the
larger fragments of cliff debris must pass before beginning their
seaward journey.
Ultimately the offshore slopes, though fluctuating from
time to time, everywhere become adjusted into approximate
equilibrium with the locally prevailing conditions. There is
then a slow seaward migration of material at the expense of
the wasting land. The transitory coarser debris is concen-
trated against the land, while the finest particles are swept out
to sea, with much to-and-fro movement on the way, until
each in turn passes beyond the range of wave and current
action and finds a resting-place, either by burial or by being
swept over the edge of the continental shelf. There is always
294
BEACH DRIFTING
a tendency for some of the material to be carried to a depth
from which it cannot be returned, because gravity assists the
outward and hinders the inward movements. Moreover, the
currents and wave oscillations rapidly weaken as the depth of
water increases. It therefore follows that for each intermediate
depth on the continental shelf there is a particular size of
particle that is just too large to be returned landwards. Thus,
during the outward migration of grains and particles of varied
sizes, each size in turn, from large to small, tends ultimately
to come to rest. There is inevitably a certain amount of over-
lapping of sizes, owing to variations of movement along the
sea floor, and also because at any depth finer particles may be
trapped in the interstices of the coarser material, so that sorting
isnever quite perfect. But with this reservation the sediments on
the sea floor are found to become finer and finer as the depth
increases ;
the full outward sequence being boulders, cobbles,
and pebbles ;
medium, and fine sands ; and mud.
coarse,
The corresponding consolidated rocks are conglomerates, sand-
stones, and shales or mudstones.
Transport and Deposition along the Shore
Longshore drift of sediment is brought about in two ways :
by beach drifting, due mainly to oblique waves, and farther

out, by transport due to longshore currents. When waves are
driven obliquely against the coast by strong winds, debris
is carried up the beach in a forward sweeping curve. The
backwash may have a slight forward movement at the start,
owing to the swing of the water as it turns, but otherwise it
tends to drag the material down the steepest slope, until it is
caught by the next wave, which repeats the process (Fig. 158).
By the continual repetition of this zig-zag progress sand and
shingle are drifted along the shore.
The direction of drifting may vary from time to time, but
along many shores there is a cumulative movement in one
direction, controlled by the prevailing or most effective winds.
A subsidiary factor which aids or hinders beach drifting is the
295
direction of the advancing flood tide. Some of the transported

material dropped before or during the slack at high tide is not
carried back again by the sluggish beginnings of the ebb-tide
current. Farther out from the shore, wherever the general
configuration of the coast is favourable, oblique winds and
waves co-operate in generating intermittent and fluctuating
longshore currents which again may be strengthened or
weakened by the set of the tides. Thus longshore drift is by
no means confined to the immediate shore line.
In the English Channel the dominant winds and the ad-
vancing tides both come from the south-west, and the prevailing
direction of drift is therefore up channel. Along the east coast
the drift is mainly southwards, as there the effective winds are
Fig. ] .58
Diagram beach drifting, showing the path along a sloping beach,

to illustrate
followed bv a pebble under the influence of the uprush and backwash of successive
oblique waves during an advancing tide
from the north-east, and the flood tide advances from the
north. Wherever it is deemed desirable to protect the coast
by checking the drift of sand and shingle, barriers, known as
groynes, are erected across the beach. On the windward side
of a groyne the debris is heaped up, while from the lee side it
is washed away, to be retained in turn by the next groyne
(Plate 70b). The groynes as a whole interrupt the natural

flow of material necessary to maintain the beaches farther along
the coast. Where the beaches are starved and drift away, as
a result of interference with the balance of deposition
artificial
and erosion elsewhere, the coast is exposed to more vigorous
wave attack. Thus to the east of such places as Brighton and
Worthing the wastage of the cliffs has seriously increased since
the groynes were built to protect these resorts.
When shore drift is in progress along an indented coast,
296
.
SPITS AND HOOKS

spits and bars are constructed as well as beaches. Where the
coast turns in at the entrance to a bay or estuary the material
transported by beach drift and longshore currents is carried
more or less straight on and dropped into the deeper water
beyond. The gradually raised into an
shoal thus started is
embankment. This grows in height by additions from its

landward attachment until a ridge of sand or shingle is built
above sea level in continuity with the shore from which the
additions are contributed. The ridge increases in length by
Fig. 159
Diagram to illustrate the development of a hooked spit by the refraction of
oblique waves
successive additions to its end, until waves or currents from

some other quarter forward growth.
limit its
If the ridge terminates in open water it is called a spit.

Storm waves roll material over to the sheltered side, especially
when they approach squarely, and some spits thus tend to
migrate landwards, often becoming curved in the process.
Curvature is also brought about by the tendency of oblique
waves to swing round the end ( i.e to be refracted) in places
where the sea floor beyond shelves rapidly into deeper water.
A spit may thus be developed into a hook, as indicated in
Fig. 159. Cross currents may assist or modify hook formation,
297
but it is clear that they cannot be essential, for hooked spits

are equally characteristic along the shores of large lakes where
tides are absent and currents are negligible (Plate 71a).
A good example of a curved spit is Spurn Head (Fig. 155),
which extends into the Humber in stream-lined continuity
with the Holderness coast. The latter is almost everywhere
fringed with sand and shingle that drifts steadily from north
Fig. 160 Fig. 161

River Yare, Norfolk River Aide, Suffolk
Examples of river deflection and extension in East Anglia by the southerly

growth of sand and shingle spits. (Both drawn to same scale)
to south, fresh supplies being constantly furnished

by the rapid
erosion of the coast. Most of the transported material is
carried beyond Spurn Head, cumbering the estuary with
shoals on its way towards the Lincolnshire coast, where it is
added to the seaward-growing coastal flats.
Southward drift is also very active along the east coast of
Norfolk and Suffolk. Ten centuries ago the Yarmouth sands
had already spread across the estuary of the Yare, forming an
298
SPITS AND BARS
obstruction which was largely responsible for the silting up
of the Broads. About the year 1016 a barrier spit began to
grow to the south, hugging the coast as closely as possible, and
the river, diverted into a channel between the mainland and
the spit, was obliged to extend itself southwards. By 1347
the end of the spit and the outlet of the river had reached
Lowestoft. Since 1560, however, an artificial outlet has been
maintained at Gorleston, where the spit now terminates, since
the rest of it has long ago drifted south (Fig. 160). At Aldeburgh
Fig. 162
Spits and bars along the south coast of the Baltic
the longest spit on the east coast has similarly diverted the
outlet of the River Aide (Fig. 161).
A bar a spit which extends from one headland to another,
is
or nearly so. If the bay inside is completely enclosed it be-

comes a shore-line lake. More usually, however, a narrow
channel is kept open by tidal scour and outflowing drainage.
Between Danzig and Memel, on the south Baltic coast, there are
two very long bars, surmounted by sand dunes, with extensive
tidal lagoons, called “ haffs,” on the landward side (Fig. 162).
A beautiful example of a bar in Iceland is illustrated in Fig.
163. A similar construction of sand or shingle connecting a
headland to an island, or one island to another, is a connecting
bar or tombolo (Italian).
299
[L. Ha.wkes
Fig. 163
Bar, 10 miles Iona:, across Lon Bav, Iceland
By far the most impressive shingle bar in Britain is Chesil

Beach (Fig. 164). For six miles south-east of Bridport the
Beach fringes the shore. Near Abbotsbury the shore recedes
and for the next eight miles the Beach continues in front of the
tidal lagoon of the Fleet as a bar well over 20 feet in height.
Finally it crosses two miles of sea to the “ Isle ” of Portland,
which is thus tied to the mainland. Chesil Beach is a com-

posite structure, its shingle having accumulated from local
sources as well as by drift from each end. At the north-west
end there are pebbles of rocks from Cornwall and Devon ;
at the south-east end much larger pebbles have been supplied

from the Portland promontory (Plate 71b), where the beach
drift is northwards ;
while between these extremities the
pebbles are relics from the land, now eroded away, that
formerly stretched in front of the Fleet. Although towards the
Portland end the Beach rises to the quite exceptional height
of over 40 feet, the sea sometimes bursts over it during great
storms and pours through breaches into the low-lying area
300
COASTAL TYPES
Bridport
0 I 2
1 i i
Fig. 164
Map of Chesil Beach, Dorset
beyond. Since two villages were demolished in 1824, the worst

disaster of this kind occurred late in 1942, when the railway
between Portland and Weymouth was partly washed away
and the lower parts of Portland itself were seriously flooded.
Shore Lines of Submergence
At the present day shores of submergence are the commonest

types, because of post-glacial changes in sea level. Although
many coasts are actually of compound origin, partial recovery
from former submergence being indicated by raised beaches
(Plate 6), on balance they still remain partially drowned.
Moreover, the occurrence of buried forests locally demonstrates
a quite recent phase of submergence. If a future change of
climate led to the melting of the ice sheets of Greenland and
Antarctica still further submergence would result.
Along coasts of “ Atlantic ” type, that is, where the trend
lines of an orogenic belt are transverse to the coast, drowning
gives an alternation of long promontories and estuaries. The
latter are called rias, the name given them in Spain, where they
occur south of Cape Finisterre. The south-west coast of Ireland
(Cork and Kerry) is a perfect example of this type (Fig. 165).
Along drowned coasts of “ Pacific ” type, where the structural
grain is parallel to the coast, long islands and inlets following
the trend lines are characteristic, as exemplified by the Dal-
301
matian coast of the Adriatic (Fig. 166). Where deeply cut

glaciated valleys have been invaded by the sea after the melting
of the glaciers a fjord type of coast results (Fig. 117), as described
already on page 224.
The first effect of marine erosion on a newly formed coast
of submergence is often to intensify all the initial irregularities
of outline. Where the rocks vary in structure and resistance
the waves pick out all the differences. Soft and fissured rocks
Fig. 165
Atlantic type of coast, south-west Ireland
are worn back into coves and bays, while the harder and more
massive rocks stand out conspicuously. The Dorset coast
north-east of Portland shows this process in active operation.
Here there is a long coastal strip of soft Lower Cretaceous beds,
backed on the landward side by an upland of Chalk, and
formerly protected from the sea by a continuous rampart of
hard upfolded Jurassic limestones. The sea has breached the
latter in places and scooped out the softer rocks behind. The
Stair Hole (Plate 7) illustrates the breaching stage, and Lul-
302
SHORE-LINE DEVELOPMENT
worth Cove (Plate 70a) is a beautiful example of a scooped-out

bay.
Eventually, however, erosion and deposition co-operate to
smooth out all the intricate outlines of a youthful shore line.
As the headlands recede before the concentrated attack of the
waves the stretches of cliffs become longer and straighter.
Spits and bars bridge across the bays and inlets, and gradually
encroach upon them as they keep in line with the retreating
Fig. 166
Pacific type of coast, Yugoslavia
cliffs. Thus protected, the embayments are shoaled up by

additions from streams on the one hand and from wind-blown
sand on the other, until with the aid of salt-marsh vegetation
the coast is built out to coalesce with the outer beach and its
sand dunes. Acoast of smoothly flowing outlines is thus
evolved. Thereafter the shore line continues to retreat as a
whole, and finally all signs of the original embayments may
be obliterated, at which stage the lengthening cliffs become
continuous (Plate 69b).
The barrier reefs and atolls of warm seas where corals
303
flourish are constructions which have been built up under

conditions of submergence. Coral reefs, however, are products
of life and are more appropriately dealt with in the next
chapter.
Shore Lines of Emergence
Typical shore lines of emergence are not common at

present, because they require uplift to an extent sufficient to
overcome the effects of submergence due to recent changes of
sea level. Finland, for example, is steadily rising isostatically,
and the south and south-west coasts are fringed with tens of
thousands of islands as a result of the emergence of the higher
parts of a hummocky ice-moulded surface. But although this
archipelago owes its existence to emergence it is merely part
of a drowned land surface that is now less drowned than
formerly.
A really typical shore line of emergence is one in which
the sea floor with veneer of sediments has been uplifted
its
to form a nearly flat coastal plain with a smoothly flowing

shore line margined by widespread stretches of shallow water.
The Atlantic and Gulf coasts of the south-eastern United States
and the Argentine shore of the River Plate are of this type.
In accordance with the principle illustrated in Fig. 157 loose
sediment from the sea floor is driven landwards in the process
of restoring a profile of equilibrium. But here the waves may
begin to drag the bottom many miles from the low-lying main-
land. As the waves lose energy in crossing the shallows, much
of the sediment is deposited long before the shore is reached,
and thus the foundations of a barrier beach or off-shore bar are
laid. How these constructions come to have their heads raised
above water (Fig. 167) is not yet clearly established. An
ordinary storm beach is formed at the upper landward limit
reached by the waves. But as soon as a growing off-shore bar
has reached a certain level it is swept by breakers, and material
isthen simply transferred from the outer to the inner side.
The bar thus advances towards the shore like a submarine
sand dune, and it cannot become other than a sub-
304
OFF-SHORE BARS
Fro. 167
Barrier beaches and swamps along the coast of North Carolina
marine feature until some other constructive factor begins to

operate.
This additional factor is probably beach drifting. Accord-
ing to this view (which still requires to be verified by observa-
tion) one stretch of a migrating off-shore bar soon reaches an
outward bulge of the coast, and is there constructed into an
ordinary beach. From this point one arm of the bar can then
be built above sea level like a lengthening spit, until it reaches
3°5
the next bulge of the coast and becomes a complete barrier

beach, stretching across a broad shallow embayment which
thus becomes a lagoon. The bar may be breached occasionally
by storm waves, but any such break is sooner or later repaired
by later drift.
Where the bar is in contact with the coast, the latter is
eroded back and the bar, being tied to it, continues to encroach
on the lagoon. Meanwhile, the lagoon itself is silted up with
material thrown over the bar by the waves and brought in at
the head from the land, until it becomes a salt marsh with a
seaward belt of sand dunes. Eventually all these features
continue to be cut back by the sea and a nearly straight
beach-fringed coast is developed. The famous beaches of
Daytona, Palm Beach, and Miami in Florida are off-shore bars
which are now nowhere far from the land and in many places
make contact with it.
Submarine Canyons
More than half a century ago it was discovered by soundings

that the Hudson and Congo valleys continue over the sea floor
as submarine trenches which interrupt the supposedly feature-
less floorof the continental shelf and become comparable in
their dimensions with deep canyons where they traverse the
continental slope beyond. One hundred and twenty miles
south-east of New York the Hudson submarine canyon is now
known to be 50 miles long, 6 miles across from rim to rim,
and 3,700 feet deep (measured from the rim). The Congo
example is on an even vaster scale. These submarine canyons
were naturally a source of great perplexity, and as other
examples came to be found elsewhere the mystery of their
origin presented geologists with a challenge of increasing
urgency.
About 1930 sounding ceased to be a slow and tedious
process with the invention of an instrument that makes it
possible to measure the depth by timing the interval required
for a sound to reach the sea floor and to be reflected back as
an echo. Since then detailed explorations of the sea floor by
306
SUBMARINE CANYONS
echo sounding have been actively carried out in a number

of selected but widely separated regions. Although less than
one per cent, of the continental shelf and slope has so far been
systematically surveyed in this way, more than a hundred
submarine canyons have already been accurately charted.
They occur off coasts of all kinds, and every continent provides
examples, except Antarctica, the margins of which have not
yet been tested. Some are in line with great rivers, but far
more begin on the continental shelf, often far out, without
relation to the drainage from the land, and all of them become
features of vigorous relief on the steeper surfaces of the conti-
Chart of the submarine canyons off the Californian coast between San Francisco
and Los Angeles, showing the maximum depths to which they have been explored
nental slope, which thus turn out to be far more rugged than
anyone could have suspected. A
dendritic system of tributary
valleys, as illustrated in Fig. 168, is characteristic. The result-
ing submarine topography closely resembles that of a land
surface dissected by river erosion. The main canyons are
broad, steep-sided, V-shaped gashes excavated in the sea floor
to depths of as much as 4,000 feet below the rim. Some
examples have been traced to depths of over 10,000 feet below
sea level, but their terminations on the deep ocean floor still
remain unfathomed. Samples dredged from the walls show
that these amazing canyons are geologically quite young.
Pliocene marine beds have been cut through and a coating of
fresh mud suggests that at least some of the canyons are now
307
being filled rather than excavated. It is therefore probable

that the canyons originated during Pleistocene time. But how?
The canyons must have been formed either above or below
the sea. If above, then either the sea floor must have been
raised or the sea level must have fallen by at least 10,000 feet.
Each of these alternatives is equally preposterous. There is
not the slightest evidence that any of the coasts concerned
were pushed up a couple of miles and then restored to their
present positions shortly afterwards, and the idea that this
could have happened all over the world is purely fantastic.
Lowering of the sea by 10,000 feet would imply that about
thirty times as much water was precipitated on the lands in
the form of continental ice sheets as w e have any right to assume.
r
The actual lowering of sea level during the glacial epochs was
of the order 300 feet, an estimate that is confirmed by the
depths of the lagoons of coral atolls and barrier reefs (p. 328).
Thus we are driven to consider the only remaining possibility :
that the canyons were formed under the sea, presumably by

the erosive effects of some kind of submarine current.
Among the many hypotheses that have been proposed to
account for submarine canyons only one, suggested by Daly
in 1936 and subsequently developed, appears to be reasonably
satisfactory. It may be briefly summarized as follows. During
the glacial epochs the continental shelf was everywhere exposed
down to about 300 feet below present sea level. Waves and
currents, then specially strong because of the stormy weather
of glacial times, churned up the muds of the outer part of the
shelf and gave rise to an undertow and bottom layer of un-
usually turbid water. The mud-laden water, being heavier
than the clear water above it, began to flow down the shelf
as a suspension current with a velocity determined by the density
of the suspension and the angle of slope. By guiding and
concentrating the flow of the loaded water chance depressions
would thus become selectively eroded into furrows which
thereafter would canalize and accentuate the currents. On
reaching the continental slope (with a gradient of about 1 in
15, a hundred times steeper than that of the average shelf) the
currents would gather speed and thus gain additional erosive
308
SUSPENSION CURRENTS
[H. S. Bell, U.S. Dept, of Agriculture

Fig. 169
A suspension current of muddy water flowing down a submerged “ delta ” of
sand in a laboratory tank
power. Once started, such submarine streams would be self-
perpetuating, and even self-accelerating, since erosion would

add to themuddiness of the water, and therefore to its effective
density. The
inference is inevitable that on the continental
slopes erosion should have been altogether more vigorous
than on the shelf where the currents were engendered. Sub-
sidiary processes which would co-operate with suspension
currents in the development of the canyons include under-
cutting of the floors and sides by emerging springs, and slump-
ing of loose, water-saturated sediment. Earthquakes might
well act as a trigger to set submarine “ landslips ” moving.
Each of these processes has been regarded as of major im-
portance by certain geologists, but it is difficult to see how
either or both (without the guiding hand of suspension currents)
could be responsible for a submarine topography that bears
all the hall-marks of erosion by running water.
There can be no doubt about the reality and efficacy of the
main process envisaged by Daly. Where the Rhone, milky
with glacial sediment, flows into Lake Geneva, its waters
visibly dive under the clear lake water and flow through a
five-mile trench towards the deepest part of the lake. Moreover,
carefully designed series of tank experiments carried out by
(396)
309 21
[//. S. Bell, U.S. Dept, of Agriculture

Fig. 170
A suspension current of muddy water approaching the dam in a laboratory
tank
Kuenen in Holland, and independently by Bell in the United

States, have demonstrated that suspension currents do, in
fact, behave as described by Daly (Figs. 169 and 170 ). The
only doubt still remaining is whether these currents could
erode resistant rocks, such as have been collected by dredging
from the walls of some of the canyons, or whether this feature
requires some additional explanation.

D. W. Johnson
Shore Processes and Shoreline Development. Wiley and Son, New York,
I
9I4 -
E. M. Ward
English Coastal Evolution. Methuen, London, 1922.
R. A. Daly
The Floor of the Ocean. University of North Carolina Press, Chapel
Hill, N.C., 1942.
E. J. Conway
Mean Geochemical Data in Relation to Oceanic Evolution
; and The
Chemical Evolution of the Ocean. Proceedings
of the Royal Irish
Academy, Vol. LXVIII b, No. 8, 1942 and No. 9, 1943.
;
310
1
Chapter XV
LIFE AS A ROCK BUILDER
Life as a Geological Agent
In earlier chapters certain aspects of the geological work

accomplished by living organisms have already been touched
on : the breaking up of the sub-soil by roots the growth of
;
soils ;
the protection of soils by forests and prairie grasses ;
the fixation of sand dunes and the

;
comminution of materials
by worms and other burrowing animals. Apart from man,
who at the present day is contributing to erosion and transport
on a gigantic scale, organisms are of limited importance as
rock breakers, but in virtue of their biochemical activities they
contribute on a very considerable scale to the chemical weather-
ing of rocks.
In particular it is worth noticing that practically all the
oxygen now present in the atmosphere and all the oxygen
that has been used up in weathering processes has been liberated
from carbon dioxide by green plants during their growth
(p. 330). The only other known source of free oxygen is due
to the action of certain bacteria which liberate oxygen from
nitrates, which they turn into ammonia, or from sulphates,
which they reduce to sulphides (such as pyrite) or even to
sulphur itself. Since oxygen has accumulated throughout
geological time as a by-product of plant life, despite the
enormous amount used up in oxidation (including breathing),
it follows that there must be an equivalent amount of carbon
elsewhere. Part of this is present in the living tissues of the

plants and animals of the present day. Both on land and in
the sea certain animals eat plants and other animals devour
the plant-eaters. Thus a great deal of organic matter is
simply transferred from one organism to another. Of the
balance, part decays, largely as a result of bacterial attack,
and passes away as carbon dioxide and water ; while
31
the remainder, in various states of preservation according

to the stage of decomposition reached before and after
burial, is found in soils and swamps, in carbonaceous and
bituminous shales, and in more concentrated deposits of

peat, coal, and petroleum. Natural fuels and the oxygen
required for their combustion are two of the most im-
portant end products of vital activities (see Fig. 177). This
aspect of the work of life forms the subject of the following
chapter.
Besides the accumulations of organic matter mentioned
above, there are immensely greater deposits which are largely
composed of the shells or other protective and supporting
structures of once-living organisms.Most of these hard parts
consist of calcium carbonate secreted from sea
essentially
water by animals such as molluscs (commonly known as shell-
fish), and the tiny single-celled foraminifera
sea urchins, corals,
(Plate 72a and and by plants of which algae (a group
b),
including seaweeds) are the chief. After death, the hard parts
persist and accumulate as shell deposits (Plate 9b), coral reefs,
deep-sea oozes, and the like all raw materials of limestones
;
in the making. Other single-celled animals and plants, known

respectively as radiolarians (Plate 72c) and diatoms, extract
silica, and encase themselves within microscopic shells of opal.
The resulting siliceous deposits constitute two important

varieties of the deep-sea oozes. Other organisms, such as
fresh-water mussels and snails, and green algae, contribute
calcareous materials to deposits forming in fresh-water lakes,
lagoons, and estuaries and fresh-water diatoms, which abound
;
in the lakes of glaciated districts, similarly provide siliceous

deposits known as diatomaceous earth. But the marine examples
are of enormously greater abundance.
Bacteria also help to form limestones by what may be
called biochemical precipitation. It was mentioned above that
certain bacteria turn nitrogen compounds into ammonia.
The latter has the effect of precipitating
calcium carbonate
from sea water, and on the shallow banks between Cuba and
the Bahamas considerable deposits are now being formed in
this way. Bacteria are also concerned in the precipitation of
3!2
ZONES OF SEDIMENTATION
limonite from the waters of lakes and marshes, notably in

Sweden and Finland, where bog iron ore of this kind has been
extracted on a commercial scale.
Phosphates are of great value because of their vital im-
portance as and because workable deposits are far
fertilizers,
less common than we could wish. Most of them have resulted
directly or indirectly from organic activities. Calcium phos-
phate is particularly concentrated in the bones, teeth, and
excrement of vertebrates, especially fishes. Vast numbers of
fish are sometimes killed by the shock of earthquake vibrations
passing through sea water, and their remains then accumulate
as bone beds. Another source is provided by the droppings of
countless generations of fish-eating birds on coasts and oceanic
islands, which thus become thickly encrusted with guano .
Marine Deposits
According to their location on the sea floor (Fig. 171),

marine deposits are classified as follows. Littoral deposits are
those formed between high and low tides. Shallow water
deposits are those which collect on the continental shelf and
LittoralZone
A ‘Shallow-wafer, Deep-waJerZont. ^
Land
‘
:
i
—
Fffprttvp simlieht-Rph/htnf Plankton
E/feCfiY£.mllgbtr.Mlf3ro(P/wko(l.Z
Continental Shelf
Twilight Zone
CONTINENTAL PLATFORM
. %
%< Completely Dark
-8
—8
I
Is
§
-1° J
~~
l
Abyssal Zone -\2
OCEANIC PLATFORM
Fig. 171
Schematic section to show the zones of marine sedimentation
at similar depths elsewhere, that is, between low tide and 100
fathoms. Below this level are the muds and oozes referred to
as deep sea or deep water deposits. The muds, etc., of the conti-
nental slope, and of similar depths around oceanic islands,
313
belong to the bathyal zone ; while the oozes of the deep ocean
floor belong to the abyssal zone.
According to the source of the materials, marine deposits
fall into two main groups : terrigenous —derived from the land
by river transport and coast erosion ;
—
and organic comprising
the calcareous and siliceous shells and other remains of marine
organisms. The terrigenous deposits are naturally found in
greatest bulk bordering the lands. Sediment that is swept
over the edge of the continental shelf comes to rest on the
slopes beyond. The abyssal ocean floor receives supplies of
land detritus only from wind-borne dust and rare icebergs.
These sources of supply are so scanty that the rate of accumula-
tion is extremely slow.
The marine organisms that contribute most conspicuously
to the sediments of the littoral and shallow water zones belong
to a group known collectively as the Benthos (bottom dwellers).
This includes seaweeds, molluscs, sea urchins and corals, and
other forms that live on the sea floor. Many of them are
firmly attached to the bottom. Deposits of shells or of their
wave-concentrated fragments are formed in great abundance
in favourable situations, while elsewhere similar remains are
dispersed as fossils through the terrigenous deposits. The
North Sea is mainly floored with terrigenous material, but off
the Kentish coast and between the Thames and the Hook of
Holland there are patches of several square miles consisting
almost entirely of large shells. The shelly sands of some of the
Cornish beaches have already been mentioned (p. 293). These
are but relatively small examples of shelly limestones in the
making. Far more extensive accumulations occur off limestone
coasts and in other situations where the organic remains are
not smothered by sand and mud. The reefs and atolls built
up by corals and their associates in the shallow water of warm,
uncontaminated seas illustrate limestone-building on so
spectacular a scale that they are reserved for a more detailed
description.
Organic deposits of the above types which have accumu-
lated off continental shores or on the flanks of oceanic islands
are described as neritic (Gr. neritos, a mussel). The organic
3X4
PLANKTON AND NEKTON
oozes and red clay of the abyssal zone are distinguished as
The oozes are largely
pelagic deposits (Gr. pelagos, the sea).
composed of the remains of marine organisms belonging to a
group called the Plankton (the wanderers). This includes the
single-celled marine plants (diatoms) and animals (foraminifera
and radiolarians) certain floating molluscs known as “ sea
;
butterflies ” or pteropods ; most of the eggs and larvae of the

benthos and other marine organisms ; and all other forms
which, unlike fishes, have no means of self-locomotion. The
pteropods are blown along the surface by the wind, but the
others, nearly all microscopically small, are passively sus-
pended in the water. Diatoms, being plants, cannot live
below the depth of effective sunlight penetration, which in the
open ocean reaches a maximum of about 650 feet. Though
individually quite invisible to the unaided eye, the diatoms are
present in such prodigious numbers that they turn the sea in
which they live into a kind of thin vegetable soup. This
forms the main food supply of the rest of the plankton, whose
habitat is therefore similarly confined to the sunlit zone.
From this prolific overhead source the sea floor receives a
slow but steady rain of plankton shells which have escaped
destruction by being eaten or by being dissolved in the sea
water as they sank. In the shallow-water zone the tiny shells
are generally lost in an overwhelming abundance of terrigenous
and neritic materials. In the bathyal zone, where the rate of
supply of terrigenous sediment is less overpowering, they make
a bigger show and can be readily found in the blue and green
muds, both of which are characteristically calcareous deposits.
In the abyssal zone, however, the plankton shells accumulate
with little contamination from other sources, to form the deep-
sea oozes which, together with the red clay, constitute the
pelagic deposits.
Fishes, whales, and other marine animals which go actively
after their food supply are grouped as the Nekton (swimmers).
These contributeto all the marine deposits on a limited scale,
but concentrated remains, such as the bone beds already
mentioned as sources of phosphates, are quite rare.
The adjoining table summarizes the leading types of deposits
3*5
life as a rock builder
now forming on the sea floor, classified according to the zones

of deposition and the sources of material as reviewed in the
preceding paragraphs. A
systematic description of all these
deposits would require more space than is available, but coral
reefs and the chief varieties of the pelagic deposits will serve as
illustrative examples of special interest.
MODERN MARINE DEPOSITS

K\ Kinds of Organic Deposits
Chemical
Material j
Terrigenous and
Deposits Biochemical Neritic Pelagic
Zones of\ Precipitates (Mainly (Mainly
Depsition \ Benthos) Plankton)
Littoral Zone Shingle Oolite sands Shell gravels
Gravel Calcareous Shell sands

muds
Shallow- Sands Coral reefs
Water Zone Cementing and
Muds materials Coral sands
< (10%)- >
Deep-Water Deep Sea Muds Coral muds Deep-Sea Oozes
Zone of the of the
Bathyal zone Abyssal zone
Green, black Cementing Pteropod ooze
and blue materials
muds Globigerina ooze
Volcanic muds
(With varying Diatom ooze
amounts of
plankton re- Radiolarlan ooze
mains) (41%)
(15%)
Insoluble Residues
from various sources
i
Red Clay (34%)
The figures in brackets represent the approximate areas covered by the various
groups of deposits, expressed as percentages of the area of the ocean floor
Pelagic Deposits
The composition and of the deep-sea oozes

distribution
depend on the temperature of the surface waters and the depth
of the underlying ocean floor. Diatoms, for example, flourish
316
PELAGIC DEPOSITS
particularly well in cold circumpolar seas which are un-

favourable to a profusion of other forms. Radiolarians, on the
other hand, are especially abundant in warm tropical waters.
Foraminifera, of which Globigerina is the commonest genus,
abound in tropical and temperate regions and the distribution
of globigerina ooze is therefore less restricted. The tiny cal-
careous and opaline shells, being extremely delicate and often
of intricate design, readily lend themselves to attack by solution
as they sink towards the bottom. The solvent power of sea-
water increases with depth : directly because of the increasing
pressure and indirectly because the proportion of gases in
solution, mainly carbon dioxide, also becomes higher as the
pressure rises and the temperature falls. The globigerina and
other calcareous foraminifera dissolve more rapidly than the
radiolarians and other siliceous forms. Only a few of the larger
varieties succeed in reaching depths of 3,000 fathoms, and in
the next 500 fathoms even these are lost. As a well individua-
lised deposit globigerina ooze is most characteristically developed
at medium depths on the ocean floor, that is, round about
2,000 fathoms. The siliceous remains persist to greater depths,
some of them down to 5,000 fathoms, the average for radiolarian
ooze being about 3,000.
Over 50 million square miles of the ocean floor lies beyond
the reach of more than traces of the plankton remains. Here
accumulate the materials that form the basis of the Red Clay :
(a) volcanic products from wind-borne dust, submarine

eruptions, and fragments of pumice that have floated far from
their source before sinking ; ( b ) non- volcanic wind-borne dust ;
(c) insoluble organic relics like shark’s teeth and the ear-bones
of whales ;
(d) the dust of meteorites and occasional larger
fragments which have fallen into the sea from the sky ;
and
locally, (e) debris dropped from far-travelled icebergs. Meteo-
rite dust falls everywhere, of course, but it is only in the red
clay that it is not smothered beyond recognition. The red
clay accumulates so slowly that some of the shark’s teeth lying
unburied on the surface are those of species now extinct.
Many of the ingredients listed above have decomposed into
clay, heavily stained by ferruginous matter which gives the
deposit a brick-red or chocolate-brown colour. Other

secondary products found in the red clay include black
nodules of manganese oxide and crystals of various silicate
and other minerals. Where the depth is not excessive un-
dissolved plankton remains appear, and as the proportion of
these increases with decreasing depth the red clay ingredients
become more and more diluted until the deposits pass into
radiolarian ooze in some or into globigerina ooze in
localities
others.
In the following table the average composition of each of
is given
the chief pelagic deposits :
Globigerina Radiolarian Diatom

Constituents Ooze Red Clay Ooze Ooze
Calcareous remains 64 6 4 23
O 2 54 41
Siliceous remains
Mineral matter 34 92 42 36
Area in millions of
square miles . 49-5 51-5 2-3 11
Radiolarian Ooze (Plate 72c) is essentially a variety of red
clay that notably rich in the remains of radiolarians. Diatoms

is
and sponge spicules are also commonly present. It occurs

beneath the tropical belts of the Pacific and Indian oceans,
where the warm surface waters favour an unusual profusion
of radiolarians, and where the great depth prevents more
than a scanty supply of calcareous remains.
Although they occur everywhere, diatoms are most abun-
dant relative to the rest of the plankton around Antarctica and
in the extreme north of the Pacific. The delicately ornamented
shells accumulate there, as Diatom Ooze, together with certain
species of foraminifera that can live in cold water. Con-
tamination by terrigenous material, much of which is derived
from floating ice, is commonly abundant.
Globigerina Ooze is by far the most widespread variety and
is especially characteristic of the Atlantic floor. It contains
318
SAMPLES OF THE DEEP-SEA FLOOR
foraminifera in great abundance and variety (Plate 72a),

among which the shells of Globigerina, each consisting of several
globular chambers, are the commonest (Plate 72b). The
proportion of calcium carbonate varies with the depth, ranging
from 97 per cent, in samples from 400 fathoms to 30 per cent,
in those from 2,600 fathoms, the balance being made up of
other plankton remains, and ingredients like those of the red
clay, into which the globigerina ooze merges at its lower
margins. In the shallower depths, over sub-tropical and
tropical submarine banks and ridges, the shells of pteropods
locally become abundant, and when they predominate the
deposit is distinguished as Pteropod ooze. Another important
variety is characterised by the button-shaped remains (cocco-
liths) of certain lime-secreting algae which flourish in sub-
tropical waters.
Geological Interest of the Pelagic Deposits
A century ago nothing was known of the deep-sea deposits.

Globigerina ooze, dredged up by one of the cable-laying
steamers in 1852, was the first to be discovered. A
systematic
exploration of the ocean floor was carried out by the
famous Challenger Expedition of the years 1872-76. The
thousands of samples then brought up are described in
one of the fifty bulky volumes in which the scientific
results of that great enterprise are recorded. Since then
certain parts of the ocean floor have been investigated
in greater detail, but only in recent years has any really
fundamental advance been made.
In 1934 Piggot successfully developed a sounding apparatus
for securing core samples of abyssal deposits. The apparatus
metal tube which, on arriving at the bottom,
consists of a long
is automatically forced into the sediment by an explosion of
cannon powder contained in a kind of gun attached to its

upper end. During 1936 cores up to ten feet long were ob-
tained along a traverse across the North Atlantic from New-
foundland to Ireland ;
seven of these came from depths
exceeding 2,000 fathoms. So slowly has the geological record
3*9
accumulated in these quiet depths that the cores represent at
least a million years of oceanic deposition. Examination of
the sequence shows that the four glacial epochs are all
clearly represented. The foraminifera of the glacial bands
are, moreover, found to be of types that live in cold waters,
whereas those of the interglacial bands indicate that the
surface waters were then as warm as, or even warmer than,
those of to-day.
A is that no representatives of the
fact of great significance
abyssal deposits of former ages have been found among the
formations now exposed on the continent, except in certain
marginal islands. The Chalk might appear to be an exception,
for it is a limestone that in some respects resembles a highly
calcareous globigerina ooze. But the Chalk foraminifera are
largely shallow-water forms, and with them are associated
many larger fossils of the benthos group. These all have
characteristically thick such as are grown only by
shells,
organisms that have to withstand the turmoil of vigorous

waves. Thus the Chalk is not an abyssal deposit. The lands
surrounding the Chalk sea were so low-lying that they provided
little or no terrigenous material to the sea floor outside the
littoral zone, and in consequence a thick uncontaminated

limestone, characterised by an unusual abundance of fora-
minifera, was able to accumulate in the thirty million years
or so during which these exceptional conditions lasted.
In the Dutch East Indies Jurassic and Cretaceous formations
have been discovered which contain alternating layers of
indurated red clay and radiolarian ooze, associated with
shallow-water and bathyal sediments. Similar uplifted de-
posits, accompanied by globigerina ooze, occur in Barbados
in the West Indies. Here they lie on Tertiary continental
sediments, including coal seams, thus presenting “ clear
evidence that portions of a continental area might be de-
pressed to oceanic depths and re-elevated.” Evidence of this
kind, however, is confined to a few tracts along the folded
margins of the continents. There is nowhere any indication
that the oceanic platform itself has become part of a continental
region.
320
CORAL REEFS
Coral Reefs and Atolls

In favourable situations in tropical seas corals, together
with all the organisms to which they give shelter and attach-
ment, grow in such profusion that they build up reefs and
Fig. 172
Coral reefs. Fringing and barrier reefs of Mayotta, Comoro Is., north end of
Mozambique Channel. The outlines of the islands suggest recent submergence
islands of very considerable size. Clothed in vivid green,

crowned by the coconut palm, and fringed with the white
foam of the ceaseless surf, the “ low islands ” of the Elizabethan
321
mariners have a reputation for dazzling but treacherous beauty.

Dangerous to navigation and difficult to explore, they have
been equally tantalising to geologists who sought to account
for their existence. Darwin was the first to face the problems
in a scientific spirit and by him coral reefs were divided into
three main classes (Figs. 172-174) :
(a) Fringing reefs, consisting of a veneer or platform of coral

which at low tide is seen to be in continuity with the shore.
The width is often half a mile or more, and the seaward side
slopes steeply down to the normal sea floor.
SW NE
Section to illustrate the relationship of the Great Barrier Reef to the coast of
Queensland {After J. A. Steers)
Reef rock, black lagoon and channel sediments, dotted
;
(b) Barrier reefs, situated up to several miles off-shore, with

an intervening lagoon. The thousand-mile complex of reefs
known as the Great Barrier Reef, which forms a gigantic
natural breakwater off the north-east coast of Australia, is by
far the greatest coral structure in the world (Plate 73a and
Fig. 173). Most barrier reefs, however, of which there are
countless examples, are island-encircling structures forming
irregular rings of variable width, more or less interrupted by
open passages on the leeward side (Fig. 172).
(c) Atolls, resembling barrier
but without the central
reefs,
island (Plate 73b and Fig. They
are essentially low-
174).
lying ring-shaped islands enclosing a lagoon which again
is generally connected with the open sea by passages on the
leeward side.
Reef-building coral live in colonies of thousands of tiny
individuals (polyps), each occupying a cup-shaped depression
322
GROWTH OF CORAL ROCK
Map of the Suva Diva atoll, Indian Ocean (50 X 40 miles), showing the depths
of the lagoon floor in fathoms [After R. A. Daly)
in a calcareous framework which is common to the whole

colony. As the successive generations of corals grow outwards
through the restless waters in their competition for food, the
stony framework also branches upwards and outwards and
grows into forms that resemble plants, some being like shrubs
and others like cushioned rock-plants. The interspaces
between the dead coralline structures are cemented and bound
together by calcareous algae called nullipores. These pre-
cipitate calcium carbonate within themselves, and still more as
incrustations which coat their surfaces and cover the coral
growths to which they are attached. Other contributions are
made by shelled molluscs, foraminifera, calcareous worms, and
323
bacteria, and the whole assemblage accumulates to form a

white porous limestone which gradually becomes more
coherent as it is buried and subjected to prolonged saturation
by sea water.
The development and maintenance of coral reefs depend
upon the conditions that favour a vigorous growth of the living
colonies. A thriving reef has to contend not only with the
waves, but also with boring organisms and voracious crustaceans
that feed on the bodies of the individual corals. The reef re-
presents the margin of success in a never-ceasing struggle
against death and extinction. Not only have the corals and
nullipores to supply material to maintain a flourishing living
face, they have also to provide the broken masses of coral rock
and other debris that accumulate to form the visible reef and
its seaward foundations. On the lagoon or landward side of
the living face there the reefflat, consisting of material thrown
is
up by the breakers to a height of 1 0-15 feet. A certain amount

of debris is also washed over into the lagoon by heavy seas that
sweep the reef. On the seaward side of the growing face the
reef passes into a talus slope thatmay descend to very great
depths on the flanks of oceanic barrier reefs and atolls.
Corals require a mean temperature of not less than 68° F.,
and reefs and atolls are therefore restricted to a zone lying
between latitudes 30° N. and S., except locally where warm
currents carry higher temperatures to the north or south of
these limits. The reefs of the Bermudas, for example, are
dependent upon the warmth of the Gulf Stream. Along the
torrid belts of the oceans the equatorial currents drift towards
the west, becoming warmer on the way, and consequently
reefs flourish far more successfully in the western parts of the
oceans than on their colder eastern shores.
The water must be clear and salt. Opposite the mouths of
rivers, where the diluted sea-water carries suspended silt and
mud, corals cannot live and no reefs appear. Conversely, reefs
grow best on the seaward edge of the reef, where splashing
waves, rising tides, and warm currents bring them constantly
renewed supplies of oxygen and food. Corals cannot long
survive exposure above the water, and consequently living reefs
324
(A) Foraminifera from Giobigerma Ooze, .South Atlantic. X 30
K. M. Cratg
(B) Globigcrina from Ooze, Porcupine Bank, we^t of Ireland. X 25
Ot Ci
tf
•Ovk I “- v w gk <s\0* '

c^y
^ 4 ^
C w0
*" r-i <tM
« # \5ffP ?ojfc' ••%
^ oz, o‘--/ i*
* r-*\i
-
feA
-
,
» »t-'c®c-v
^ '
*f>' +Q
*2&?% -J"*
>«W5jj
•* ’
-^1
^ -w4 S' -J2

'
<0 A r>
A\ .U.
(C) Radiolaria from Ooze, South Pacific.
PLATE 73
ORIGIN OF CORAL REEFS
can never grow much above low-tide level. Dead reefs are
found above sea level, but they have been uplifted into such
positions by earth movements, to which they are therefore a
most reliable index. On the other hand, reef-building corals
require sunlight and do not grow freely at depths greater
than about 25 fathoms nullipores are similarly restricted to
;
about 50 fathoms. A necessity for reef formation is therefore

the pre-existence of a platform not far below sea level. Reefs
and atolls may be “ drowned ” by rapid subsidence, and
several examples of reefs and atolls that have been killed off
in this way have been discovered on the sea floor. It follows
from the above considerations that the living corals and the
growing face of the reef tend to spread upwards and outwards
towards the surface waters of the open sea.
The Origin of Barrier Reefs and Atolls

The origin of fringing reefs is easy to understand. Minute
coral larvae drift with the ocean currents, and those that reach
suitable shores find attachment and start new reefs that
gradually develop seawards. Barrier reefs and atolls, however,
are remarkable in that they generally rise from depths where
no corals or nullipores could live. There are two possibilities :
either the reefs have grown upwards from submerged banks

not more than, say, 50 fathoms below the surface, or they have
grown upwards and outwards from fringing reefs during the
submergence of the land or island to which they were originally
attached. Another feature that calls for explanation is that
the lagoons have nearly flat floors and depths that are all of
the same order, the range being from about 45 fathoms for the
larger examples, which may be several miles across, to 25
fathoms for the smaller ones.
The first general explanation was offered by Darwin as a
result of the observations he made during his celebrated voyage
in the Beagle. He visualized all reefs and atolls as different
stages in a single process (Fig. 175). Growth begins with the
building of a fringing reef around, let us say, a volcanic island.
(396) 325 22
Subsidence of the island combined with continuous growth

converts the reef into a barrier reef. Since corals can grow
upwards at a rate of about a foot in ten years it will rarely
happen that they are unable to keep pace with the movement.
The submerged area between the island and the rim of coral
rock forms the lagoon. By further subsidence the summit of
the central island sinks out of sight, and the barrier reef be-
comes an atoll.
Darwin’s simple theory has not passed unchallenged, but
it satisfactorily accounts for most of the features associated with
reefs. —
The reality of subsidence or at least of a change of
—
sea level is proved by the drowned valleys and embayed
Fig. 175
Diagram to illustrate Darwin’s theory of the successive development of fringing
reef, barrier reef, and atoll around a subsiding island
shore lines of the land inside the lagoons of barrier reefs. The
Great Barrier Reef has grown on the edge of a down-faulted
area, which was formerly the coastal plain of Queensland and
part of New South Wales. Uplifted atolls in Timor and else-
where are found to lie unconformably on an eroded foundation,
exactly as the theory requires. The theory does not, however,
make it clear how the lagoons of the present day have come
to be so remarkably uniform in depth. Fig. 175 shows the
enormous quantity of lagoon sediment necessary to fill in the
“ moat ” around a subsiding volcanic island. Alternatively,
the flat lagoon floors of atolls and island-encircling barrier
reefs suggest that the corals grew upwards from the edges of
submerged platforms worn down by marine erosion.
In 1910 Daly showed that these features are an inevitable
result of recent and Pleistocene changes of climate and sea
326
GLACIAL CONTROL THEORY
level. He had —
already noticed the narrowness and therefore
the youthfulness — of the reefs fringing the Hawaiian Islands.
Connecting this youthfulness with the discovery that a former
glacier had left its traces on the flanks of Mauna Kea, he came
to the conclusion that corals could not have flourished along
those shores during the glacial epochs and that the existing
reefs must have grown there during post-glacial time. During
the glacial epochs the fall of temperature must have killed off
most of the pre-existing reef-builders, leaving only a few
sheltered spots from which the active reefs of the interglacial
stages, and finally those of the present day, could be colonised.
Moreover, during the height of each glaciation the level of the
oceans must have been about 300 feet or 50 fathoms lower than
to-day. As a result of the lowered sea level, pre-glacial islands
and reefs would be steadily attacked by the waves, and in many
places reduced to platforms of marine erosion near, or a few
fathoms below, the sea level of the time.
Thus, innumerable platforms— many of them being the
—
truncated summits of oceanic volcanoes were formed at about
the right depth to account for the existing floors (Fig. 176).
The latest colonisation of the platforms and the upbuilding of
the encircling reefs by corals present no difficulty. It is about
25,000 years since the melting of the ice locked up in the
continental ice-sheets of Europe and North America began to
restore to the oceans the water previously abstracted. With
a growth rate of a foot or so in ten years the corals could readily
keep pace not only with the rising sea level, but also with the
necessity to provide material for the wave-built reef-flats and
for the talus slopes on the seaward flanks.
The lagoons must, of course, have been somewhat shallowed
by deposition. The smaller the lagoon the more rapidly its
floor would be built up by sedimentation, because of the
proportionately greater length of reef across which debris
could be washed. This consideration is matched by the
observed fact that the lagoon depths increase as the widths
increase. Submerged platforms in the colder oceanic regions
where corals failed to gain a footing have received very little
sediment and their depths are correspondingly greater.
327
BANKS
MACCLESFIELD
Reef
,T ILADUMM AT I
l BANKS WITH
?S Reefs
[
INSET REEFS
J MIL AD U MM A D U LU
o 10
o
BARRIER
REEFS
Reef ^4 34 ?5
PEROS BANHOS
O 10
?3
[l;;3i!jjjS!|'F'uN Af ur j
\atous
Fig. 176
Sections demonstrating the flatness of the floors of reef lagoons and the close
similarity of their depths to those of submerged banks
( From R. A. Daly, “ The Floor of the Ocean f by permission)
Darwin’s subsidence theory leaves the origin of the plat-

forms unexplained. Daly’s “ Glacial-Control ” theory pro-
vides an adequate explanation. In all regions the rise of sea
level since the withdrawal of the ice sheets has been a definite
factor in the development of coral reefs and atolls. But
wherever earth movements have been in operation they too
have been a factor that cannot be ignored. Their importance
”
is clearly indicated by the occurrence of both “ drowned
and elevated reefs in regions such as the outer margin of the

Dutch East Indies, where orogenic movements are now or have
328
CORAL REEFS
recently been in progress. In Timor uplifted atolls furnish
evidence that subsidence and coral growth were succeeded by
upheaval and coral extinction. Moreover, it seems probable
that many volcanic islands may have subsided in consequence
either of isostatic readjustment, or of contraction of the under-
lying foundations.
Darwin’s theory refers to submergence by earth move-
ments, while Daly’s refers to submergence by a rising sea level.
Neither by itself provides an all-embracing explanation of coral
reefs, but together, as a complementary pair, they solve all the
major problems.

J. Murray and J. Hjort
The Depths of the Ocean. Macmillan Co., New York, 1912.
J. Johnstone
An Introduction to Oceanography. Liverpool University Press, 1928.
C. Darwin
The Structure and Distribution of Coral Reefs. Smith Elder and Co.,
London, 1842.
W. M. Davis
The Coral Reef Problem. American Geographical Society, Special
Publication No. g, New York, 1928.
R. A. Daly
The Changing World of the Ice Age. Yale University Press, New Haven,
Conn., 1934.
F. H. Hatch and R. H. Rastall (
revised by M. Black)
The Petrology of the Sedimentary Rocks. Allen and Unwin, London,
I938-
^9
Chapter XVI
LIFE AS A FUEL MAKER : COAL AND OIL

The Sources of Natural Fuels
Carbon dioxide the primary source of the carbon compounds
is
of all living organisms and of all those that have lived in past
ages. Under the influence of the sun’s rays green plants, includ-
ing most of the bacteria, synthesize carbon dioxide and water
into carbohydrates, such as cellulose and starch, both
(C 6 H O
10 5 ).,, and
sugar (C^H^On). Since these compounds
Fig. 177
Diagram to illustrate the Carbon Dioxide -Oxygen Cycle and its by-products
are equivalent to carbon and water, their formation involves

liberation of the oxygen which was originally combined with
the carbon. Some of the oxygen so set free recombines with
the carbon of organic matter, both living and dead another ;
part is used up in weathering processes ;

and the balance
passes into the atmosphere or into the sea. The cycle of changes
isschematically summarized in Fig. 177.
If all the decaying remains of dead organisms were com-
pletely oxidized there would be no free oxygen left over.
Under water-logged conditions, however, oxidation is not
complete. The decomposition products of vegetation, for
example, accumulate as humus in the soil and as deposits of
330
ORGANIC CARBON
peat in bogs and swamps. The buried peat deposits of former

ages have been transformed into coal seams. In marine sedi-
ments a high proportion of the organic material of plant and
animal life is either eaten or lost by oxidation, but some
escapes complete destruction and is entrapped in muddy
deposits to form the minute droplets of oil and bubbles of gas
which are the source materials of the concentrated “ pools ” of
petroleum and natural gas found in oilfields. Life is thus
responsible for all the natural fuels, including wood, peat,
coal, oil, and gas, andfor the enormously greater amount of
carbonaceous and bituminous matter that is dispersed through
shales and other sedimentary rocks.
It is of interest to attempt an approximate balance-sheet
between the carbon of organic matter and the oxygen comple-
mentary to it. The following table gives some idea of the
prodigious amounts involved, and bears eloquent witness to
the work of countless generations of countless millions of plants
and animals.
Carbon in Millions of tons Oxygen Millions of tons
Living matter . . . 700,000 In the air 1,223,500,000
Soil 400,000
Peat 1,200,000 Added to weathering
Lignite and Brown Coal 2,100,000 products and now
Bituminous Coal . . 3,200,000 in sediments 8,000,000,000*
....
. . .
Anthracite 600,000
Ordinary Sediments . 4,576,000,000* Dissolved in the ocean 12,000,000
Total . . 4,584,200,000 Total . . 9,235,500,000
The figures marked with an asterisk are only rough estimates,

as it is impossible to assess with accuracy either the total mass
of sedimentary rocks or their average composition. Never-
all
theless, the totals forcarbon and oxygen turn out to be of the
same order, with oxygen in excess, as it should be if both
have been derived from carbon dioxide. No figure is given
above for the carbon in oil and natural gas, as any reasonable
guess as to ultimate resources would be quite negligible in
comparison with the figures for the other repositories of carbon.
So far 13,000 million tons of oil have been discovered, and half
of this total has already been extracted from oilfields and used.
33 1
Peat
The development of vegetable mould and humus from

decaying plant remains has already been mentioned in con-
nection with the soil (p. 122). Dead plant debris is at once
attacked by bacteria and fungi, and in the presence of sunlight
and oxygen the cellulose which is the basic substance of nearly
all plant tissues rapidly disappears, the ultimate products being
carbon dioxide and water :
+ 60 2 = 6C0 2 -f- 5H 20
Cellulose Oxygen Carbon Dioxide Water
But in water-logged environments, such as bogs and swamps,

the degree of putrefaction is limited by the paucity or absence
of oxygen and the generation of antiseptic organic acids which
inhibit the bacterial activities. Under these conditions the
softer and macerated plant debris changes into a dark
finely
brown jelly-like humus. Part of this soaks into the cells of
fragments of wood, bark, roots, twigs, etc., which are also
being “ humified,” and the cellular structures of these remains
are in consequence often wonderfully well preserved. All
the humified products, together with a variable proportion of
the less destructible materials, such as resin and the waxy
pollen cases and spores, accumulate to form deposits of
peat.
The process of humification enriches the residue in carbon,
as indicated by the following equation, which approximately
represents the chemical changes involved :
2C 6 H 10O5 = C8 H 10 O 6 + 2C0 2 + 2CH 4 + H0

2
Cellulose Humified residue Methane
Methane, more familiarly known as marsh gas, is highly
inflammable, and pale flames are responsible for the “ will-
its
o’-the-wisp ” which is occasionally seen flickering over the

surface of a bog. In coal mines, where the gas sometimes
escapes in disastrous quantities from the coal face, it is the
chief constituent of the dreaded fire damp.
The vegetation which contributes to peat formation ranges
332
PEAT BOGS AND SWAMPS
—
from mosses (bog peat) to trees (forest peat Plate 74a), and
the environment may be a swampy lowland or a water-logged
upland with imperfect drainage. The climate must therefore
be humid and the conditions such that growth exceeds
wastage. In the bogs of cool humid regions the rate of
decay lags behind because of the low temperature, whereas
in the densely forested swamps of tropical regions the pheno-
menal rapidity of growth more than keeps pace with the high
rate of decay.
A special variety of peat accumulates at the bottom of lakes
and pools surrounded by marsh vegetation. Wind-blown
pollen and leaves fall into the water, and all manner of organic
particles drift into it. Eventually these settle down to form a
layer of organic ooze. Fresh- water algae may add further
contributions. Locally the spores and algal remains may
predominate, giving rise to a deposit that is specially rich in
the waxy and oily ingredients of vegetation. If streams are
flowing into the water the ooze is likely to be contaminated
by a certain amount of muddy sediment.
Many thousands of the shallow lakes that formerly occupied
depressions in areas strewn with glacial deposits have been
converted into peat bogs by the steady encroachment of marsh
and swamp vegetation, and in others the process of infilling
is still in progress. The rushes, reeds, and pond weeds gradually
advance over the dark gelatinous slime formed from the resi-
dues of earlier generations. Floating vegetation sometimes
grows out in thick spongy rafts across the surface. Meanwhile
the floor is being built up as organic ooze accumulates, and
finally the site of the lake becomes a swamp. The treacherous
surface may be covered with quaking tussocks of sphagnum
moss, as in the bogs of Ireland. Where the drainage con-
ditions are suitable, the plant sequence may culminate in a
forest of trees with roots adapted to the precarious foundation
through which they spread.
On a more extensive scale swamps are developed from
the shallow lagoons and lakes of low-lying coastal plains, flood
plains, and deltas. The Dismal Swamp of the coastal plain of
Virginia and North Carolina is an immense forested area, only
333
a few feet above sea level, interspersed with stretches of open

water (Fig. 167). Here 1,500 square miles have been covered
with peat averaging 7 feet in thickness. Along the north-east
coast of Sumatra there are many scattered swamps supporting
almost impenetrable tropical jungles. In one of these peat is
known to have accumulated to a depth of 30 feet.
The densely forested swamps of the Ganges and other
tropical deltas (Plate 75b) provide ideal conditions for peat
growth and serve to illustrate the climatic and geographical
conditions under which the coal seams of the Carboniferous
period originated. Moreover, borings through the Ganges
delta reveal a succession of buried peat beds with intervening
deposits of sand and clay. The sequence points to repeated
alternations of subsidence and standstill, with the actual
surface never far from sea level.
As peat accumulates year after year the entangled water is
squeezed out of the lower layers and the peat shrinks and
consolidates. It still contains a high proportion of water,
however, and before being used as a fuel prolonged air-drying
is necessary. In appearance it then ranges from a light brown
fibrous or woody material to a dark brown or black amorphous
substance.
Coal and its Varieties
Peat becomes still further compacted when it is buried

beneath a cover of clays and sands. As the overhead pressure
increases water and gases continue to be driven off, their
composition being such that the residue is progressively en-
riched in carbon until it is transformed into a variety of coal
(Fig. 178). It has been estimated that at least a foot of peat
is necessary to make an inch of ordinary coal. The essential
conditions for the development of a coal seam are thus (a)
long-continued growth of peat and (b) subsidence of the
;
area and burial of the peat beneath a thick accumulation of

sediments.
Considerable variation in the character of coal is naturally
to be expected according to (a) the nature of the plant residues
334
)
Fig. 178
Diagram from woody material through
to illustrate the variation in composition
peat to coal and anthracite. Black columns represent carbon dotted columns,
;
oxygen ; and line-shaded columns, hydrogen. (After H. G. A. Hickling
— which determines the type of coal material and ( b ) the stage

;
of chemical alteration which has been reached -which deter- —

mines the rank of the coal.
In normal coals the remains of wood and bark predominate,
indicating derivation from forest peats. Coal of the lowest
rank, that is, the variety most like peat, is called lignite. It
commonly retains visible vegetable structures, but there are
also varieties, often known as brown coal, in which the woody
tissues are obscure. Lignites and brown coals are common
in the Cretaceousand Tertiary coalfields of Europe and North
America, but are of no importance in Britain.
The familiar shining black or dark grey coals of the British
and other Carboniferous coalfields belong to a group of fuels
of higher rank known collectively as bituminous coal. This term
does not imply the presence of the material properly called
bitumen, but has reference to the fact that in the manufacture
of coal gas and coke one of the distillation products, coal tar,
is of a bituminous nature. Coal of the highest rank called —
335
anthracite—occurs in certain localities (e.g. in South Wales)

where unusually high pressure due to deep burial and subse-
quent earth movements, combined with a moderate increase
of temperature, has facilitated the driving-off of nearly all
the volatile constituents. Anthracite is hard, brittle, and stony-
looking ; does not soil the fingers ; burns with a smokeless
flame, and has a high heat-producing capacity.
Coals of special types are formed from a variety of organic
ooze rich in spores or algal remains. These contain less carbon
than the normal coals of corresponding rank, but are notably
richer in hydrogen. When heated they give off abundant
supplies of hydrocarbons and on distillation they yield oil.
Varieties in which algal remains predominate are called
boghead or bogheads, after the name of an estate in West
coals,
Lothian where they were first mined. They are close grained
and tough and often resemble dark brown or nearly black
leather. They leave a considerable residue of ash, most of
which is the muddy material that contaminated the original
ooze. By increase of inorganic sediment the bogheads pass into
oil-shales.
Between bituminous coal and the bogheads there are
intermediate varieties known as camel coal. In some of these
spores and blebs of resin are very abundant, while others
contain algal remains in addition. Cannel coal is dull black
and appears quite structureless to the unaided eye. It occurs
as individual seams and also as lenticles and bands in seams
of ordinary coal. The name refers to the fact that splinters
of cannel can be burnt like a candle, a fact that in turn demon-
strates the richness of the material in inflammable hydro-
carbons.
The Constitution of Coal

Almost any block of bituminous coal can be seen to have
a well-marked banded or stratified structure. The commonest
bands are composed of soft bright coal which readily breaks
into approximately right-angled pieces with smooth brilliant
surfaces. Most of the bands appear to be quite structureless,
336
CONSTITUENTS OF COAL
and since the material bears a close resemblance to glass it

has been called vitrain or vitrinite. Others are finely laminated
and consist of shreds and films of vitrinite in a very fine-grained
matrix. This type of coal material is known as clarairt. The
bands of bright coal are separated by layers of a dull grey-
black type of coal which, being relatively hard and tough, is
distinguished as durain. None of the three types already men-
tioned soilsthe fingers. The dirtiness of coal is due to the
presence in quite small amounts of a fourth type of material
called fusain or fusinite. It consists of thin flakes of extremely
friable “ mineral charcoal ” which are distributed over certain
surfaces at intervals through the seam. Coal naturally splits
very easily along these surfaces, and as the fusinite readily
crumbles to powder the broken coal becomes dusty.
All the types of coal are intensely opaque, and it is necessary
to make sections ten times as thin as those of ordinary rocks
before light can penetrate the material so that microscopic
examination becomes possible. A translucent film of vitrinite
1 / 10, 000th of an inch in thickness is no longer black, but has a
rich golden brown or reddish colour. It is found to consist of
bark or wood, each band representing a single fragment.
Section a of Plate 76 shows a highly magnified sample of
typical vitrinite in which the cellular structure of bark is
perfectly preserved. The material clearly corresponds to the
humified bark found in peat.
Fusinite, similarly examined, also turns out to be wood or
bark, but only the carbonized cell walls remain, the cells being
empty or occupied by ash. It represents woody and other
fibres that escaped humification, though how they became
altered into charcoal is a problem that remains unexplained.
Some of the fusinite may be the charred relics of forest fires,
but most of it appears to have been formed after burial during
the process of “ coalification.”
Unlike these two types, which are unit constituents of
uniform composition, durain is found to be an assemblage of
minute particles, like the organic ooze of peat. It contains
the resistant coats of spores, more or less crushed and flattened,
microscopic shreds of vitrinite, lenticles and grains of fusinite,
337
and blebs of resin, in an obscure matrix of debris too finely
macerated to be identified. Section b of Plate 76 is cut from
a band of dull spore-coal, and shows durain consisting of
microspores (male) and macrospores (female) embedded in a
dark matrix. In Section c, cut from a banded coal, the upper
portion is part of a band of durain showing the same features.
The lower middle portion is a band of vitrinite formed mainly
from compressed bark. Part of another band of durain, with
microspores, is seen below.
Clarain contains the same ingredients as durain, but in
very different proportions. The well laminated structure and
brightness are due to the presence of abundant closely packed
strips of vitrinite. The intervening laminae of durain-like
material are extremely thin. A
very highly magnified thin
section of clarain would thus have an appearance not unlike
the more coarsely bandedcoal illustrated in Plate 76c. Clarain,
indeed, a sort of microscopic replica of a seam of bright coal
is
in which bands of vitrinite predominate.

Cannel coal is essentially durain which is especially rich
in spore cases and other waxy and resinous remains. There
are transitional varieties towards clarain on the one hand
and towards the algal-rich bogheads on the other.
Lignites and anthracites are found to consist of the same
structural types of material as bituminous coals. The variation
in properties throughout the series depends partly on the
proportions in which the type ingredients are present, and
partly on the degree of alteration which they have suffered,
i.e. on the rank of the coal. Fusinite, for example, is highly
inflammable, because its friability and high porosity make for
easy oxidation. —
Coal-dust explosions formerly a serious
menace to mining before precautions were enforced result —
from this dangerous property. Spore cases and resins are rich
in hydrocarbons consequently dull coals and cannels yield
;
far more gas and tar than the bright coals of intermediate ranks.
The latter, however, are excellent for household purposes.
Steam coal, suitable for use in locomotives and ships, is of
higher rank, transitional towards anthracite. It bums with
litdesmoke, but ignites more easily than anthracite and has a
338
) —
CARBONIFEROUS FORESTS
high heat-producing capacity. The properties of anthracite
slow ignition and slow burning, with intense heat and no
—
smoke are determined entirely by its high rank. Even such
spores as may still be detectable have been reduced to ghostly
carbonized relics.
Coal Seams and Coalfields

The development of a coal seam
essential conditions for the
are (a) long-continued growth of peat and ( b subsidence
;
of the area and burial of the peat beneath a thick accumulation

of sediments. The peak period of coal formation was the
Upper Carboniferous. Hercynian movements then provided
the extensive, rhythmically subsiding basins in which the sedi-
ments and coal seams of the Coal Measures were deposited ;
and provided them, moreover, along a belt running through

North America (bordering the Appalachians), the British Isles,
and parts of Europe and Asia, where the climate was then hot
and the vegetation luxuriant.
There are no coalfields earlier than the Carboniferous.
Land plants capable of preservation make their first feeble
appearance only towards the end of the Silurian. By Carboni-
ferous times, however, a rich and prolific flora had developed,
and the fossilized remains of more than three thousand species
are already known. The chief coal-makers were tall forest
trees ( Lepidodendron and Sigillaria, with widely spreading roots
known as Stigmaria) which grew to heights of as much as a
hundred feet ;
and giant reeds called Calamites (the ancestors
of the little horse-tails of to-day) which flourished in bamboo-
like thickets toa height of fifty feet or more together with an
;
undergrowth of smaller rushes and ferns, and slender plants

of trailing or climbing habits (Fig. 179). No flowers or birds
enlivened these gloomy jungles, but insects again of extra- —
—
vagant size were abundant.
Practically all seams of bituminous coal and anthracite
have certain characteristics which are consistent with the
theory that each seam represents the actual site of the swamp
in which the parental vegetation lived and died.
339
*
jp* ** >.
m-
ru-wf* /
Fig. 179
Reconstruction of a Carboniferous forest
The“ seat earth ” which forms the floor

(a) of a seam is a
high-grade fireclay (useful for making refractory bricks). It is
riddled with innumerable rootlets of the plants which first
colonized the swamp, and may even contain casts of stigmarian
COAL SEAMS AND THEIR ORIGINS
roots, often in the position of growth. The trunks either
rotted away above water level or fell into theswamp to con-
tribute to the lower part of the seam.
b The roof of the seam sometimes contains casts of the
( )
trunks of great forest trees. These represent the generation

that was drowned when the swamp conditions were brought
to an end, for the time being, by subsidence and general
inundation.
(c) Seams are generally of very wide extent, sometimes
covering areas of thousands of square miles with but little

variation in thickness.
(d) Seams are locally interrupted by “ wash-outs,” that is,
by the sandstone-filled channels of streams that flowed through

the forest swamps, like the distributaries of modern deltas.
(e) The coal contains no fish remains or other fossils of
aquatic animals, and (except in certain bands of durain) is

uncontaminated by muddy sediment. Such ash as remains
when the coal is burnt is derived either from the vegetation
itself or from carbonate minerals and pyrite (the brassy-
looking material sometimes seen in coal) subsequently deposited
in cracks from ground-waters. Seams and bands of the durain
type may, however, leave a little sedimentary ash. These dull
coals seem to have accumulated in stretches of stagnant water
into which a limited amount of fine sediment might be intro-
duced while the delta rivers were in flood.
—
In the special coals the cannels and bogheads muddy —
sediment is much more abundant than in durain. Moreover,
these varieties contain the remains of fish and other aquatic
organisms. The water in which the mud-contaminated ooze
accumulated was therefore not stagnant, but was continually
renewed and oxygenated. Evidently these coals were formed
not in situ, but from plant debris that was drifted by wind and
running water into lakes with a through drainage. Such con-
ditions would also arise locally in the hollows of a peaty
surface just beginning to subside. Matching expectation, thin
lenticles of cannel of limited area are found to occur at the top
of many seams of otherwise normal coal.
(396)
34 x 23
Fig. 180
Map of the coalfields (black) of the British Isles. Areas of younger rocks are
indicated by shading ; those of older strata are left unshaded. (The leading
tectonic structural lines of Scotland and their continuations into Ireland are
added for convenience ; see pp. 367 and 436)
In the Upper Carboniferous coalfields the successive seams

are separated by a characteristic sequence of sediments which
is commonly repeated dozens of times, and in some localities
hundreds of times. A coal seam is usually roofed with shales

in which fossil leaves and bands of freshwater mussels fre-
342
BRITISH COALFIELDS
quently occur. As they are followed up the shales become

sandy and pass into sandstones, sometimes with shaly
interruptions. Then follows the seat earth that underlies
the next coal. The picture is clearly one of repeated alter-
nations of phases of subsidence and phases of comparative
rest.
The subsiding regions developed into widespread tectonic
basins lying between rising tracts of country which supplied
the basins with sediments and kept them filled. The floor of
part of the South Wales coalfield sank more than 10,000 feet
in all, while to the north a persistent ridge of higher ground
separated the southern basin from the vast area of irregular
depression in which the northern coalfields originated (Fig.
180). The sediments of this group of coalesced basins came
mainly from an upland region now represented in part by the
Highlands of Scotland.
Evidently at this time both the British area and the
adjoining parts of Europe (Fig. 200) were being warped into
basins and swells by pulsations of pressure associated with the
Hercynian mountain-building movements. The latter were
already in active progress farther south. By later movements
the basins and their contents were themselves buckled, folded,
and faulted. The dominantly upfolded portions, being ex-
posed to denudation, have since lost their original covering of
Coal Measures. The downfolded portions, however, have been
preserved as the isolated coalfields of to-day either where the
;
Coal Measures are exposed at the surface, or, like the coalfield
of Kent, where they are concealed by a blanket of later
deposits.
Petroleum
Petroleum (Gr .petra, rock, L. oleum, oil) is the general term

for all the natural —
hydrocarbons whether gaseous, liquid, or
solid —
found in rocks. In common usage, however, it refers
more particularly to the liquid oils. Gaseous varieties are dis-
tinguished as natural gas. Highly viscous to solid varieties are
called bitumen or asphalt, but the latter term is also applied to the
343
bituminous residues when petroleum is refined, and to

left
natural and paving materials composed of sand,

artificial
gravel, etc., with a bituminous cement.

Petroleum consists of an extremely complex mixture of
hundreds of different hydrocarbons, generally accompanied by
small quantities of related compounds containing nitrogen,
sulphur, or oxygen. The hydrocarbons fall into a number of
natural series of which the paraffin series is the most familiar.
Its members, all of which can be expressed by the formula
C„H 2(l+2 range from light gases ( e.g methane, CH 4 , the chief
,
constituent of natural gas), through a long series of liquids (the

chief ingredients of successive products of distillation such as
and lubricating oil), to paraffin wax (in-
petrol, paraffin oil,
cluding C 20 H 42 and
higher members). Crude oils in which
these hydrocarbons predominate are said to have a paraffin
base they are generally of pale colour with a yellowish or
;
greenish hue. The darker brown and greenish oils generally

contain a high proportion of the naphthene series, each
member having a composition of the type C n 2 „. H
These
furnish heavy fuel and, as they leave a dark asphaltic
oils
residue on being refined, they are said to have an asphaltic

base. Intermediate varieties have a mixed base of wax and
asphalt. In all crude oils there are also smaller propor-
tions of several other series, including acetylene and its
higher members, C„H 2n _ 2 , and a great variety of aromatic
hydrocarbons, of which the benzene series, C„H 2„_ 6 is an ,
example.
To avoid confusion it should be clearly understood that
neither oil shales nor the cannel and boghead coals contain
petroleum as such. If they did, it could be dissolved out by
carbon disulphide. They do, however, contain pjro-bituminous
substances which can be altered into oil and bitumen by heat.
Such deposits can therefore be made to yield a group of
petroleum products by destructive distillation. Petrol and
related products can be obtained in commercial quantities
from ordinary coal only by highly technical processes involving
the intimate introduction of hydrogen into suitably prepared
coal at high pressures and temperatures. Petrol can also be
344
:
SOURCES OF MINERAL OILS
made from the heavier and less valuable oils by a similar but
less elaborate process of hydrogenation. The following table
summarizes the sources of oil and related products :
Bituminous deposits Pyro-bituminous deposits Carbonaceous deposits

of requiring requiring
Petroleum destructive distillation hydrogenation
Natural gas Special Coals Ordinary coals

Crude or mineral oil Cannels
Bitumen and mineral Bogheads
wax
Tar sands and asphalt Oil-shales —
Being fluids, oil and gas behave very much like ground-
waters. They occupy the interstices of pervious rocks, such
as sand and sandstone and cavernous or fissured limestones,
in places where these “ reservoir rocks ” are suitably enclosed
by impervious rocks, so that the oil and gas remain sealed up.
Accumulations on a scale sufficient to repay the drilling of wells
are referred to as oil or gas pools. The “ pool,” however, is
merely the part of a sedimentary formation that contains oil
or gas instead of ground-water.
The Origin of Petroleum

Unlike coal, petroleum retains within itself no visible
evidence of the nature of the material from which it was
formed. It has been suggested as a purely speculative possi-
bility that oil may have been formed by volcanic or deep-
seated chemical processes akin to the production of acetylene
by the action of water on calcium carbide. But these hypo-
theses are quite incompatible with the geological distribution
of oil and with certain peculiarities of its composition and pro-
perties. All the relevant evidence points convincingly to an
organic origin.
(a) Some of the constituents of petroleum have the pro-

perty of altering the direction of vibration of light rays. This
“ optical activity ” is characteristic of certain substances pro-
345
duced by plants and animals, but is not shared by the hydro-

carbons and related compounds generated by purely chemical
reactions.
(b) The nitrogen-bearing constituents of petroleum include
a group of compounds called porphyrins which can be formed

only from the green colouring matter (chlorophyll) of plants
or from corresponding colouring substances of animal origin.
Oil-shales rich in algal remains also contain vegetable por-
phyrins. In the presence of oxygen the porphyrins are quickly
destroyed, and their persistence in oil indicates that the latter
must have originated in an environment from which free
oxygen was absent. Although the porphyrins derived from
vegetation are more easily oxidized than those of animal
origin, they are, nevertheless, the dominant type. It is therefore
probable that plant life has contributed to the raw material of
petroleum more than animal life.
c
( )
Oil is not found in association with volcanoes or
igneous rocks, except accidentally. West of Edinburgh, for
example, oil shales have been invaded by intrusions and
volcanic necks, with results comparable to those obtained
when oil shales are distilled. Oil so liberated by metamorphism
would naturally migrate into overlying sandstones, and there
it is still occasionally found. No major oilfield, however, has
originated in this way. About 70 per cent, of the world’s
known oilfields have been located in marine sediments of
Cretaceous and Tertiary age, generally along the flanks and
in the less closely folded portions of the Tertiary orogenic belts.
Most of the remaining fields occur in the Palaeozoic sediments
of North America and the U.S.S.R. in the geosynclines and
marginal basins of earlier ages (cf. Fig. 182). No oil pools of
commercial importance have been found in sediments older
than the Ordovician.
d The existence of pre-Carboniferous oilfields suggests
( )
that land plants were not essential to oil formation, and this
inference is strengthened by the important fact that no signifi-
cant lateral connection between coal seams and oil pools has
anywhere been traced. The two may occur in close association
346
ORGANIC ORIGIN OF PETROLEUM
by some accident of faulting, and one may lie above the other
in a sequence of varied strata, but in neither case has the
association any bearing on the origin of oil. While it is not
impossible that drifting relics of land vegetation, swept into
the sea by great rivers, may have contributed to oil formation,
it is such a source was quite subsidiary to the con-
likely that
tributions furnishedby marine algae and diatoms. Moreover,
wherever the conditions were such that organic residues from
marine plants could survive, they must equally have favoured
the accumulation of similarly unconsumed remains of foramini-
fera and other forms of animal life.
The various lines of evidence all lead to the conclusion that

petroleum has originated from organic matter which became
incorporated in the sediments of depressed regions of the sea
floor. The organisms flourished in the surface waters, and
their dead remains sank into stagnant depths with a marked
deficiency of oxygen, where bacteria alone could live, and
where they were safeguarded from being either oxidized or
eaten. Indeed, under these conditions oxygen would be
actually abstracted by bacteria, and the organic matter trans-
formed into fatty and waxy substances. For an example of
the early stages of oil formation, in progress at the present day,
reference may be made to the muds now collecting in the
deeper parts of the Black Sea. Samples dredged up contain
as much as 25 per cent, of organic matter, and 10 per cent, of
this has already been changed into heavy hydrocarbons,
soluble in benzene. The solution, moreover, is green and
contains traces of chlorophyll. The oil first generated in
mud is mainly bitumen composed of large molecules. The
lighter hydrocarbons of the paraffin series appear to be later
derivatives, produced by a kind of natural refining brought
about during burial by increased pressure, rising tempera-
ture and continued bacterial action. Bacteria are known
to exist in the ground-waters of certain oilfields. The details
of those later developments are, of course, very difficult to
trace.
347
Migration and Concentration of Petroleum

The sediments which petroleum had its origin are called
in
the source rocks, to distinguish them from the reservoir rocks in
which oil and gas are now found on a commercial scale. The
reservoir rocks carry far more oil than could possibly have
originated within them and, moreover, they commonly contain
fossils of the benthos group which lived on the bottom in shallow
oxygenated waters where no appreciable quantities of organic

matter could have survived. The source rocks, on the other
hand, must have been muds and calcareous deposits with
when preserved, corresponding to the organisms (mainly
fossils,
of the plankton group) which contributed the raw materials

of petroleum. An oil pool is, in fact, a concentration of oil
which has migrated from the source rocks into places where it
could draw to a head and accumulate.
Source beds such as clay and shale are now compact and
impervious. But while they were still unconsolidated they
contained a high proportion of sea-water carrying dispersed
globules of oil. During this stage circulation of the mixed
fluids would sooner or later become possible, in response to
pressure differences set up by a varying overhead load or,
more effectively, by earth movements. As the source beds
become compressed, the squeezed-out fluids pass into more
coarsely porous and less compressible formations, such as
sands. Once the oil has been flushed into these permeable
beds it may be carried through them for long distances, but
it cannot again escape from them, unless the overlying rocks
are fissured. If the mixed fluids encounter a sediment with

very fine pores, the water may filter slowly through, but the
oil is held back. Within the sand or other reservoir bed the
oil trickles upwards through the water until it comes up against
an impervious barrier and collects there.

In general, then, oil migrates outwards and upwards from
the source beds, passes into coarsely porous or fissured reservoir
beds, rises to the highest possible level, and collects into an oil
pool wherever the structure provides a trap which impedes
further migration. Gas, if present in excess of the amount
348
OIL TRAPS
that the can hold in solution, bubbles to the top and forms
oil
a gas cap over the pool. Beneath the pool the pore spaces are
occupied by ground-water (often salt) which is commonly under
a very considerable hydrostatic pressure. If the pressure and
gas content are sufficiently high, the oil gushes out like an effer-
vescent fountain when the pool is tapped by drilling. But
when the pressure conditions are insufficient to drive the oil
to the surface —
or become so as the initial pressure falls off-
pumping necessary to bring it up.
is
In accordance with the principles of oil concentration a

dome or anticline of alternating pervious and impervious
sediments makes an efficient trap for oil migrating towards it
(Fig. 181a). open anticlines surrounded by extensive

Isolated
gathering grounds have a much better chance of being pro-
ductive than more closely packed folds. Not only have the
latter to share a limited supply, but they are likely to be too
much broken and fissured to retain any oil and gas that passed
into them.
“
Although the anticline theory ” of oil concentration
dominated the search for oil for many years, it gradually
came to be realized that anticlines are far from being the only
traps, or even the most productive. The early discovered oil-
fields of Pennsylvania, for example, occur in a broad sedi-
mentary basin in which the formations still remain practically
horizontal over widespread areas. Here the oil pools occupy
lenticular bands of porous sandstone which pass laterally, as
well as vertically, into shales thus the oil is sealed within an
;
impervious envelope (Fig. 181c). Oil pools occupying the

upper ends of tilted reservoir beds are also of great importance.
The tilted bed may pass laterally into shale, or it may be more
abruptly cut off by an impervious barrier. The obstacle
may be a fault throwing an impervious bed against the reser-
voir bed (Fig. 1816) or a hill belonging to an ancient land
;
surface which was unconformably overlapped and buried by

a later series of petroliferous strata (Fig. 181c?) ;
or a salt dome
which has been intruded into a thick series of sediments
(Fig. 181c).
Salt domes are curious structures occurring in great
349
Fig. 181
Sections to illustrate various types of structural traps favourable to the accumula-
tion of oil and gas (gas is omitted except in a)
numbers along the Gulf Coast of the United States and in

other regions where salt deposits have been deeply buried.
Being plastic under high pressure, the salt is squeezed towards
places of weakness in the sedimentary cover. It then develops
into a plugwhich ascends through the cover, perforating and
doming up the beds as it advances. Some salt domes rise from
depths of four or five miles. In Iran some of them bulge up
the surface, and those that are still actively rising pierce the
roof and escape down the slopes as “ glaciers ” of salt. The
simple case illustrated in Fig. 181 e shows oil accumulating
against the walls of a salt dome. Oil may also be dammed
back by faults produced in the surrounding rocks by the upward
drag of the intrusive salt, and finally, it may collect in the
domed sediments over the roof.
Oil not necessarily, or even generally, confined to a single
is
reservoir bed in a given field. Any suitably placed formation

may have been fed either from an outlying primary source, or
350
ESCAPE OF NATURAL GAS
via an underlying pool, the oil from which escaped upwards

through fractures in the intervening impermeable beds. Gas,
in particular, readily migrates to higher levels, and in many
places vast quantities occur alone. This reflects the natural
tendency of some varieties of petroleum to differentiate into
asphaltic and gaseous fractions. If the reservoir begins to leak
after such fractionation has taken place, the mobile gas moves
on and leaves the sticky asphalt behind. Most of the gas
encountered by drilling was formerly discharged into the air
and wasted. But now it is welcomed, not only as an easily
distributed source of power and illumination, but still more
because its heavier constituents can be condensed into petrol.
The Discovery of Oilfields
Wherever the cover of a reservoir is perforated by fissures

leading up to the surface, leakage of gas and oil becomes
inevitable. Moreover, denudation may strip off the covering
rocks until gas can force a passage through the roof and so open
a way for the subsequent escape of oil. Thus the most obvious
indications that a region is petroliferous are (a) seepages and
springs of gas or oil ;

and (b) superficial deposits or veins of
asphalt and other more or less “ solid ” residues of petroleum
leftbehind after the volatile constituents have evaporated.
Inflammable gas may be found bubbling through the water
of springs and wells or streams. In a few places gas escapes
on a more spectacular scale. Historical records show that
vigorous gas jets may persist for many centuries. Blazing jets,
like the “ perpetual fires ” along the Caspian coast near Baku,
were long regarded with veneration by Fire Worshippers.
Where gas erupts through water-bearing strata and clays it
carries up wet mud and spatters it around the vent until a
mound is built up with a crater at the summit. Groups and
rows of these “ mud volcanoes,” occurring on the crests of
anticlines or along faults, are well known in Burma, Trinidad,
California, and various other localities. Eruption may be
continuous or spasmodic, according to the gas pressure and
35i
the nature of the obstruction to be overcome. There is, of
course, no connection with ordinary volcanic activity.

Flowing springs of oil emerge at the surface in some
localities, but more commonly the exudations are sluggish,
and the oil may be seen only as iridescent films on water. The
“
largest surface “ shows ” are sands cemented with residual tar”
(bitumen), or localized deposits of more concentrated asphalt.
These residues may eventually plug up the outlets and so pre-
vent further losses from depth ; but, on the other hand, they
may be all that remains of an otherwise dissipated oil pool.
Evidence of the use of bitumen in Mesopotamia dates back
to very ancient times. Long before Noah caulked his Ark with
asphalt, the value of bitumen for cementing and waterproofing
purposes had become familiar to the Sumerians (4000 B.C.).
The whole region of the Middle East, from the Caucasus to
Iran, is now known to be richly endowed with concentrations
of oil. The famous asphalt lake of Trinidad, first described
by Lery in 1786, is still being replenished from underlying oil
sands, but much too slowly to keep pace with the immense
quantities removed each year. Another well-known occurrence
is near Los Angeles, where crater-like depressions blown out
by gas have been filled with deposits of asphalt. Here the

skeletons of a great variety of prehistoric animals are found in
a perfect state of preservation, their unfortunate owners having
broken through the surface crust and become helplessly mired.
The first oilfields to be discovered were naturally found by
digging wells in the neighbourhood of surface “ shows.”
Many centuries before the Yenangyaung field of Burma
(Plate 77a) was developed by modern methods the Burmese
collected oil from surface exudations and later from shallow
hand-dug wells. Yenangyaung is the classic example of the
concentration of oil in the sands of an elongated dome. Oldham
was the first geologist to survey the area, and as early as 1855
he pointed out the importance of anticlinal structures as oil
traps. Nevertheless little call was made on the services of
geologists until after 1900, when the demand
began to
for oil
accelerate, first motor transport, then for aeroplanes, and
for
eventually for war purposes and national security. It was then
352
THE SEARCH FOR OIL
realized that the systematic discovery of oil involves the search
for potential oil traps, that is to say, it demands the detailed
geological survey of all the regions in which oil might con-
ceivably lie hidden. The search for oil became a geological
enterprise.
It is obvious, however, that in addition to the favourable
structures that can be located from outcrops at the surface,
there must be some that cannot be detected in this way, and
others that lie concealed by tropical forests or marshes, by
spreads of alluvium, boulder clay, desert sands or loess, or even
by the sea. Hidden anticlines and structures such as the
buried hills and salt domes illustrated in Fig. 181 can be
detected by the gravity anomalies (p. 404) to which they give
rise. all kinds can be explored by
Significant structures of
their effects on earthquake waves (p. 371). Such
artificial
waves are generated to order by exploding a charge of dynamite
in the ground, and the waves reflected or refracted back to
the surface by the rocks encountered in depth are then recorded
by seismographs placed at suitable distances from the point of
explosion. Many hundreds of salt domes, for example, have
been successfully located by this method. Magnetic and
electrical methods of exploring underground structures have
also been devised.
The discovery of a favourable structure does not, of course,
guarantee that oil will be found. On the other hand, the
absence of surface indications is no proof that oil will not be
found ;
it may also mean that the oil, if there, is sealed in so
efficiently that it cannot escape. Whether oil is present in

commercial quantities or not can be finally determined only
by the practical test of sinking wells to strike the parts of the
suspected reservoirs where oil is most likely to be concentrated.
In the early days of the oil industry a few important oil-
fields were found more or less by accident. Wells drilled for
water, for example, sometimes strike oil. Moreover, there have
always been optimistic operators willing to risk their capital
and take a chance by sinking “ wildcat ” wells on sites selected
for some quite unscientific reason. Only 3 or 4 per cent, of
these speculative ventures have proved successful, but the
353
fabulous profits that reward success continue to encourage
wildcat prospecting. It is a remarkable fact that the first
indication of the existence of a great oilfield in East Texas
was discovered in 1930 merely by drawing a line between two
of the already developed fields of Texas and Louisiana and
drilling along it. In twelve years this prolific field produced
over 300 million tons of oil from about 30,000 wells in an area
of 7 miles by 40, and it will probably yield at least as much
again before it is exhausted. This is a world record so far.
The following production figures for the last full year before
the War serve to illustrate the general distribution of oil by
countries :
Petroleum Production (in tons) in 1938

U.S.A 167,705,000 Peru .... . 2,186,000
U.S.S.R. . . . 29,630,000 Burma and India. . 1,435,000
Venezuela 27,657,000 Bahrein 1,117,000
....
.
.
Iran 10,192,000 Canada . 883,000

Dutch East Indies. 8,194,000 Germany . . 599,000
Rumania . 6,761,000 Poland.... 541,000
Japan ....
.
Mexico 5,434,000 344,000

Iraq .... 4,298,000 Equador .
.
291,000
Egypt ....
.
Colombia . 3,068,000 . 222,000

Trinidad 2,541,000 All other countries . 321,000
Argentina . 2,386,000
Total. 275,805,000
The total world production up to the end of 1938 is estimated

to have been about seventeen times the 1938 total, and the
proved reserves at that date amounted to about sixteen years’
supply at the 1938 rate of extraction. However, as new fields
are continually being discovered, not likely that any
it is
serious shortage of oil will be many
years to come.
felt for
A particularly significant fact is that 54 per cent, of all the
oil already won has been contributed by the United States,
although its territories include only 15 per cent, of the world’s

total area of unmetamorphosed marine sediments of Palaeozoic
and later ages in which oil might be expected to occur. Let
us call 54/15 or 3*6 the relative productivity of the United
354
States. the corresponding productivity for all the rest

Then
of the world is 46/85 or 0-54. This means that seven times
more oil has been obtained from the marine sediments of the
United States than from the similar sediments elsewhere. Yet
from a geological point of view there is no obvious reason why
the United States area should have been so specially favoured.
It may be conceded as probable that the United States has
more than an average share of the world’s oil, but it must not
be overlooked that its citizens have sought for oil far more
actively than those of any other country. Geologists have
been employed in making systematic surveys on a scale rivalled
only by Britain and the U.S.S.R. (who, however, began late)
and the efforts of wildcatters have provided much additional
information, as well as a surprisingamount of oil, despite the
wastage of effort involved in the attempt to short-circuit the
preliminary search for structures. Altogether more than a
million wells have already been drilled and the number grows
by about 30,000 a year. Moreover, as the shallow pools
became exhausted, wells have been sunk to ever-increasing
depths. One exploratory well in California has been carried
down just over 15,000 feet. In the light of this record of
enterprise it is impossible to resist the conclusion that the high
rate of oil production and discovery in the United States

the intensity of the search as well as the actual resources.
reflects
It may not unreasonably be anticipated that other lands
favoured with suitable sediments and structures may yet be
rewarded with successful discoveries when the problem of
finding oil is tackled with corresponding energy and efficiency.

W. Gibson
Coal in Great Britain. Arnold, London, 1927.
A. Raistrick and C. E. Marshall
The Nature and Origin of Coal and Coal Seams. English Universities
Press, London, 1939.
E. S. Moore
Coal. Wiley and Sons, New York, 1940.
356
lC. E, Mat shall
(A) Bark preserved as \itiinite. x 3.“>U
(B) Dull coal ^durain) with spores, x 30
.
[C. Mat shall
(C) formal banded coal, x 30
PHOTOMICROGRAPHS OF COAL
V .-i.
(A) Part of the Yenangyaung Oilfield, Burma (in 1922)
\Burmah Oil Co., Ltd.

(B) Part of the Digboi Oilfield, Assam
COAL AND OIL
G. Stutzer and A. C. Noe
Geology of Coal. Chicago University Press, 1940.
J. G. Crowther
About Petroleum. Oxford University Press, 1938.
W. H. Emmons
Geology of Petroleum. McGraw-Hill, New York, 1931.
W. E. Pratt
Oil in the Earth. University of Kansas Press, 1942.
V. C. Illing
Geology applied to Petroleum. Proceedings of the Geologists’ Association,
Vol. LIIL, pp. 156-87, 1942.
PART III— INTERNAL PROCESSES AND
THEIR EFFECTS
Chapter XVII
EARTHQUAKES
The Nature of Earthquakes
When a stone is thrown into a pool, a series of waves spreads
through the water in all directions. Similarly, when rocks
are suddenly disturbed, vibrations spread out in all directions
from the source of the disturbance. An earthquake is the
passage of these vibrations. In the neighbourhood of the
disturbance itself the shaking of the ground can be felt and
the effects may be catastrophic, but further away the tremors
die down until they can be detected only by delicate instru-
ments called seismographs (Gr. seismos, an earthquake).
Vibrations are set up in solid bodies by a sudden blow or
rupture, or by the scraping together of two rough surfaces.
Corresponding causes of earthquakes in the earth’s crust are
volcanic explosions, the initiation of faults, and the movements
of the rocks along fault planes. Perceptible tremors are set
up by the passage of trains and tanks, by avalanches and
landslides, by rock in mines and caverns, and by explosions
falls
of all kinds. When
a munition factory explodes, the intensity
of the resulting earthquake may be comparable with that of
volcanic earthquakes. The majority of earthquakes, however,
including all the most widespread and disastrous examples,
are due to sudden earth movements, generally along faults ;
these are distinguished as tectonic earthquakes. The term
tectonic (Gr. tekton, a builder) refers to any structural change
brought about by deformation or displacement of rocks
(cf. architecture).
The cause of tectonic earthquakes is thus the application
of stresses to rocks until they are strained to breaking point,
358
ASSOCIATED CRUSTAL MOVEMENTS
when they suddenly rupture and move. The fault movements

themselves, as already described on page 78, may be either
vertical or horizontal or oblique. After the great Alaskan
earthquake of 1899, it was possible from the presence of
barnacles clinging to the uplifted rocks of Disenchantment
Bay to measure the uplift, which in this case reached an
exceptional maximum of 47 feet. In Japan the crustal blocks
Fig. 183
Map of Sagarai Bay, Japan, showing movements associated with the earth-
quake of September 1, 1923. Vertical displacements ranging from a few inches
to several feet are indicated by shading on the downthrow side of the fault
lines. Horizontal rotation in a clockwise direction is indicated by dotted
lines, with numbers representing the displacement at various localities in feet
oftenmove obliquely, both vertical and sideways movements

being observed. Surveys carried out after the Sagami Bay
earthquake of 1923, when Tokyo and Yokohama were wrecked,
showed that the floor of the Bay and the surrounding mainland
had twisted round a little in a clockwise direction, the observed
shift of Oshima Island being over 12 feet (Fig. 183). Both in
Japan and the Philippines the horizontal fault displacements
indicate a general southerly movement of the Pacific side.
On the other hand, the horizontal movements along the San
359
EARTHQUAKES
Andreas fault in California point to a northerly movement on
the Pacific side (Fig. 184). Along one San stretch of the
Andreas fault, deep ravines in the on the continental
hillsides
side of the fault are abruptly displaced, reappearing on the
seaward side 150 feet to the north-west. This probably repre-
sents the cumulative effect of several comparatively recent
Fig. 184
Map of part of California and Nevada showing the San Andreas fault and the
other chief faults of the area. Many of these have been active in recent time
faultmovements, each like the one responsible for the 1906

earthquake when the maximum lateral shift was 21 feet.
When the rocks are nearly at their breaking point, an
earthquake may be precipitated by some minor, but quite
extraneous agent, such as a high tide, a rapid change of
barometric pressure, or shaking by an independent earthquake
originating elsewhere. Thus arises the occasional association
360
EARTHQUAKE DAMAGE
of earthquakes with great storms or with other earthquakes.
The principal shock, which generally lasts only a few seconds
or at most a minute or two, may be preceded by fore-shocks,
and invariably followed by a series of after-shocks. The fore-
is
shocks represent the preliminary loosening and removal of

small obstructions along a fault plane or zone. When these
have been overcome, the main movement occurs. But complete
stability is not restored immediately, and moreover, the jolt
often disturbs adjoining fault-blocks. The after-shocks repre-
sent the minor movements as the region gradually settles down
again. Considering the whole earth, earthquakes of one kind
or another take place every few minutes, but many of these
are so slight that they are known to have occurred only from
instrumental records. Really severe earthquakes, which would
be catastrophic in populated areas, take place about once a
fortnighton an average. Fortunately most of these originate
beneath the continental slopes and do little damage.
The Effects of Earthquakes
One of the most alarming and disastrous features of a

great earthquake near its place of origin is the passage of large
“ surface ” waves over the ground, which is thrown into ever-
changing undulations. The latter may beonly a foot high
and 30 feet or so from crest to crest, but the rapidity of their
rise and fall often gives the terrifying impression that the
ground is writhing “ like a storm-tossed sea.” Fissures gape
open at the crests, only to close again as the waves pass and the
crests turn into troughs. Roads are cracked open (Plate 78a),
railways are buckled and twisted (Plate 78b), bridges collapse
(Plate 79), and buildings are shaken down. On the sea floor
telegraph cables may be broken. The surface waves are a
local by-product of the normal earthquake vibrations, which
are of a much smaller order. In the Assam earthquake of
1897 the ground vibrated more than two hundred times a
minute, each to and fro movement being more than a foot
in range. Such rapid shaking through even an inch or two
361
EARTHQUAKES
would be highly destructive. A range of a quarter of an inch
would suffice to wreck most buildings not specially constructed
to resist earthquake shocks, and one of T6>0 ^ inch can be felt.
The effect on the feet of the vertical vibrations in a strong
earthquake has been described as like the “ powerful upward
blows of a monstrous hammer.”
In the region of destruction landslides are set moving on
valley sides, and avalanches are started in snowy mountains.
Glaciers are shattered and where they terminate and break
off in the sea icebergs become unusually abundant. Vast
masses of wet sediments may be so loosened by submarine
shocks that they slump for miles down the continental slope.
In Sagami Bay in 1923 parts of the floor were thus lowered by
1,000 to 1,500 feet, other parts being correspondingly raised.
Underground water is greatly disturbed by earthquakes, and
new lakes or swamps may be formed and old ones drained.
Compression of water-filled sands, especially in alluvial
the water to ascend through fissures and often
districts, forces
to issue at the surface in powerful sandy jets around which
sand-craters develop.
Strong submarine earthquakes are followed by seismic sea
waves, technically called tsunamis (Japanese). The celebrated
Lisbon earthquake of 1755, probably the greatest on record,
originated in a sudden subsidence of the sea floor to the west.
At Lisbon the sea withdrew immediately after the principal
shock, only to return as a gigantic wave, about 40 feet high,
which swept across the lower parts of the city and completed
the ruin and desolation. The ebb and flow of the sea continued
for some time after the first wave, this being an inevitable
characteristic of all tsunamis. The Lisbon earthquake was of
such exceptional severity that lakes were set oscillating as far
away as Loch Lomond and Loch Ness, where the water
continued to rise and fall through a range of two or three feet
for about an hour.

The appalling losses of human life that accompany great
earthquakes in populated areas are mainly due to secondary
causes such as the collapse of buildings, fires, landslides, and
tsunamis. Gas mains are torn open and fires, once started,
362
EARTHQUAKE INTENSITIES
rapidly spread beyond control, since the water mains are also
wrenched apart. In San Francisco in 1906 far more damage
was done by fire than by the earthquake itself. The Sagami
Bay earthquake of 1923 occurred just as the housewives of
Tokyo and Yokohama were cooking the midday meal. Fires
broke out in all directions and completed the toll of death and
destruction. Two hundred and fifty thousand lives were lost
and over half a million houses destroyed. In the loess country
of Kansu in China, 200,000 people were killed in 1920, and
another 100,000 in 1927 by catastrophic landslips of loess which
overwhelmed cave dwellings, buried villages and towns, and
blocked river courses, so causing calamitous floods.
Isoseismal Lines and Depth of Origin
Within the area disturbed by an earthquake which may —

be anything up to millions of square miles the intensity at —
any place is gauged by the effects on buildings and on the
ground (fissures and landslips), on people, and on seismographs.
The intensity is stated by reference to an arbitrary scale of
twelve degrees (originally ten), of which the following is a
brief summary :
MODIFIED MERCALLI SCALE OF EARTHQUAKE INTENSITIES
I (< 10) Instrumental, detected VII (> 500) Very strong, cracking of
only by seismographs walls, general alarm
II (> 10) Very feeble, noticed only VIII (>1,000) Destructive, chimneys
by sensitive persons fall
III (> 25) Slight, felt by people at IX (>2,500) Ruinous, houses begin to
rest fall
IV ( > 50) Moderate, felt by people X (>5,000) Disastrous, many build-

in motion ings destroyed
V (> 100) Rather strong, people are XI (>7,500) Very disastrous, few
wakened, bells ring structures left stand-
ing, ground fissured
VI (> 250) Strong, slight damage XII (>9,800) Catastrophic, total de-
struction, o bj e c t s
thrown into air,
ground badly twisted
363
EARTHQUAKES
The actual intensity of the vibrations is measured by the maxi-
mum acceleration of the ground ;

approximate values for the
latter are given in each case in brackets (the acceleration of
gravity in the same 9,800 mm. /sec. /sec.). From the
units is
centre of the disturbance the intensity decreases outwards

inversely as the square of the distance.
A line drawn through all places with the same intensity
Block diagram showing isoseismal lines and their relation to the epicentre and
to the wave paths radiating from the focus of an earthquake
is an isoseismal line (Fig. 185). Each one generally encloses

a roughly circular or elliptical area, according as the place
of origin of the earthquake is a point-like or elongated area.
The place of origin is called the origin or focus, and the point
or line on the surface vertically above is the epicentre or epicentral
line. By comparing the intensities at the epicentre and
along an isoseismal line, Oldham showed how the depth of
focus could be determined (Fig. 186).
Intensity at E = m (known)
Intensity at G —n (known)
G is ata known distance d from E
G is at an unknown distance r from F
E is at an unknown distance h above F
Fig. 186
Diagram to illustrate Oldham’s method for estimating the depth of the focus
of an earthquake
364
DEPTHS OF ORIGIN
By the inverse square law,
njm = h 2 jr 2 = sin 2
9 ; the angle 9 being thus determined,
h = d tan Q = the depth of the focus.
From Oldham found that
the records of 5,605 shocks in Italy,
90 per cent, of the earthquakes originated at depths of less
than 8 km. (5 miles) nearly 8 per cent, at depths between
;
8 and 30 km. and the rest at greater depths. Tectonic earth-

;
quakes are now classified as
Normal, when the depth of origin is 0-50 km.

Intermediate, when the depth of origin is 50 - 250 km.
Deep-focus, when the depth of origin is 250 - 700 km.
Volcanic earthquakes, which may be due to gas explosions or

to the formation and injection of fractures by magma, are
generally of relatively shallow origin. For this reason the area
of disturbance is correspondingly small (a few hundred square
miles at the most), though the intensity may be high near the
volcano.
Distribution of Epicentres : Earthquake Belts

Earthquakes may be recorded anywhere, but the places
where they originate are confined to regions where earth
movements or volcanoes are active. Fig. 187 shows the
distribution of known epicentres during the present century.
The distribution turns out to be essentially the same when all
known epicentres are plotted and, moreover, the distribution
of the most severe shocks corresponds closely with that of the
most frequent. It will be noticed that most earthquakes
originate in two well defined belts a Circum-Pacific belt (68 per
:
cent, of all earthquakes) with offshoots from Japan to Yap

Island and from Central America through the West Indies ;
and a Mediterranean belt (21 per cent, of all earthquakes)
extending from the Cape Verde Islands and Portugal, through
Asia Minor and the Himalayas to the Dutch East Indies, with
a side branch from Tibet through China. Only 11 per cent.
3 65
.
DISTRIBUTION OF EARTHQUAKES
of earthquakes originate elsewhere. Expressed in another

way, the belts form two rings (cf. Figs. 209-10) : one enclosing
North America and most of Asia and Europe (known collec-
tively as Laurasia), and the other enclosing South America,
Africa and Arabia, India, Australia, and Antarctica (known
collectively as Gondwanaland) From Portugal to Burma the
two rings coalesce. These belts or rings are also those of
Tertiary and recent mountain building. Judged by the
frequency of catastrophic shocks, the most dangerous lands
are, in order, the Philippines, Italy, China, Asia Minor, Japan,
Mexico, and the Balkans.
In the immediate neighbourhood of active volcanoes
tectonic earthquakes are rare, though the latter may be
numerous within a hundred miles or so. Earthquakes tend to
occur most frequently and severely on the outer sides of
mountain and island arcs, especially where the slopes are
steep (as off Japan and the Philippines), whereas volcanoes are
generally aligned at some distance on the inside of the arcs
(see Figs. 211 and 213). In the earthquake belts, steep slopes
are direct consequences of geologically recent mountain build-
ing, and tectonic earthquakes often originate beneath steep
slopes because they result from the present-day continuations
of the same movements.
A minor belt of epicentres extends from Spitzbergen and
Iceland, along the mid-Atlantic “ swell,” to Bouvet Island in
the far south. Another runs from the Nile through the rift
valley region of eastern and central Africa (see page 433) with
,
a side branch from the Gulf of Aden through the Indian Ocean
east of the Seychelles. Even in the more stable regions of the
continents and ocean floors, sporadic shocks occasionally occur.
In 1929, for example, a powerful earthquake originated be-
tween Nova Scotia and Newfoundland. No place can be
regarded as permanently immune from shocks. Earthquakes
are rare in Britain and most of those that do occur can be traced
to belated movements along ancient faults such as the Great
Glen Fault along the Caledonian Canal and the Highland
Border Fault between the Grampians and the Midland Valley
of Scotland (see Fig. 180).
367
EARTHQUAKES
Seismographs and Seismic Waves
From the focus of an earthquake, waves are propagated

through the earth in all directions, and when they arrive at a
seismological station they are recorded on seismographs, pro-
vided they are not too vigorous to put the instrument out of
action. Fig. 188 shows the essential parts of a common type
of horizontal seismograph. The vibrations of the ground are
transmitted to a delicately poised, weighted boom which
swings horizontally against a massive support which is firmly
attached to the ground. The weight tends to remain stationary,
Fig. 188
Diagram to illustrate the essential parts of a horizontal seismograph of the
Milne-Shaw type
and the movement of the end of the boom thus reproduces

the horizontal components of the vibrations of the support.
A small mirror attached to the end reflects a beam of light on
to photographic paper wrapped around a drum which rotates
on a long screw, so that while rotating it carries the paper
along at right angles to the reflected beam of light. The
vibrations are thus continuously recorded on the paper with
a magnification that depends on the length of the reflected
beam of light.
The record of a distant earthquake has the appearance
illustrated in Fig. 189. A
first or primary pulse P is followed
by rapid oscillations ;
then comes a second pulse S, followed
368
EARTHQUAKE WAVES
by more oscillations ; and finally a third pulse L initiates the
“ long ” or “ main ” vibrations. The P or “ push-and-pull ”
waves are compressional waves like those of sound, in which
each particle vibrates in the direction of propagation. The
S or “ shake ” waves are distortional waves, in which each
Fig. 189
Record at Pulkovo, Russia, of an earthquake in Asia Minor (February 9, 1909).
The time interval S-P is3 minutes 43 seconds, corresponding to a distance of
1,400 miles from the epicentre. ( After B. Galit~in
)
particle vibrates at right angles to the direction of propagation.

The velocity of P depends on the density and resistance to
compression of the rocks traversed ; that of S on the density
and resistance to distortion. The deeper the waves go (until
the earth’s core is reached), the faster they travel. At each
Focus of
Earthquake* Record! ng
'/^5%g£gary -"'R' Siaiio

(S) Wav g
rel faS
iminary (P) Wave
<-
Fig. 190
Section through a segment of the earth showing the paths followed by the P, S,
and L waves generated by an earthquake at F and recorded at R
depth, P more rapidly than S, and is thus recorded first

travels
(Fig. 190). The L waves are those that are confined to the
crust by reflections up and down at its lower and upper sur-
faces ;
they follow a zig-zag path, and thus arrive later than
the others. Their existence shows that there is a crust, and
that the crust overlies a substratum with different properties.
3 69
EARTHQUAKES
When the times of travel of P and S to various stations are

plotted against the corresponding distances from the epicentre,
they fall on smooth curves which are the same for all distant
earthquakes (Fig. 191). An important feature is that the time
To illustrate how the epicentre of an earthquake can be fixed when its distances
from three suitably placed stations are known. A circle is drawn on a globe
around each of the three stations ( e.g Bombay, Tokyo, and Wellington), with
.
a radius corresponding in each case to the respective distance of the epicentre.

The epicentre lies at the point of intersection of the three circles
with the distance. Thus, when

interval, S-P, steadily increases
an earthquake recorded at any station, a measurement of
is
the interval, S-P, determines the distance of the epicentre and

370
'
LAYERS OF THE EARTH’S CRUST

the time of origin. Three distance determinations at three
well-spaced stations serve to fix the actual position of the
epicentre (Fig. 192).
The Structure of the Earth’s Crust

Seismograph records are of great interest to geologists
because they provide the most powerful available means of
exploring the earth’s interior. The P and S waves of distant
earthquakes descend far beneath the crust, but those of near
earthquakes (within a few hundred miles of the stations where
they are recorded) travel through the crustal rocks with velo-
Recording
BHi WMM
(a)
1
Lower o”
n Ultrabasic Layer
i
_ *
p
_ ,11
Si
6 , —> Pg S S i
j Sg,,<n
'l)
j
I'
ji
LjnslldVdS
^Contmag
Fig. 193
(a) Diagrammatic section through the crystalline layers of the continental crust
to show the probable paths followed by the six waves of P and S types observed
in the record (b) of a near earthquake. [After H. Jeffreys.) Waves traversing
the upper layer are distinguished as P and S those traversing the inter-
t K ;
mediate layer as P* and S* ;

and those traversing the lower layer as P and S
citiesthat are characteristic of the rocks through which they

pass. The waves that travel through the upper layer of the
continental crust generally have velocities of 5-4 (P and 3-3 g)
(Sg) km. per second (Fig. 193). These velocities agree with
those calculated from the densities and elastic constants of
granitic rocks, and it is therefore inferred that the upper layer
of the continents is made of granitic rocks. Waves that have
traversed overlying “ veneers ” of sedimentary rocks, P s and
S s have lower velocities, and thus give on the records pulses
,
which follow those of Pg and Sg and can therefore be distin-

,
guished. Other waves dive below the granitic layer before

37 1
—
EARTHQUAKES
n — yy
y
s
/
-O
c ^y' /
/
0 /
u
01 «> 4y /
/
<D
o_ V / J. y / /
/
4
s — /y s
X V <?/
<>/
-v /
c
X
/
/ 4 /
/
'
VJ
o /
<7/
CD /
> /
/ /
/
Soot /
s
f Incr easing De bth
[I
0 10
|
20
Lyl
L.-- .
I
30 40
1
Period of
50
i
Waves ir
60
i
seconds
I
70
Fic. 194
Diagram showing the velocities of long waves of different periods which have
traversed the crust of the regions indicated. The longer the period the deeper
is the penetration of the wave concerned. The results demonstrate the contrast
in structure between the continental crust (with a granitic layer) and the Pacific
crust (without a granitic layer), and show that the granitic layer is relatively
thin or patchy beneath the other oceans
returning to the surface, and pass through a series of inter-

mediate layers with velocities ranging from 6-0-7-2 (P*) and
3-5-40 (S*) km. per second. These velocities correspond
approximately to those of plutonic and metamorphic rocks
having a composition like basalt. Waves that penetrate still
deeper, like those of distant earthquakes, pass through the
lower layer with velocities of 7-8 or more (P) and 4-35 or more
(S) km. per second. These correspond to the heavy rock
materials of the sima, possibly comparable in composition to
peridotite but probably differing in many respects from any
known surface rocks.
The thicknesses of the continental layers can be estimated
372
(B) BuckFing of railway lines near Tokyo
JAPANESE EARTHQUAKE OF 1st SEPT. 1923

EXPLORING THE EARTH’S INTERIOR
from the travel times of the various waves, the results being as
follows :
Sedimentary layer .... 0-10 km. (widely variable)

Upper or Granitic layer . . 10-15 km. (becoming much thicker under
the Alps and the Caucasus)
Intermediate or Basaltic layers 20-30 km. (becoming much thinner under
the Alps and the Caucasus, but thicker under
the Sierra Nevada (U.S.A.)
Lower layer Continuing down to the earth’s core at a depth
of about 2,900 km.
The structure of the oceanic crust is less well known,

because there are few island stations at which earthquakes
can be efficiently recorded. From the velocities of L waves
which have travelled along the oceanic crust before being
recorded it appears that the granitic layer is missing from the
crust underlying the greater part of the Pacific, and that
it is present only in relatively thin patches below the Atlantic
and Indian oceans (Fig. 194). Most of the oceanic crust seems
to be like that of the deeper parts of the basaltic layers beneath
the continents. At greater depths the materials have the same
properties beneath both oceans and continents.
The Structure of the Deep Interior
The P and S waves of distant earthquakes reach great depths

and their travel times indicate that their velocities increase
with depth from the figures given above to about 13 (P) and
7 (S) km. per second at a depth of 2,900 km. Relatively rapid
changes occur at about 400 and 700 km., but whether these
correspond to changes of composition or of state (e.g. from
crystalline to glassy) is not yet known. However, at 2,900 km.
there is a most conspicuous change (Fig. 195). The waves
that just attain this depth emerge at the surface at places
about 11,000 km. from the epicentre of the earthquake con-
cerned. Stations lying up to 5,000 km. beyond this distance
record no P or S waves, though the L waves come along as
usual. At distances more than 16,000 km. from the epicentre
of the earthquake the P wave again appears and it continues
(39«)
373 25
MINUTES
Fig. 195
Section through the centre of the earth showing the wave paths (firm lines with
arrows), wave fronts (dotted lines), and arrival times (in minutes reckoned from
the zero time of the shock). Since there is a shadow zone free from P and S
waves for each such earthquake, it is inferred that the earth has a core which
refracts the deeper waves as shown in the diagram
to do so right up to the antipodes of the epicentre. Corres-

ponding to each earthquake there is thus a ring-like shadow,
free from P and S waves, as illustrated in Fig. 196. The part
of the earth which casts this shadow is called its core.
P waves pass through the core with a greatly reduced
velocity, but S waves are not transmitted at all. Since it is
characteristic of liquids that they have no distortional elasticity,
and that S waves cannot pass through them, it is inferred that
the earth’s core is essentially liquid. The material itself is
thought to be mainly iron with a small percentage of nickel,
partly because such a composition matches the high density
of the earth and the velocity of the P waves through the core ;
NATURE OF THE EARTH’S CORE
The shadow zone cast by the earth’s core in the case of an earthquake
originating in Japan
partly because it corresponds with the composition of iron

meteorites,which are probably samples from the interior of a
disrupted planetary body and partly because the presence
;
of abundant iron oxides in the crustal rocks implies a high

concentration of iron in the interior, just as in a blast furnace
a little oxidised iron remains in the slag, while most of the iron
sinks to the bottom.

C. Davison
The Japanese Earthquake of 1923. Allen and Unwin (Murby),
London, 1931.
Great Earthquakes. Allen and Unwin (Murby), London, 1936.
375
EARTHQUAKES
J. Milne ( revised by A. W. Lee)

Earthquakes and other Earth Movements. Kegan Paul, Trench, Trubner
and Co., London, 1939.
P. Byerly
Seismology. Prentice-Hall, New York, 1942.
R. A. Daly
Our Mobile Earth. Charles Scribner’s Sons, New York, 1926.
H. Jeffreys
The Earth : Its Origin, History, and Physical Constitution. Cambridge
University Press, 1929.
Earthquakes and. Mountains. Methuen, London, 1935.
B. Gutenberg (Editor)
Internal Constitution of the Earth. McGraw-Hill, New York, 1939.
376
Chapter XVIII
EARTH MOVEMENTS : MOUNTAIN BUILDING

The Nature of Orogenic Belts
As already indicated on p. 106, an orogenic belt is an elongated

structural zone of the earth’s crust in which the rocks have
been folded, crumpled, contorted or overthrust on a spectacular
scale. As a consequence of the thickening of the crust due to
intense lateral compression, long sections of each orogenic belt
are uplifted, thereby becoming mountain ranges, like the Alps
and the Himalayas, with a varied relief of high peaks and deep
valleys carved by the agents of denudation. Most of the great
mountain ranges of to-day are parts of orogenic belts that came
into existence at various times of crustal unrest since the
Jurassic period, and particularly during the Tertiary period.
Others, like the Appalachians and the mountains of Scandi-
navia and Britain, represent older orogenic belts that were
deeply eroded long ago, but have since been rejuvenated, so
to speak, by uplift associated with the crustal compression of
Tertiary times.
While all the present-day ranges of folded mountains re-
present uplifted orogenic belts, it must be clearly realised that
by no means all orogenic belts are now mountainous. Most
of the Pre-Cambrian orogenic belts and certain stretches of
the later ones have lost their original mountainous relief by
long continued denudation, so that the rocks now exposed
to view ( e.g along the mainland and island shores of southern
Finland (Plates 10 and 11) are those which were formed by
metamorphism and igneous activity far below the surface at
the time when the region was undergoing active deformation.
Some sections of orogenic belts are submerged below sea level.
Between Scandinavia and Britain, for example, the Caledonian
orogenic belt beneath the North Sea.
lies The Black Sea,
which has only recently been formed by subsidence, hides the
377
EARTH MOVEMENTS ! MOUNTAIN BUILDING
submerged connecting links between the Caucasus and the

Crimea and the Balkan Mountains. It is therefore essential
to discriminate carefully between the geographical concept
of a mountain range or system and the geological concept of an
orogenic belt the one refers to the height and relief of the
:
land ;
the other to the structure of the rocks, whether the
region be high, low, or submerged. A mountain system is
the whole series of ranges belonging to an orogenic belt. The
term cordillera is sometimes used for a broad assemblage of
—
ranges such as that of western North America belonging to —
more than one system (see Figs. 201 and 219).
We have already learned that mountainous orogenic belts
have deep sialic roots which go down to depths comparable
with the whole thickness of the crust. It follows that the
compression responsible for the development of an orogenic
belt disturbs all the rocks down to a very great depth. In all
the greater crustal revolutions the previous structures of the
rocks are entirely altered by folding and thrusting in the
:
upper levels, and by flowage and metamorphism below, cul-

minating in fusion and igneous activity. Indeed, it is only in
the orogenic belts that the complete cycle of rock transforma-
tion (page 66) is achieved on a regional scale. Changes of
—
chemical composition as in the transformation of sediments
to schists and migmatites, and even to granites are brought —
about by hot migrating fluids. Magmas are generated in the
crust others ascend from greater depths
; and valuable
;
deposits of metallic ores are locally introduced. Orogenesis

thus involves not only great lateral compression, but also the
heating of the rocks and the soaking through them of chemi-
cally active fluids. The tectonic, thermal, chemical, and
magmatic changes that accompany the great crustal revolu-
tions are thus in striking contrast with the relative passivity
of the wide intervening regions where the dominant movements
are epeirogenic (p. 107). In these areas only slow fluctua-
tions of level take place, accompanied by fracturing and
faulting. Folding, if it occurs at all, is local and on a limited
scale.
378
DOWNWARPING AND SEDIMENTATION
Geosynclines
By the work of several generations of geologists it has been

firmly established that the orogenic belts of each geological
era originated in long downwarps of the crust in which extra-
ordinarily thick deposits of sedimentary rocks accumulated
during the era (or eras) that preceded the orogenic revolution.
The first important step towards understanding the natural
history of folded mountains was taken nearly a century ago
A B
Initial After sedimentation
condition up to sea level
Base of crust
depressed by
displacement of
(h-iooY of sima (3-4)
Fig. 197
Diagram to illustrate the isostatic response of the earth’s crust to sedimentation
by the brothers W. B. and H. D. Rogers. From their study

of the Appalachians these two geologists discovered that the
folded sediments out of which the ranges are built are shallow
water marine types which locally reach a thickness of 40,000
feet. In the unfolded regions of the Interior Lowlands to the
west the sediments of corresponding age are only a tenth or
a twentieth as thick.
The accumulation of several miles of sandstones, shales,
379
and limestones clearly implies that the original floor of the belt
must have subsided by a like amount. The mountains were
evidently preceded by the gradual development of a deep
trough in which sedimentation more or less kept pace with
the downwarping of the crust. Such elongated belts of long
continued subsidence and sedimentation were called geosyn-
clines by Dana in 1873.
The early pioneers thought that the weight of the accumu-
lating sediments was itself sufficient to depress the crust, so that
room was automatically provided for still more sediments.
Any such effect, however, is strictly limited. Suppose the
initial depth of sea water to be 100 feet (stage A in Fig. 197)
and that marine sediments of density 2-4 accumulated, and
depressed the crust isostatically, until the region became com-
pletely silted up (stage B). Let the maximum thickness of
sediments so deposited be h feet. The crust is depressed by
(h - 100) feet, and this must also be the thickness of the deep-
seated sima, density 3-4, displaced at the base of the crust.
At stage B
the weight of sediment added is proportional to 2-4 X k
the weight of water displaced is proportional to 1-0 X 100

the weight of sima displaced is proportional to 3-4 X ( h —
100).
For be maintained the weight

isostasy to lost must be equal
to the weight gained. Thus, we have
2-4 A=100-j-3-4 h - 340 ;

whence k = 240 feet.
For 40,000 marine sediments to accumulate under such

feet of
conditions, the initial depth of water would have had to be
nearly 17,000 feet. Actually, however, the water was very
shallow to begin with, as shown by the abundance of shore
and deltaic deposits. It follows that the weight of sediments
does not depress the crust to any significant extent. It is the
independent downwarping of the crust that makes room for
the sediments to accumulate. A geosyncline is essentially a
result of earth movements, the precursors of the vigorous
revolutionary movements that follow at a later date. It may
be noticed in passing that the average rate of sinking is ex-
380
NATURE OF OROGENIG BELTS
tremely slow. In the Appalachian example, 40,000 feet in
300,000,000 years (Cambrian to early Permian) is only one
foot in 7,500 years. However, this very slowness makes the
cumulative effect seem even more remarkable.
The continual supply of sediments implies that the land
bordering at least one of the shores of a geosyncline must have
been rising while the floor of the geosyncline was sinking.
Near the shore, or wherever the curvature of the warped crust
is greatest, fractures and faults may develop. Volcanic
activity, evidenced by the occurrence of lavas and tuffs, inter-
bedded with the sediments of many geosynclines, would be
expected to take advantage of such fractures through the crust.
In Britain vulcanism broke out on a large scale during the
Ordovician period, while the Caledonian geosyncline was
developing (Figs. 51 and 200). Volcanic rocks first appeared
at Rhobell Fawr in Wales and on both sides of the Midland
Valley of Scotland (Fig. 180). Later, the activity spread
towards the middle of the geosyncline and brought to the
surface the lavas and tuffs which are responsible for the rugged
scenery of Snowdonia in North Wales (Plate 54b), and of
Borrowdale (Plate 37) in the Lake District. Such volcanic
activity indicates that the rocks beneath the geosyncline were
being heated up as well as depressed.
Structures of Orogenic Belts and their Implications
Some of the structures characteristically impressed by sub-

sequent compression upon the vast thicknesses of sediments
(and of volcanic rocks, when present) have already been illus-
trated. These structures include alternations of more or less
open anticlines and synclines (Plates 12 and 80) ;
tightly
compressed isoclinal folds (Fig. 26 and Plate 69a) ; recumbent
folds (Fig. 27 and Plate 81) ;
and thrusts and nappes (Fig. 33).
A folded mountain range is a linear tectonic unit, straight or
arcuate, in which the axes of the folds are generally parallel to
the main trend of the range. The axial planes, recumbent
folds, and thrusts are all (except for parts of an occasional
381
anticlinorium or synclinorium, Fig. 26) directed upwards and
outwards from the interior of the geosyncline towards one of
the unfolded crustal blocks which margined the geosyncline.
The crustal block towards or over which the structures splay
Folded and Overf-hrusf
Folded and Overfhrusl-
NE Border W. Border Ranges-
Median Mass-
Ram Iranian Plateau Zagros Ml? Alluvial Plain
Turkmen of Mesopotamia
Fio. 198
Hypothetical section across the symmetrical orogenic belt of Iraq and Iran,
indicating the approach of the forelands and the outward thrusting of the margins
of the geosynclinal block (folding of the marginal ranges omitted)
out is called the foreland. Most geosynclines have two forelands,

but where a geosyncline developed alongside a continental
margin (as in the Dutch East Indies) the outer foreland is the
ocean floor beyond the zone of depression. The resulting
mountain system is thus characteristically bilateral, and con-
sists of two unilateral bordering ranges (or series of ranges),
each having its structures directed outwards, away from the

axis of the geosyncline. Each bordering range, taken by itself,
Overthrust Folded and Overrhrust

N Border Ranges- Median Mass: S Border Ranges-.
Kwen Lun Tibetan Plateau Himalayas
Indo-GanqeNc
Alluvial Plain
Fig. 199
Hypothetical section across the asymmetrical orogenic belt of the Himalayas and
Tibet (folding of the marginal ranges omitted)
is a unilateral range, in which all the structures splay out the

same way, that is, towards the neighbouring foreland.
This unity of structure is particularly well displayed by the
Alpine system of ranges (Figs. 198 and 199). The Alpine-
Himalayan system as a whole extends from each side of
382
BILATERAL STRUCTURE OF THE ALPINE SYSTEM
Gibraltar on the west, to the Himalayas and the East Indies

on the east. It originated in a long, wide, and composite geo-
syncline which developed between the northern foreland of
Europe and Asia (Eurasia) and the southern foreland of Africa,
Arabia, and India. To distinguish this immense seaway from
the present Mediterranean it was called the Tethys by Suess.
By subsequent compression two main sets of ranges originated :
a northerly set, including the Pyrenees, the Alps, the Car-

pathians, and the Caucasus ; and a southerly set, including the
Atlas Mountains, the Apennines, the Dinaric Alps, the Tauric
ranges, and the Himalayas. Where these bordering ranges
lie far apart there may be a broad intervening region of sea,
plain, or plateau. Examples of these are the Western Mediter-
ranean, the Hungarian plain, and the high plateau of Tibet.
Such intermontane, relatively unfolded regions are called
median areas. Where there is no median area and the bordering
ranges are closely compressed together, back to back, the
bilateral structure may be roughly symmetrical, but it is
generally conspicuously asymmetrical. The W estern Alps, for
example, are so highly asymmetrical as to be almost unilateral,
all the thrusts and recumbent folds, except in the extreme
south, being directed towards the European foreland (Fig. 204).
The direction of overfolding and thrusting, and of the
crustal compression responsible for the over-riding movements,
is naturally described as it is seen at the surface ;
that is to
say, with reference to the underlying rocks, which are im-
plicitly regarded as having remained stationary. Thus the
overthrusts of the North-West Highlands (Fig. 33) or the nappes
of the Western Alps (Fig. 204) are said to be directed towards
the north-west as a result of pressure from the south-east. But
this description is Exactly the same effect
purely relative.
would be attained if the foreland rocks had been underthrust
towards the orogenic belt. Now in bilateral systems we
observe that on each side the overthrusting and the correspond-
ing pressure appear to have come from within the system.
Pressure could not, however, have operated outwards in two
diametrically opposed directions unless the deeper levels of
the geosynclinal belt had been powerfully compressed from
383
outside, that is, by the inward movement of the two forelands.
The crustal blocks of the forelands have, in fact, acted like the
jaws of an irresistibly closing vice, underthrusting the geosyn-
cline,and so causing its sediments and floor to crumple up and
splay out on both sides. The crustal shortening involved in the
folding and thrusting can often be roughly estimated by imagin-
ing the structures to be all straightened out ;
commonly
it
amounts to many tens of miles. As the forelands move in-

wards, the rocks in depth, already downwarped during the
growth of the geosyncline, continue to buckle downwards
(Fig. 212) with consequent formation of a deep mountain
root. Only the upper strata of the geosyncline itself tend to
wrinkle upwards. The general uplift of the region into a belt
of highlands follows later, as a result of isostasy (page 15).
Orogenic Belts of Europe
The general outlines of the tectonic framework of Europe

are shown in Fig. 200. The oldest part of the continent is the
Baltic Shield. broad region of geologically ancient
This
crystalline rocks hasremained a stable land area ever since
pre-Cambrian times. It has fluctuated in level from time to
time, but on the whole the movements have been slow gentle
uplifts which maintained the surface above sea level, despite
the ravages of denudation. Towards the east and south,
however, the old rocks of the Shield are covered by a veneer
of flat-lying sediments, deposited during the Palaeozoic and
later periods. This buried extension of the Shield, known as the
Russian Platform, evidently subsided a little at intervals and
was flooded by shallow seas, just as part of the Shield is flooded
to-day by the Baltic. North of the Black Sea, and in a few
other isolated spots where the sedimentary blanket has been
removed, the shield rocks reappear at the surface. The north
German plains and probably the English Midlands represent
a westerly continuation of the Russian Platform, although in
this section the crust was less stable, and certain basins of
—
sedimentation that were deeper than usual including most
384
Tectonic framework of Europe
Fig. 200
Tectonic map of Europe. The Alpine orogenic belt is outlined by thick black
lines with arrows indicating the outward thrusts and overfolding towards the
forelands
of the great coalfields of Britain —have suffered a certain amount

of folding.
Thestable triangle of the Shield and Platform is bordered
on three sides by clearly defined orogenic belts, towards
its
each of which it acted as a rigid foreland. On the north-west

the Caledonian belt extends through Scandinavia to Britain.
Here the rocks of a Lower Palaeozoic geosyncline were intensely
and invaded by granites during late Silurian
folded, overthrust,
and early Devonian times. In Scandinavia the south-eastern
front of the belt is well preserved. Outward thrusting towards
the Shield is conspicuous in many places and has involved
movement of possibly as much as 80 miles. Within the belt
itself the rocks are contorted, metamorphosed, and granitized
385
EARTH MOVEMENTS t MOUNTAIN BUILDING
so intensely that over large tracts the structures baffle analysis.
The north-western front of the belt is cut off by the Atlantic,
except in the N.W. Highlands of Scotland, where the thrusts
illustrated in Fig. 33 splay out over another shield area, of
which only a narrow strip now remains as land. This relic
of the north-west foreland probably represents the extreme
limit of the great Canadian Shield of North America and
Greenland (Fig. 201). The south-eastern front of the Cale-
donian belt is poorly defined in Britain. Much of it is hidden
by later sediments, and where the older rocks are exposed
as in Shropshire — the folding is open and broadly undulating.
While the Caledonian movements were in progress, other
geosynclines began to develop and to fill up with sediment :
(a) along the site of the Urals and (b) across Central and
;
Southern Europe from the promontories of south-west Ireland

to north of the Sea of Azov. These became transformed into
orogenic belts during Carboniferous and early Permian times.
This orogenesis and the structures and mountains that resulted
areknown by the name Hercynian (after the Harz Mountains).
Most of the Uralian belt is continuously preserved, but the
much wider Hercynian belt of the south has become broken
into a series of isolated blocks or massifs.
These include south-
west Ireland, South Wales, and Cornwall and Devon Brittany
;
and the Central Plateau of France the Ardennes, Vosges, and

;
Black Forest (Plate 82 a) and the Harz and Bohemian Moun-

;
tains. The depressed regions between are buried beneath later

sediments, but here and there ( e.g in south-east England)
.
the Hercynian foundation has been encountered in borings

and mining operations. There is no doubt that all the massifs
referred to are relatively uplifted parts of a belt that is con-
tinuous in depth.
The southern margin, and locally the whole, of the Hercy-
nian belt has vanished from sight for another reason. In
Mesozoic times considerable stretches of the Hercynian tract
became submerged beneath the Tethys, so forming the greater
part of the floor of the Alpine geosyncline. Other portions,
such as the massifs of Mt. Blanc and the Aiguilles Rouges,
remained as islands near the northern shores of the Tethys
386
Caledonian and hercynian belts
(Plate 82b). These, together with parts of the floor itself,
were incorporated in the Alpine ranges which later completed

the southern framework of Europe.
In most places the northern Hercynian front lies well to
the north of the northern Alpine front, but the Carpathians
were thrust forward beyond the Hercynian front, so that here
the Russian Platform became the foreland. In Silesia the
structural relations between the mountains and their foreland
have been made clear by mining operations. The broad
coalfield — —
of Silesia belonging to the foreland was partly
overridden by the advancing nappes of the Carpathian arc.
Nevertheless, the buried half of the coalfield has been located
beneath the nappes by borings sunk through them into the
coal seams beneath.
Just as the Alpine structures encroach on the older Her-
cynian belt, so in the British Isles (see Fig. 258) the Hercynian
structures in turn encroach on the still older Caledonian belt.
From South Wales to the south-west shores of Ireland the
Lower Palaeozoic rocks with their typical Caledonian structures
disappear beneath the folded and overthrust Devonian and
Carboniferous sediments which represent the northern Her-
cynian front. The two orogenic belts, gradually converging
across Europe, ultimately meet, and the northern Hercynian
front begins to cross the south-eastern Caledonian front.
What is probably the completion of the crossing is found in
the Appalachian Mountains on the other side of the Atlantic.
The Appalachians
These mountains (Fig. 201) appear to be a closely knit
complex of two systems which roughly correspond to the
Caledonian and Hercynian of Europe, though in each case
the main orogenesis was slightly later in age. Originally, the
Appalachians were regarded as the standard example of a
unilateral mountain system, with the structures all directed
towards the Canadian Shield (a stable region of the same type
as the Baltic Shield) and its buried continuation beneath the
387
Fig. 201
Tectonic map of North America
Interior Lowlands (a region corresponding to the Russian

Platform). This apparent simplicity, however, only reflects
the fact that along the Atlantic seaboard the south-eastern
side of the orogenic belt is everywhere concealed by the
later sediments of the Coastal Plain (Fig. 202). The foreland
on the south-east is not seen, because the Atlantic now occupies
the site where weshould naturally look for it. But that there
certainly was such a forelandis firmly established by the ob-
servation that the geosynclinal sediments become coarser as

they are traced towards the Atlantic and include deltaic
deposits which spread out towards the north-west. Clearly
the rivers that supplied these sediments must have drained a
388
STRUCTURE OF THE APPALACHIANS
land that lay to the south-east, a land that must have been
undergoing denudation throughout the greater part of
Palaeozoic time.
In the north the overthrust Appalachian front begins in
Newfoundland. Continuing along the line of the St. Lawrence
as far as Quebec, it then turns south towards New York.
During the Devonian all this northern section was folded,
overthrust, invaded by granite, and uplifted by an orogenesis
that corresponds with the later phases of the Caledonian
revolution. South-west of New York the Caledonian part of
the chain is probably represented by the “ Older Appala-
chians ” (Fig. 202), where the old basement rocks, repre-
senting the floor of the geosyncline, intensely deformed and
penetrated by granite, were overthrust towards the north-west.
In this region only a few infolded and highly metamorphosed
NEWER APPALACHIANS OLDER APPALACHIANS
Folded Sedimentary Rocks Upthrust Crystalline Rocks

CambrUirCar bom f* rout
Fig. 202
Section across the composite orogenic belt of the southern Appalachians
remnants of the geosynclinal sediments now remain. We see

deep into the heart of the former mountains.
Inland, however, between the Blue Ridge and the Appala-
chian Plateau, the sediments are largely preserved in a broad
series of deep folds. These are generally open, as shown in
Fig. 202, but in places, e.g. the slate regions of West Virginia,
the folding is isoclinal and is locally broken by thrusts, again
indicating movement towards the north-west. Still farther
inland, in the Plateau, the folding gradually flattens out. In
these “ Newer Appalachians ” not only Lower Palaeozoic
sediments are found, but also a far thicker sequence belonging
to the Upper Palaeozoic. The whole assemblage was folded
during the main Appalachian revolution, which began in the
Carboniferous and reached its climax in the Permian. The
“ Newer Appalachians ” thus correspond with the later phases
of the Hercynian orogenesis. By measuring the folds it has
t 396 26
) 389
been estimated that the folded “ ridge and valley ” tract, now
about 65 miles across, had an original width of about 100
miles. The forelands thus approached by at least 35 miles.
Many batholiths and stocks were emplaced in the heart of the
growing range, particularly in the adjoining “ Older Appala-
chians ” which must therefore have shared in the later oro-
genesis.
Farther north, in the New England States, relics of strongly
folded Carboniferous rocks mark the continuation of the
“ Newer Appalachians.” But here they lie on the inner side
of the Caledonian front of the Appalachian system, whereas
”
in the south the folded belt is on the outer side. The “ Newer
or Hercynian front appears on the coast near Boston, reaches
the “ Older ” or Caledonian front behind New York, and then,
as Bailey puts it, “ steps clear of its Caledonian predecessor.”
The crossing of the Caledonian belt by the outer front of the
Hercynian belt, begun in Britain, is finally accomplished in
America (Figs. 201 and 258).
The Western Alps

The folded structures of the Appalachians are simple and
straightforward compared with the amazing complexity of the
recumbent folds and far-driven nappes of many of the Alpine
ranges. Since Escher’s discovery in 1841 of gigantic thrusts
in the Swiss Alps, many brilliant geologists have devoted their
professional lives to the unravelling of structures which, for
many years after they were first described, appeared to be
quite incredible to those who had not actually traced them
from peak to peak. By patient mapping, supplemented by
underground observations made possible by an incomparable
series of tunnels, the intricate tectonic pattern of the Swiss
Alps now known with a wealth of detail superior to anything
is
that has been achieved in the other great mountain systems of

the world.
Geologically the Alps are divided into the Western Albs,
which curve in a broad arc from the Mediterranean to Lakes
390
STRUCTURE OF THE ALPS
Tectonic map of the Alps. Hercynian massifs are indicated by close dotting
(A=Aar Massif; G=St. Gotthard Massif ;R= Aiguilles Rouges; B=Mt. Blanc)
Constance and Como, and the Eastern which continue
Alps,
in a gentler curve towards the Danube
Beyond(Fig. 203).
Vienna the vast bow of the Carpathians begins, while on the
southern side the ranges of northern Italy swing round into
the Dinaric Alps. It is in the Western Alps, and particularly
in Switzerland, that the key to the general structure has been
revealed. The essential feature, as portrayed in Fig. 204, is
the occurrence of a series of gigantic recumbent folds and

nappes, each of which has been driven forward for many
miles towards the foreland, and in many cases far across it.
The Alpine rivers have cut deeply into the nappes, thus
exposing the underlying rocks in many a steep-walled valley.
If the nappes were everywhere at the same level, only the
39 1
outer parts of the structure would be exposed to view in this

way, and even the tunnels would add little more. But as the
nappes are traced along the trend of the ranges from S.W. to
N.E. they are found to undulate up and down in an alternating
Matterhorn
Weisshorn ' Monte Rosa Nappes Nappes
High Calcareous .. ofthe otrne
Nappes .... VJ ''[pstem Dmarlc
Pre-Alps Alps
'
Eiinpion Nappes Z ,
African S.E.
ti i c /-;// Zone of Roots rorehnd
Mai s f tfjs'ef
Fig. 204
Tectonic section across the Western Alps ( After R. Staub)
successionof broad culminations and depressions (Fig. 205).

Where the depressions carry the nappes downwards, the
uppermost ones are still well preserved in the mountain peaks.
Where the culminations raised the upper nappes to levels now
far above the highest peaks —
that is to say, where they have
already been swept away by erosion the lower nappes come —
Fig. 205
Longitudinal section showing the nappes of the Fligh Calcareous Alps exposed
in the tectonic depression between the culminations of the Aiguilles Rouges and
the Aar Massif. The nappes advanced at right angles to the section in the
direction away from the observer
to the surface and are themselves cut through by the valleys.

Thus, although no single section across the mountains provides
more than part of the picture, the whole complicated structure
can be visualized by taking a series of several sections in order
across the successive culminations and depressions. Finally,
392
PLATE 81
[/'. A\ Asha oft
Recumbent Calcareous Alps. The Axenstrasse, near Fluelen,

folds in the Hie^h
south end of Lake Lucerne. Switzerland
-
SUBDIVISIONS OF THE WESTERN ALPS

in the Eastern Alps, all the structures seen in the Western Alps
disappear beneath a widespread cover of a still higher series
of nappes. Thus, because the evidence is largely hidden, the
structure of the Eastern Alps is much less well known. Only
here and there, as in the Engadine and the High Tauern
(Fig. 203) has erosion removed the cover and so opened
“ windows ” in the surrounding framework through which
the underlying structures are locally seen.
The chief subdivisions of the Western Alps as seen in plan,
subdivisions representing successive zones which provide a
first clue to the general structure, can be clearly detected from
the air. Each has its distinctive topography, and each comes
into view in turn from north-west to south-east during a flight
from, say, the Jura Mountains to Milan. In order, these
zones are as follows :
1. The Jura Mountains an arcuate bundle of hills, like a

:
crumpled-up tablecloth, pushed forward (while the Alps were

advancing) across the gap between the Central Plateau of
France on one side and the Vosges and Black Forest on the
other, and thus standing well in front of the Alps proper.
On the inner side the strata are mainly folded into bold, open
anticlines,some of which form the actual hills (Plate 80), but
the outer side is a tableland broken by faults into an irregular
mosaic of strips and blocks.

2. The Swiss Plain a broad lowland, filled with soft Tertiary
:
sediments called molasse, derived from the denudation of the

rising Alps. Far to the south-east the edge of the High Cal-
careous Alps appears in the distance like a great wall (Fig. 206).
On each side of Lake Geneva a broad patch of foothills pro-
trudes on to the plain. These are —
3. The Pre-Alps :extending between Lake Thun and the
River Arve. They consist of an isolated pile of nappes, much
folded and sliced by minor thrusts, which were driven over
the molasse far from their roots in the south-east. The strata
include types which are entirely different from anything seen
in the Jura or the High Calcareous Alps. The Pre-Alps are
thus completely foreign to the district where they came to
393
[P. Hausei
Fig. 206
The Swiss Plain and the High Calcareous Alps viewed from the Jura Mountains
near Solothurn
rest. Exactly where they came from remains the chief un-
solved problem of Alpine tectonics. The uppermost of these
far-travelled rock sheets may even represent the remnants of a
series of nappes that formerly covered the Western Alps in
continuation of the upper nappes of the Eastern Alps. It is
possible that the later part of the forward journey of the Pre-
Alps was accomplished by down-sliding —like a gigantic
landslip.
4. The High Calcareous Alps :a high range of rugged moun-

tains whose structure is dominated by a series of clean-cut
overthrusts (Fig. 204). The zone includes the Bernese Ober-
land with its snowfields and its many familiar
and glaciers
peaks ( e.g . the Jungfrau,
13,669 ft.). Each of the nappes
is composed of sediments which were deposited along the
northern margin of the geosyncline. Locally, many of the

nappes are themselves intensely folded (Plate 81).
394
SUBDIVISIONS OF THE WESTERN ALPS
5. The Hercynian Massifs : an arcuate chain of long isolated
blocks consisting largely of crystalline rocks. These have
been sheared by innumerable small thrusts, and except for the
more massive granites they a ready prey to the splintering
fall
action of frost (Plate 82 b). The skylines are in consequence

characteristically jagged, as in the appropriately named
Aiguilles Rouges. The latter massif, together with the adjoin-
ing massif of Mt. Blanc (15,732 ft.), emerges along the crest
of a great culmination. To the north-east the Hercynian
foundation disappears beneath the nappes of the High Cal-
careous Alps, to emerge again in another culmination as the
Aar and St. Gotthard massifs (Figs. 203 and 205). Here the
Rhine and the Rhone have their sources. The Rhone flows
to the west-south-west through a long trough-like valley
which, after leaving the Hercynian massifs, follows the bound-
ary between the High Calcareous Alps and the broad zone of
the Pennine Alps.
6. The Pennine Nappes an involved series of six great nappes,

:
all of which were squeezed out of the main geosyncline as

gigantic recumbent folds. Each nappe has a core of older
rocks, mainly gneisses, representing the floor of the geosyn-
cline ; wrapped round by an envelope of lustrous schists and
crystalline limestones, representing the sediments (and volcanic
rocks) of the geosyncline. The Pennine Alps (from which our
own modest Pennines take their name) is a lofty region of
boldly carved mountains, rising above the snowfields into
pyramidal peaks, of which the Matterhorn (14,705 ft.) is the
most famous (Plate 50), although the less shapely Monte Rosa
(15,215 ft.) is higher. To the east the Pennine nappes continue
at a lower level as the Lepontine Alps. Beyond the margin
of the Eastern Alps they remain unexposed except in the
windows of the Lower Engadine and the High Tauern, where
the two upper members of the series have been recognized.
As numbered in Fig. 204 from bottom to top, the Pennine
nappes are as follows :
I— I II The Simplon Nappes (comprising the three lower

nappes), which were pressed against the upstanding mass
395
of Mt. Blanc and the Aiguilles Rouges and thereby folded

into such intricate convolutions that their structure was
never made clear until they were penetrated by the Simplon
tunnel.
IV The which rides over the

Great St. Bernard Nappe,
Simplon nappes and also sends out a remarkable backward
bulge in response to the pressure exerted by the Monte
Rosa nappe.
V
The Monte Rosa Nappe, which, struggling to expand
within a writhing complex of plastic rocks, all competing
for space, found it easier to plunge into the back of the
St. Bernard Nappe than to ride over it.
VI The Dent Blanche Nappe, which drove far forward

over all the nappes in front, forming above them a wide-
spread carapace, of which the greater part has since been
removed by erosion.
7. The Zone of Roots a long narrow zone near the Italian

:
frontier,where the Pennine nappes turn vertically down, and

so appear to be rooted in the ground. Some idea of the
titanic compression which was involved in the making of the
PENNINE. NAPPES *
Simplon Great St. Bernard Monte Rosa Dent Blanche
High Calcareous Alps Nappes Nappe Nappe Nappe
Fig. 207
Reconstruction (in section from N.W. to S.E.) of a stage in the development
of the Tethys geosyncline and its northern shores, to illustrate the environments
from which the nappes of the Western Alps were driven
Alps can be gained by contrasting this tightly squeezed belt

with the great width of the geosyncline that is implied by the
sedimentary strata of the nappes (Fig. 207). While the great
squeeze was in progress, culminating in Miocene times, the
rocks must have been unusually plastic, and evidence that they
39 6
THE ZONE OF ROOTS
were lubricated by hot migrating fluids is furnished by the
prevalence of migmatites, with swirling structures due to
flowage, in the deepest parts of the roots. Later on, great
masses of granite were emplaced at various places along the
zone of roots (Plate 83).
8. To the south and west of our line of section the Pennine
Alps drop steeply to the alluvial plain of Lombardy. On the

east, however, the zone of roots gradually widens with the
incoming of still another series of structural elements. These
represent the southern side and foreland of the geosyncline,
and this time all the thrusts are directed towards the south.
From the obscurity of this highly metamorphosed complex the
upper nappes of the Eastern Alps emerge, and ultimately turn
over to the north, in similar fashion to the Dent Blanche Nappe.
Farther south the backward thrusts maintain their direction
and mark the beginning of the Italian ranges which pass round
the head of the Adriatic into the Dinaric Alps.
The probable relationships between the nappes of the

Western Alps and the geography of the obliterated geosyncline
are indicated in Fig. 207.
Orogenic Belts of the Alpine Revolution
For convenience the term Alpine Revolution is adopted to

cover the cumulative effect of all the orogenic movements
which have occurred at intervals since late Jurassic times.
As indicated by the following examples (which are far from
complete), the resulting orogenic belts are not all of the same
a§e -
Late Jurassic ( Nevadan orogeny ) Atlas, Caucasus, Sierra

:
Nevada of California, Japan, New Zealand.

Late Cretaceous ( Laramide orogeny) Pyrenees, Dinaric Alps,
:
Tauric Mountains, mountains of Malaya, Sumatra, and New

Zealand, Rocky Mountains, Greater Antilles, and Andes.
Lower and Middle Tertiary ( main Alpine orogeny) : Pyrenees,
Alps, Carpathians, Himalayas, and Asiatic Island festoons.
397
Upper Tertiary to Present Day ( Cascadian orogeny) Foothills:
of the Himalayas (Siwaliks), Banda arc of the Dutch East

Indies, St. Elias and the other Pacific Coast ranges of North
America, Greater Antilles, New Zealand.
In most of these Alpine orogenic belts two or more phases of
movement can be recognised, and in certain regions, particu-
larly in the East and West Indies, orogenic movements seem
still to be actively in progress. But taken as a whole the re-
sulting assemblage of mountain systems combines into a
Fig. 20S
Map showing the distribution of the Alpine and Circum-Pacific orogenic belts
pattern of striking simplicity. The usual way of expressing

this pattern is to describe it as a great ring encircling the
Pacific, combined with a Mediterranean belt, of which the
chief part is the Alpine-Himalayan system (Fig. 208). On the
east the latter swings into the Circum-Pacific ring by way of
the Dutch East Indies (the Burman or Banda arc, Fig. 211).
Here three arcs come together in spiral fashion in the “ Medi-
terranean ” region between Australasia and the extreme south
of Asia (a) the New Zealand-New Guinea arc
: b the
( ) ;
Philippine-Borneo arc and (c) the Banda arc. It is possible

;
that the Kwen Lun and neighbouring ranges north of Tibet

398
PATTERN OF THE ALPINE OROGENIG BELTS
continue across China to join the Circum-Pacific ring south-
west of Japan (Fig. 222). At the other end of the Mediter-
ranean belt the Sierra Nevada of Spain and the Atlas are
linked through Gibraltar by an inner connecting arc (Fig. 200),
but the main ranges remain separate and are cut off abruptly
by the Atlantic. Across the Atlantic two corresponding
mountain systems appear on each side of the Caribbean Sea
(the American Mediterranean) ; a northern one which can be
traced from the Cordillera of North and Central America
through the Greater Antilles as far as the Virgin Islands, and
a southern one, which swings round from the Andes through
Venezuela to Trinidad. The two systems, though originally
separate, are now being linked by an actively growing arc
surrounding the volcanic arc of the Lesser Antilles (Fig. 213).
Here the Circum-Pacific ring is being completed by a vast
loop which encroaches on the Atlantic. A similar, though
much older loop, connects the Andes of Patagonia with the
Antarctic Andes of Grahamland. It should be noticed that
the Costa Rica-Panama land bridge is of volcanic origin and
may possibly not belong to the Circum-Pacific orogenic ring.
Another way of expressing the pattern is to describe it in
terms of two gigantic rings (Figs. 209 and 210), one surrounding
the northern continents of North America and Eurasia (known
collectively as Laurasia), the other surrounding the southern
land masses of South America, Africa, Arabia, India, Australia,
and Antarctica (known collectively as Gondwanaland ) . The
first of these rings includes the northern ranges of the Medi-
terranean belts and the northern half of the Circum-Pacific

ring, while the second includes the southern ranges of the
Mediterranean belts and the southern half of the Circum-
Pacific ring.
Now, we have seen, the Tethys was a great “ Medi-
as
terranean ” geosyncline which developed between Laurasia
and Gondwanaland. Figs. 209 and 210 show that each of the
land masses referred to above is rimmed by an orogenic belt
wherever it is bordered by the Pacific or by the former site
of the Tethys. In plan Laurasia, Gondwanaland, and the
Pacific floor all appear to have moved radially outwards,
399
Fig. '209
Map showing the interrupted orogenic ring peripheral
to the continental masses
(unshaded) of Laurasia. The adjoining blocks of
Gondwanaland are dotted.
The probable continental movements directed outwards
towards the Pacific and
the Tethys are indicated by arrows
thus buckling the crust at their margins and forming orogenic

belts along these highly compressed
peripheral zones. The
Tethysitself was obliterated by the approach
of Laurasia and
Gondwanaland and transformed into the Mediterranean belts.
Within the peripheral ring of Laurasia lie the disruptive
basins of the North Atlantic and Arctic
oceans, the coasts of
which are essentially due to fracture and faulting. Similarly,
within the peripheral ring of Gondwanaland lie the South
Atlantic and Indian oceans. Here again the coasts are
essentially due and faulting, except in the case of
to fracture
the Burman where the continuation of the Tethys lay
arc,
along what is now the north-east margin of the Indian ocean.
Suess was the first to recognise that
the coastal structures of
the world were of two contrasted types,
which he distinguished
as Atlantic and Pacific. The coasts of Atlantic type are deter-
mined by fractures and subsidences which
characteristically
cut across the “grain” of the lands
Fig. (cf. 165), though
400
i .-l. G. Kihhbcrz
(B) Frost-splintered peaks of the Mt. Blanc Massif, with the Glacier de Tacul
and a large glacial erratic in front
HLRCYXIAN MASSIFS
Ashcroft
foreground
N.
[F. the
in
Soglio
of
illage
(lertiary).
Massif
Bergello
the
of
peaks
granite
ihe
and
Bondasco
Val
The
COASTS OF PACIFIC TYPE
Fig. 210
Map showing the interrupted orogenic ring peripheral to the continental masses
(dotted) of Gondwanaland. The probable continental movements directed out-
wards towards the Pacific and the Tethys are indicated by arrows
locally they may happen to be roughly parallel to one of the

older orogenic belts (as in Norway). The coasts of Pacific
type are determined by folding and are, in their general out-
“ grain ” of the lands
lines, parallel to the (cf. Fig. 166 ), being
fringed by bordering mountain chains, including island
festoons like those of Asia.
The Orogenic Cycle

In every continent there are mountain systems and older
orogenic belts of widely different ages. Each of the Pre-
Cambrian shields represents a coalescence of successive orogenic
belts from which the mountainous superstructure has long ago
been removed by denudation. In the Baltic shield there are
401
four such belts, and six are known in the Canadian shield.
The continents have developed by successive orogenic additions
to their shields, as illustrated in Figs. 200 and 201.
Detailed study of many of the orogenic belts has shown
that, although no two are exactly alike, all can be regarded as
variations on a common theme. The theme itself, or in other
words, the usual sequence of events involved in the evolution
of an orogenic belt, is called the orogenic cycle. In general terms
it may be summarised as follows :
(a) Development of a geosyncline with heavy sedimenta-

tion and occasional volcanic activity.
(
b) Compression of the belt by a orogenic phase, in-
first
volving root formation in depth, folding and overthrusting of

the superstructure, and uplift of the compressed zone in re-
sponse to the buoyant (isostatic) effect of the root.
(c) Lateral growth of the geosyncline by development of
a new subsiding tract outside the rising mountains (Fig. 199).
As the latter are carved into peaks and valleys by denudation,
they provide much of the sediment which fills up the depression.
(i d) Renewed orogenic compression of the whole belt.
Stages (c) and (d) may occur twice or even three times (rarely
more) in the more complex belts.
During the more vigorous orogenic phases, and par-

(e)
ticularly during the climax of the revolution, the deeper rocks

are intensely metamorphosed, and migmatites are formed by
hot migradng fluids. Later, granite batholiths are emplaced,
followed in some cases by the introduction of valuable ore de-
posits. These are gradually uncovered by denudation if the
cumulative uplift of the completed orogenic belt is sufficiently
long maintained.
It sometimes happens, however, that a mountain system
isreduced to a lowland or even submerged before much of the
superstructure has been denuded (Plate 21a). In such cases
the buoyant action of the roots must soon have ceased to
function. In other words, the roots themselves must have dis-
appeared, probably by becoming so plastic that, unable to
402
A PRESENT-DAY GEOSYNCLINE
support the overhead pressure, they flowed out laterally. In

response to subterranean changes of this kind the later be-
haviour of an orogenic belt is generally dominated by epeiro-
genic movements, often accompanied by volcanic activity.
These topics are dealt with in succeeding chapters.
Present-day Orogenesis
As already mentioned, the evolution of certain stretches of

the orogenic belts belonging to the Alpine revolution is still
actively in progress. Fig. 200 makes it clear that, unlike the
Western Mediterranean, which is a sunken median area, the
Eastern Mediterranean, the Adriatic, and the plain of Lom-

bardy are depressed regions lying along the “ African ” front
of the Alpine system. South of the Atlas this marginal de-
pression continues on the land, where it is marked by the
shotts of Tunisia and Algeria. Towards the east a similar
zone of depression can be traced through the alluvial plain
of Mesopotamia, the Persian Gulf, and the Gulf of Oman.
Still farther east, in front of the Himalayas, the depression
reappears in the alluvial plains of the Indus and Ganges. This

long zone of subsidence and sedimentation is probably the best
example of a marginal geosyncline now in course of develop-
ment. It illustrates the latest repetition of stage ( c ) in the
orogenic cycle.
In the Dutch East Indies the continuation of this geosyn-
cline can be traced for over 4,000 miles in front of the Banda
arc (Fig. 211). But here the active belt is very narrow, and
—
stage ( d ) —
the stage of orogenic compression has already been
reached. The trend line of this long strip is indicated by an
arcuate submarine ridge from which rows of rising islands
emerge above sea level at intervals. Proof of recent and con-
tinuing movements is furnished by the occurrence on the
islands of terraces of upraised coral reefs at heights ranging
from sea level to as much as 4,000 feet in Timor ; by transverse
fractures and tear faults, pointing to horizontal movements,
probably symptomatic of the advance of nappes or recumbent
403
Fig. 211
Tectonic map of the Banda Arc and
its links with the Philippine and New Guinea
arcs. The belt of negative anomalies of gravity discovered by Vening Meinesz

is shown (as far as it has been mapped) by a shading of fine dots. Active
volcanoes are indicated by black or white crosses
folds;
and by the extreme liability of the strip to frequent and
powerful earthquakes, some of which originate at very great
depths.
To all this Vening Meinesz has added a discovery of
fundamental significance, made in the course of an expedition
(1926) carried out in a submarine lent by the Netherlands Navy
for the purpose of making gravity measurements over the sea
floors of the East Indies. Meinesz found that the long strip
under discussion is characterised throughout its length, but
to a width of only 60 or 70 miles, by surprisingly great defi-
ciencies of gravity (Fig. 211). This band of what are called
“ negative anomalies of gravity ” implies
that there is a cor-
responding deficiency of density in the materials of the crust
beneath. Only one geological explanation of such a deficiency
is consistent with the observed facts the lighter layers of the
:
crust must have buckled into a great downward fold or root.

Fig. 212 illustrates the inferred structure. Obviously a crustal
404
AN EMBRYONIC MOUNTAIN CHAIN
Isostatic Anomalies
of Gravity
Fig. 212
Crustal section through Java and the adjoining floor of the Indian Ocean to
show the relation between topography, gravity anomalies (broken line with
actual determinations indicated by heavy dots), and the inferred down-buckling
of the crust. The white arrows in the lower part of the crust suggest the
directions of crustal movements thought to be initiated and maintained by
convection currents in the underlying substratum (cf. Fig. 216)
stripof this kind is far from being in isostatic equilibrium.

If were free to do so, it would be buoyed up into a high
it
mountain range and, indeed, as the islands bear witness, a long

series of upheavals has already occurred. The islands are the
firstvisible symptoms of an embryonic mountain chain. But
the uplift would be far more spectacular and continuous than
it actually is, there were not, even now, some intensely
if
powerful compressive and downsucking process at work,
restraining the tendency of the strip to rise into a position of
equilibrium. All the evidence points to the same conclusion :
the processes responsible for orogenesis are here still in opera-

tion.
Agravity survey of the Caribbean region (1928-1937)
carried out by Meinesz and a group of American collaborators
in submarines lent by the U.S. Navy, has disclosed the existence
of a very similar band of negative anomalies, extending from
<**» 405 27
) )
Fig. 213
Tectonic map of the area around the Caribbean Sea, showing the peripheral belt
of negative anomalies of gravity (dotted). Volcanoes are indicated by white
stars and dots
north of Haiti, round the volcanic arc of the Lesser Antilles, to

Trinidad and South America (Fig. 213). Had the war not
interrupted this new method of exploring the depths, many
other promising belts, such as the Asiatic island festoons, the
New Zealand-New Guinea arc, and the loop of the Southern
between South America and Antarctica, might by now
Antilles
have been investigated in this way.
The Cause of Mountain Building

Many attempts have been made to “ account for ” the
remarkable phenomena involved in mountain building, that
is, to find some mechanism in the earth competent a to pro-
(
vide forces of sufficient magnitude to compress, buckle, and

thicken the crust ; and ( b to explain the sequence of events
in each orogenic cycle, the succession of orogenic cycles during
geological time, and the distribution of the orogenic belts over
This is a tall order, and discussion of the
the face of the earth.
matter has not yet passed the speculative and controversial
406
THE THERMAL CONTRACTION HYPOTHESIS
stage. In a body like the earth gravity tends to maintain
equilibrium and stability. The only known agency capable
of disturbing this equilibrium to any important extent is heat.
Increase of temperature leads to expansion and fusion, and
decrease to consolidation and contraction. We therefore look
to thermal changes within the earth as the most promising line
of attack.
We feel reasonably sure that the earth was originally molten,
and it is therefore generally believed that the earth has attained
itspresent thermal state, at least on balance, by cooling. This

idea provides the basis of the time-honoured contraction hypo-
thesis. Once a relatively cold crust was formed, the cooling
interior tends to shrink away from it. Obliged by gravity to
settle down on the shrinking substratum, and so to fit into a
smaller area than before, the crust is inevitably thrown into a

state of compression, to which it responds by folding and
thrusting. In a similar way the skin of an apple is thrown into
wrinkles as the apple dries, and shrinks by loss of moisture.
The contraction hypothesis satisfies condition (a) qualita-
tively if not quantitatively, but is much less satisfactory in
relation to (
b ). Some of the objections are as follows :
1. One would expect the crustal wrinkles produced by
uniformly distributed compression to be in pattern rather — —

like those of a drying apple, instead of being strongly localised,
as the actual orogenic belts are. Experiments made to imitate
the process as closely as possible confirm this expectation
(Fig. 214) and also show that nothing resembling the pre-
liminary geosynclines is reproduced.
2. The cooling of the earth must have been relatively rapid

in its and have slowed down ever since. Thus
early history
the time intervals between the climax of each orogenic cycle
and the next should have become steadily longer. But the
actual intervals, though all of the same order, suggest a speeding
up rather than a slowing down (cf. page 109). The Alpine
Revolution and all the associated volcanic activity indicate that
during the latest cycle the earth has been far more vigorous
than at any earlier time for at least 1,000 million years.
407
3. It is improbable that the earth can have

highly
cooled during the last 200 million years to
sufficiently
furnish more than a small proportion of the contraction
necessary to match the crustal shoitening involved in the
Alpine Revolution.
At least in the early stages of the earth’s history cooling

would be brought about by convection in the substratum. In
this process (see Fig. G) currents of relatively hot and light
material ascend in certain places, so carrying heat up to the
Fig. 214
Network of anticlinal ridges

produced in a layer of
collodion on a stretched
circular sheet of rubber
allowed to contract in its
own plane. (See A. J. Bull,
“ The Pattern of a Contracting
Earth,” Geological Magazine,
1932, p. 73)
[A. J. Bull
base of the crust, through which some of it escapes by con-

duction. Towards the top, the currents spread out in all direc-
tions from each centre, until they encounter similar currents
from neighbouring centres and turn downwards. The de-
scending currents consist of somewhat cooled and slightly
heavier material. The driving force arises from the difference
in density between the central and marginal columns. This
kind of circulation continues until the temperature falls nearly
to the freezing point of the material concerned.
Beiore the discovery of radioactivity it was thought that
the stage of convective cooling would
not last very long, but
408
RADIOACTIVITY AND CONVECTION CURRENTS
there is now good reason to believe that it may not even yet
have ceased. The only essential condition for the maintenance
of convection in the substratum is that there should be a small
supply of heat available, to make good the heat lost through
the crust by conduction and igneous activity. That the earth
actually has such a source of internal heat became evident in
1906 when Lord Rayleigh discovered the presence of small
,
amounts of the radioactive elements in a great variety of rocks.

The radioactive elements give out heat as they distintegrate
(see page 103 ), whence it follows that heat is continuously being
generated within the rocks that contain them. Of the many
hundreds of samples that have been tested from all parts of
the world not one has been found to be free from these heat-
generating elements. The whole of the heat lost from the
earth could be supplied by one ounce of radium in every
1,000 million tons of the material of the earth. The rocks
themselves contain, on an average, fifty times as much. This
startling result indicates that the radioactive elements must be
largely concentrated in the crust —
otherwise the earth would
be much hotter than it is, and we should not be here to discuss
—
the matter but it provides no reason for supposing that the
material of the substratum can be absolutely devoid of these
elements. Even minute traces would suffice to keep convection
going. Moreover, there is the additional possibility that the
base of the substratum may receive a supply of heat from the
underlying liquid core. Thus we are free to explore the possi-
bilities that arise from the hypothesis of sub-crustal convection
currents.
Where currents are flowing horizontally along the under-
surface of the crust, they exert a powerful drag on the latter,
throwing it into tension where they diverge and into com-
pression where they converge. Thus we should expect orogenic

belts to be formed where two approaching currents turn down.
This mechanism is particularly well adapted to account for
root development and for the localised folding and thrusting
of the overlying sedimentary layers. Figs. 209 and 210 strongly
suggest that currents arose beneath Laurasia and Gondwana-
land and spread out towards their margins, where they en-
409
countered similar currents belonging to a vast Pacific system.
The evidence that roots are still being held down in the East
and West Indies indicates that locally, at least, the currents
may still be active.
The convective mechanism is not a steady process, but a
periodic one which waxes and wanes and then begins again
with a different arrangement of centres. After a particular
distribution of centres has been established, the rate of flow
for a very long period must be extremely slow (Fig. 215).
As hotter and lighter material from the base rises to become
the ascending columns, and cooler and heavier material from
the top turns downwards into the descending columns, the
—
driving force of the currents and therefore their velocity
— is increased. For a comparatively short period the currents
move relatively quickly. Towards the end of this stage the
hotter material begins to spread out at the top, while the cooler
material begins to flow along the bottom. This slows down
the currents and as hot material moves into the sinking
;
columns and cool material into the rising columns, the currents
finally come to rest. Thereafter, a new arrangement of currents
begins to develop.
The crustal effects brought about by the three stages of the
ideal convection cycle just described correspond closely with
the three stages of the simplest type of orogenic cycle :
Convection Cycle Orogenic Cycle

Stage 1 A long period of slowly A long period of geosynclinal subsidence
accelerating currents over the site of the descending currents
Stage 2 A short period of relatively A shortperiod of root formation and

rapid currents orogenic compression. Downdrag pre-
vents the root from rising into isostatic
equilibrium
Stage 3 A period of waning currents, A period of gradual uplift and restoration

bringing the cycle to an end of isostatic equilibrium
In order to study the effects on the crustal layers of con-

vection currents in the substratum Griggs has made a very
effective series of experiments with small-scale models, using
materials with properties related to the size of the model
exactly as the properties of the earth’s materials are related
410
Geosyncline
Fig. 215
Sections through the earth to illustrate the possible correlation between
the
successive stages of an orogenic cycle and those
of a hypothetical convection
current cycle
41
to the size of the earth. In one model, designed so that an
earth process requiring one million years could be reproduced
in one minute, the crust was made of a mixture of sand and
heavy oil, while the substratum consisted of very viscous water-
glass. To generate the currents rotating drums were used
(Fig. 216). When the drums are slowly rotated the crust is
gently downwarped by the descending currents (stage As 1).
the rotation speeded up, outward directed thrusts develop
is
near the surface, while the greater part of the crust is dragged
inwards and downw'ards to form a root, which is kept down by
the sinking currents (stage 2). As rotation is
slowed down and
rit;. 210
gram of Griggs
dynamic model to simulate the action of subcrustal
convection currents and the response
of the crust. The stage illustrated shows
e eseopment o a crustal downfold (root
or tectogene) with outward thrust-
near
™L n ° e rums
,
T
(C/ Fig 197 > in response to the currents set up by the
- -
substratum in the model is very viscous waterglass,

’
(
rom IJ. ^
continentaI crust ^ a mixture
Griggs :
.
of heavy oil and sand
American Journal of Science, vol.
237, 1939, p. 642)
stopped, the root exerts its

buoyant effect, and the surface is
upliited well above the original
level.
Thus in convection currents
we have found a mechanism
a c osely
matches the requirements. Actual
currents in
he substratum must, of course,
be far more complicated than
Se 0 * e 1 ea ^ c c ^ G
? y which alone has been considered here
u so are le orogenic
t
phenomena we seek to understand.
412
SUB-CRUSTAL MOVEMENTS
In particular, eddies and swirls are to be expected, and one

has only to glance at a map of the islands around the Banda
Sea (Fig. 211) to realise that their distribution may well be the
surface reflection of sub-crustal spiral-like movements.

W. H. Bucher
The Deformation of the Earth’s Crust. Princeton University Press, 1933.
L. W. Collet
The Structure of the Alps. Arnold, London.
E. B. Bailey
Tectonic Essays : Mainly Alpine. Oxford University Press, 1935.
H. A. Brouwer
The Geology of the Netherlands East Indies. Macmillan Co., New York,
!
925 -
J. H. F. Umgrove, and Ph. H. Kuenen

F. A. Vening Meinesz,
Gravity Expeditions at Sea,
1923-1932, Vol. II. Publication of the
Netherlands Geodetic Commission, Delft, 1934.
H. H. Hess
Gravity Anomalies and Island Arc Structures with Particular Reference to
the West Indies. Proceedings of the American Philosophical Society,
Vol. LXXIX., pp. 71-96, 1938.
A. Holmes
Radioactivity and Earth
Movements. Transactions of the Geological
Vol. XVIII., pp. 559-606, 1931.
Society of Glasgow,
The Thermal History of the Earth. Journal of the Washington Academy of
Sciences, Vol. XXIII., pp. 169-95, 1933.
D. Griggs
A Theory of Mountain Building. American Journal of Science, Vol.
CCXXXVIL, pp. 611-50, 1939.
J. H. F. Umgrov e
The Pulse of the Earth, Nijhoff, The Hague, 1942.
4*3
—
Chapter XIX
EARTH MOVEMENTS PLATEAUS AND RIFT

:
VALLEYS
Surface Expressions of Epeirogenic Movements
We have now to consider the crustal structures,and the result-

ing modifications of surface relief, brought about by movements
of uplift and subsidence. The framework and “ grain ” of the
continental crust are determined by the orogenic belts, but
superimposed on this primary framework there are later
epeirogenic effects due to warping, fracturing, and faulting of
the crust. Here the movements are essentially up and down.
The crust behaves somewhat like a flagged or badly cracked
pavement on a shifting foundation. Widespread swells and
sags e.g. plateaus and basins —
are produced by differential
warping on a regional scale, characteristically accompanied by
marginal and internal faulting ;
and ridges and troughs— e.g.
—
block mountains and rift valleys are produced by differential
movements of the fault-margined blocks and strips into
which the dislocated crust is shattered. The net effect of
all these epeirogenic and related movements in recent geo-
logical time has been, despite local sinkings and inbreaks,

to elevate the greater part of the continental surface well
above sea level.
Plateaus are broad uplands of considerable elevation.
Tibet and the Colorado and East African plateaus are out-
standing examples. Basins are relatively depressed regions of
roughly equidimensional outline. The term is used very
widely and is applied to all broad sags of the crust, whatever
the surface levels may be ; from sea basins, like the Black
Sea or the Celebes Sea, to mountain-rimmed plateaus which
are often, like the Great Basin of Nevada, characterized by
internaldrainage. Ideally, the drainage from a plateau
would be outwards and that of a basin inwards but as many
;
414
BLOCK FAULTING
plateaus have locally dimpled or down-broken surfaces and
many basins have drainage exits through marginal depressions
in the rims, this simple criterion is far from being of general
application. The term “ basin ” is also given to ancient crustal
Little Kharas Great Kharas

N.W. Mountains Mountains S.E.
Uncontormily _
OWYKA Senes-,
UnconformilY-
nama System
Granites and Schists
of Basement Complex
Fig. 217
Diagrammatic section (with minor modifications due to denudation and deposi-
tion omitted) to show the faulted structure of the Kharas Mountains, south-west
of the Kalahari Desert, South-West Africa. Length of section about 60 miles.
{After C. M. Schwellnus)
sags which have been filled with sediments and in some cases,
as in Africa, subsequently uplifted into plateaus (Fig. 223).
Regions which have been divided by faulting into relatively
elevated or depressed blocks are said to be block faulted. The
upstanding fault blocks, which may be small plateaus or long
ridge-like block mountains, are called horsts (Fig. 217). The
Hercynian massifs of Europe, such as the Vosges and the
Section across the Rhine Rift Valley, north of Mulhausen, showing the structure
of the rift valley strata as determined by numerous borings. Miocene sediments
appear here and there below the alluvium, but most of the beds represented are
of Oligocene age, resting on Jurassic, as shown. Length of section about 30 miles
Black Forest (Fig. 218) and the Harz Mountains, are horsts.
Blocks which have been tilted, like many of the Great Basin
ranges (Fig. 221), are sometimes distinguished as tilt blocks.
The North Pennine region, as shown in Fig. 39, is an up-
4*5
EARTH MOVEMENTS : PLATEAUS AND RIFT VALLEYS
tilted block sloping gently down towards the North Sea. Fault
blocks depressed below their surroundings form minor
basins
or fault troughs. A
long fault trough, forming a tectonic valley
bordered by parallel fault scarps, is known as a rift valley.
Forest the
Between the horsts of the Vosges and the Black
Rhine flows through a rift valley (the Rheingraben —Fig. 218).
The river occupies the valley, but did not excavate The
it.
most renowned system of rift valleys, however, is that which

traverses the East African plateaus from the Zambesi to the
Red Sea and beyond (Figs. 223 and 226-28).

In the course of time the surface relief which results from
epeirogenic movements becomes greatly modified by denuda-
tion and deposition, and locally by volcanic activity. Never-
theless, as illustrated by the examples already mentioned,
the
face of the earth of to-day is diversified by many new or
boldly preserved topographic features which are primarily
due to vertical movements. The reason for this is that such
movements have been unusually active during Tertiary and
late geological time, right down to the present day. Deep
gorges are being cut by rivers in high plateaus fault scarps ;
are being gradually worn back by the sculpturing hand of

erosion block mountains are being carved into hills and
;
valleys, and sediments are accumulating in basins and troughs ;
but long ages must elapse before the broad outlines of the
topography of these regions cease to reflect the latest epeiro-
genic dislocations of the crust.
The irregular surface of a fractured pavement is due mainly
to the varying subsidence of different parts of a poorly laid
foundation. Influenced by the idea of a contracting earth,
Suess considered the verticalmovements of the crust to be
also downwards, “ sunken ” and “ uplifted
essentially
regions being simply those which had subsided to a greater
or less extent than their surroundings. Horsts, according to
Suess, are merely blocks behind in the general down-
left
settling of the crust on the shrinking interior. This conception
of the underlying mechanism is no longer acceptable. There
isno justification for making any assumption as to whether
the up-and-down movements carry the surface of the region
416
FLUCTUATIONS OF SEA LEVEL
affected away from the centre of the globe or towards it.
Some standard of reference there must be, however, and for

general purposes sea level is the most appropriate one, whenever
it can be appealed to. In dealing with movement along a
fault plane such reference may not be possible. The block
on one side is said to be downthrown or upthrown relative to
the other. Here, each of the adjoining blocks serves as the
standard of reference for the other. In the same way a basin
or rift valley is naturally described as a region of depression
relative to an adjoining plateau or range. The floors of some
basins and rift valleys have indeed subsided below sea level
(Fig. 225), while in other cases the “ depressed ” tract may

have been actually uplifted, but less so than its surroundings.
Fluctuations of Sea Level
Although sea level is the nearest approach we have to an

absolute standard of reference, it must not be overlooked that
the level of the sea itself fluctuates from time to time. If the
frozen water locked up in the Greenland ice sheet, and
now
in the far greaterone of Antarctica, were to be melted in
consequence of a change of climate, the volume of the
oceans would be increased and the sea level (allowing for
isostatic reactions) would rise by about 100 feet. All the
lowlands of the world would be inundated up to this height,
and to envisage the resultant consequences to humanity
baffles the imagination.
Yet mankind has already lived through even greater
changes. During the Pleistocene ice ages, when vast ice sheets
covered immense areas of Europe and North America, the
level of the sea was about 300 feet lower than it is now (again
allowing for isostatic reactions). Broad stretches of land then
extended beyond the present shores of all the maritime countries
which lay outside the blanketing ice fields. The bottom of the
Persian Gulf, for example, must then have been a fertile plain,
floored with alluvium from the united waters of the Tigris and
Euphrates. This vanished land was, in all probability, the
4G
—
home of the ancestors of the Sumerian people who immigrated

into Chaldea thousands of years ago. We
can easily under-
stand why the migration was inevitable. As the last of the
former ice sheets began to melt away the sea gradually rose and
occupied the Gulf, driving the dispossessed population into the
higher ground of Mesopotamia, where the remains of their
settlements, long buried by river silt and wind-blown sand, have
been disinterred by archaeologists.
World-wide changes of sea level, whether due to the growth
and decay of ice sheets, to displacement of water by accumu-
lating sediments, or to other causes, are described as eustatic
changes. In general these are always more or less involved
with slow crustal movements due to isostatic readjustments.
The removal of an ice sheet, for example, reduces the mass
of the crustal column beneath the glaciated area, and in re-
sponse the region is slowly heaved up until isostatic equilibrium
is restored. At the same time, but still more slowly, the ocean
floor, now loaded with the weight of the restored water,
responds by sinking a little. Moreover, in certain places
independent earth movements may be simultaneously affecting
the level of the crust. For these reasons the recent changes
of level indicated by land emergence e.g. raised beaches or —
— —
land submergence e.g. submerged forests can only in part
be referred to eustatic changes. The actual displacement at
any particular locality is the algebraic sum of the changes of
level due (a) to the recent eustatic changes of sea level ;
(
the degree of isostatic readjustment so far accomplished
b ) to
at that place ; and quite commonly ( c ) to independent earth
movements which in some places are still going on.
When the relative movement between land and sea is a
few hundred feet or less, it is often difficult or even impossible
to disentangle the separate effects of these three factors. In
dealing with the major effects of epeirogenic movements,
however, this difficulty does not arise, for it commonly happens
that the changes of level involved are measurable in many
thousands of feet. Moreover, wherever faulting or tilting has
taken place it is clear that earth movements have operated,
since eustatic changes of level are everywhere uniform.
418
RISE AND FALL OF MOUNTAIN RANGES
Dislocations of Orogenic Belts
As we have already seen, the compression of an orogenic

belt normally followed by uplift. This first upheaval of
is
the mountains can be reasonably interpreted as a direct effect

of isostatic readjustment, due to the buoyancy of the root.
But the root is sometimes quite short-lived. If its heated
material softens and spreads out, the mountains subside again
over the weak and unstable foundations. Thereafter, the
history’ of the folded tracts and the adjoining areas is one of
continued epeirogenic movements of such variety and com-
plexity that no adequate explanation is yet forthcoming. In a
general way subsequent uplifts can perhaps be referred to
thickening by compression of the weaker parts of the softened
foundations, accompanied by widespread uparching and
cracking of the overlying layers of the crust. But deep sub-
sidences also occur, and these present a problem for which no
satisfactory solution has yet been offered.
The Rockies and the Andes are examples of ranges that
illustrate some of these remarkable vicissitudes. They reached
their present elevations by comparatively recent uplifts which
took place only after the first mountains had already been
reduced to low-lying plains. In the high elongated plateaus
of the Andes the uplifted plain, now surmounted by great
volcanoes, can easily be recognized. Moreover, it is clear that
parts of the plain must have been submerged below sea level
before the uplift took place, for unfolded Pliocene marine beds
are still preserved at heights up to 6,000 feet.
Turning to Europe (Fig. 200) it is apparent that the
Hercynian belt has been severely faulted and broken up into
isolated upstanding blocks with intervening depressed areas.
South-west Ireland is separated from Cornwall and
Devon by
the western approaches of the Atlantic. Between the horsts
of the Vosges and the Black Forest lies the Rhine rift valley.
Moreover, as we have seen, much of the broad Hercynian
belt subsided at a very early stage to form the floor of the
Alpine geosyncline.
Even the Alpine ranges have had their ups and downs.
4i9
The iEgean archipelago provides a good example of the

fragmentation and partial collapse of a mountain range.
Before the Pliocene Greece was directly connected with Crete
and Asia Minor. This is proved by the widespread occurrence
in the islands and surrounding countries of thick deposits of
fresh-water marls crowded with the fossil shells of snails.
Lying on these lake deposits there are raised beaches and
marine strata of Pleistocene age which indicate subsidence at
the time of the first advance of the Western Mediterranean
into the fEgean area. At this time also volcanoes broke out,
some of which are still active. As a result of more recent
faulting and localized uplifts the lake deposits and raised
beaches now stand at various levels up to 3,000 feet in the
islands and over 5,000 feet near the Gulf of Corinth. Along
the shore lines these beds are abruptly cut off, without change
of thickness, showing that they must continue beneath the sea.
The islands are horsts and between them the sea occupies the
intervening regions of collapse. The prevalence of severe
earthquakes indicates that the region is still very unstable.
The Black Sea is another region of quite recent subsidence,
due in this case to the inbreak of roughly circular areas within
a series of arcuate faults. Samples brought up from the floor
of the western basin at a depth of over 6,000 feet prove to be
Pliocene land deposits. Nearer the present shores there are
recent marsh deposits which are now submerged to depths of
at least 300 feet. These must have gone down within the last
few thousand years.
The Cordilleran Plateaus of North America

The Cordillera of North America includes all the ranges
and plateaus west of the Great Plains and of the coastal low-
lands of the Gulf of Mexico. The general structure of this
long and complex highland region is shown in plan by Fig. 201
and in section by Fig. 219. The Rocky Mountains and their
Alaskan and Mexican continuations (folded at the end of the
Cretaceous) rise boldly all along the eastern margin (Plate 42 b).
420
[(j. JL Grunt, i .S. Xational Park Semce
(B) Aerial view of the Grand Canyon of the Colorado, looking o\ cr the Painted
Desert and Marble Canyon to the Vermilion cliffs and Paria Plateau beyond
>*
'Darien Leigh, Ltd.

(B) Salt marsh ( yalina i in Death Valley. California isee Fig. 184)
CORDILLERAN PLATEAUS
To the west the plateaus are bordered by the Cascade and

Sierra Nevada ranges (folded at the close of the Jurassic).
Still farther west the Coast Ranges (folded in the late Tertiary
and Pleistocene) margin the Pacific.

In the United States the Rocky Mountains represent only
one side of the orogenic belt to which they belong. The other
side swings out to the west, where it has been transformed by
erosion and block faulting into the ranges and depressions of
the Great Basin. Farther north a similar tract, buried under
a thick cover of plateau basalts, has become the Columbia
Plateau. Between the Great Basin and the Southern Rockies
lies the unfolded median area of the Colorado Plateau. South
of this median area the two sides of the Rocky Mountain
orogenic belt unite, the two fronts being represented by the
Coast Great
Rocky M r-S p
W. Ranges Sierra Basin Wasatch Co'orado Park Front
California Nevada M Cs Plateau Range Range
Pacific
Fig. 219
Schematic section across the Cordillera of the western United States. Length
ot section about 1,900 miles
eastern and western Sierra Madre. Between these mountain

ranges there is no unfolded median area, but the intermontane
region has nevertheless developed into the Mexican Plateau.
The tectonic history of this plateau is thus somewhat like that
of the Great Basin, but it differs in having widespread lava
tracts on which, in the south, a series of lofty volcanic cones
(including Popocatepetl, 17,520 feet) have been built. In
Canada and Alaska the intermontane plateaus, though more
constricted, extend through British Columbia to the Yukon.
The Colorado Plateau is distinguished by its great height
and by the fact that it is built of horizontal or gently un-
dulating sediments. The region was first uplifted with its
bordering mountains at the end of the Cretaceous. Deep
erosion and accumulation of lake and alluvial deposits occurred
during the Eocene. A
second uplift, with much dislocation by
faulting, took place in Middle Tertiary times, after which the
(39«) 4.2 1 28
plateau was again worn down to a plain. The latest uplift

has raised the surface to heights of 7,000 to 10,000 feet since
the beginning of the Pleistocene.
The Colorado River and its tributaries have cut deep
canyons through the Plateau (Plates 1 and 84b). Northwards
from the Grand Canyon (see p. 199 for description) the surface
rises by successive cliffs and terraces to over 11,000 feet in the
high plateau of Utah, the edge of which has been sculptured
by erosion into such fantastic landscapes as those of Bryce and
Zion canyons (Plate 32a). West of Bryce canyon are the
famous laccoliths which are responsible for the updomed hills
of the Henry Mountains. To the south of these, on the other
side of the Colorado River, is Monument Valley, so called
because of its obelisks and towers and other castellated erosion
remnants, carved out of red Triassic rocks (Plate 64b). South
of the Grand Canyon a considerable area is covered by lava
flows and dotted with hundreds of small volcanic cones, now
extinct, but so perfectly preserved that they are obviously
very youthful features. Not far away from the volcanic area
is Meteor Crater, a great depression 4,000 feet across and
500 feet deep, caused by the explosion which followed the
impact of a giant meteorite that fell from the sky. This brief
outline gives but a faint idea of the variety and interest of the
magnificent scenic and geological features for which the region
is so justly renowned.
Bordering the Plateau on the north, the Uinta Mountains

extend at right angles to the Wasatch Range for 160 miles
along the Utah side of the boundary with Wyoming. They
provide a classic example of subsidence and uplift on a
stupendous scale. As shown in Fig. 220, the structure is a
broad open anticline, 45 miles across, cut by steep east-and-
west faults. The sedimentary rocks are all shallow water
types which were deposited, in a deeply subsiding part of the
Rocky Mountain geosyncline, to a total thickness of 35,000
feet. Since the Pre-Cambrian foundation of this immense
pile is now exposed at a height of 10,000 feet above sea level,
it follows that the subsequent uplift has amounted to 45,000
feet. Late Cretaceous uparching of the sediments raised the

422
THE GREAT BASIN
crest to about 10,000 feet. During and after this elevation
the surface was actively eroded, and Eocene sediments were
deposited in the depressions bordering the great arch. The
main faulting began at the end of the Eocene and the range
was uplifted 25,000 feet, the Eocene deposits being dragged
up and strongly tilted at the same time. Erosion continued
to lower the surface. Finally the range received a further
elevation of 10,000 feet by sharing in the regional uplift which
raised the Plateaus during the last million years or so.
The Great Basin is, for the most part, an area of internal
drainage made up of more than a hundred undrained troughs
and basins lying between long block mountains which trend
approximately north and south. In the extreme south west
Section across the Uinta Mountains, Utah. C= Cambrian ;

T = Triassic
of the area the drainage escapes into the Colorado River at a

point which is now the site of the great Boulder Dam. The
upstanding blocks, known as the Basin Ranges, rise to heights
of 7,000 to 10,000 feet (Plate 84a). The bordering ranges of
the Sierra Nevada and the Wasatch Mountains are gigantic
tilted blocks of the same type, with their uplifted scarps facing
inwards in each case. The floor levels between the ranges
rarely exceed 5,000 feet but are usually above 3,000. There
are two exceptionally low down-faulted troughs along the
western margin. North of the mouth of the Colorado River
at the head of the Gulf of California —
itself a great depression
— is the Salton Sink, 273 feet below sea level. The floor of
Death Valley, 200 miles to the north, is just a few feet lower
(Plate 85b).
Like the neighbouring areas, the Great Basin region was
423
EARTH MOVEMENTS I PLATEAUS AND RIFT VALLEYS
reduced to a plain during Tertiary times. The movements

and most of the differential faulting which have determined
the present topography are post-Tertiary. Along some of the
faults there are lines of youthful volcanic cones and, locally,
thin sheets of lava have been faulted into strips which now
stand at very different levels. Most of the Basin ranges are tilt
blocks with steep fault scarps on one side and gentler back
slopes on the other (Fig. 221). The older scarps have been
considerably eroded into ravines and spurs and the lower
slopes are often aproned with screes and fans of rock-waste.
Both these and the spurs are truncated by triangular facets
along the lower slopes of some of the ranges, showing that
Fig. 221
Diagram to illustrate the fault-block structure of the “ basin ” ranges of the
Great Basin, Utah. (After W. M. Davis)
quite recent fault movements have occurred. The severe

earthquake of Owens Valley in 1872 was due to faulting of this
kind at the foot of the Sierra Nevada.
The basins and troughs are often coated or levelled up
with sediment from the ranges. Some are barren desert wastes,
others support scanty vegetation, and some contain lakes,
mostly temporary. If the basins filled with water until over-
flow from one to another became possible, an external drainage
could be established. But the climate is arid, as the region
falls within the rain shadow of the Sierra Nevada. The supply
of water is limited and the rate of loss by evaporation is high.
Lakes are therefore shallow. Most of them are saline, and
surrounded by mud flats known as playas or by salt-encrusted
flats called salinas (Plate 85b). Great Salt Lake, near the
424
STRUCTURE OF ASIA
Wasatch Range, is the salt-saturated relic of a much larger

fresh-water lake, known as Lake Bonneville, which during the
Ice Age expanded to a height of about 1,000 feet above the
present level and found an exit northwards into the Snake
River. The terraced shore lines of this ancestral lake (cf.
Plate 63b)form conspicuous horizontal features along the

slopes of the many ranges which formed its margins or stood
as islands above its surface.
The Plateaus and Basins of Central Asia
The surface of Asia has been compared to a bas-relief with

a raised pattern of folded and block-faulted mountain ranges
on a background of plateaus and basins and marginal plains
(Fig. 222). The ranges
represent the compressed sediments
of former geosynclines. Although the elevation of the present
ranges is due mainly to Tertiary faulting and uplift so that —
most of the ranges north and north-east of the Himalayas are
—
block mountains their orogenic folding is far older, except
in the south. The Caledonian orogeny was responsible for the
folded arcs around Angaraland and for the eastern part of
the Altai Mountains. The Hercynian orogeny is well repre-
sented in the western Altai and the Tien Shan, and in the
bordering ranges north of Tibet, and, less markedly, in their
eastern continuations. The folding of the latter, from the
Tsinling Shan to Japan was completed in middle Mesozoic
times, forming a central mountain belt which now divides
China into two natural regions the desert and loess-covered
:
country of the north and the green and tree-clad lands of the
more genial south. The folds of the Himalayas, Burma, and
the Dutch East Indies, and some of the Pacific island festoons
were added by Tertiary mountain building. Thus the inner
ranges are the oldest and the marginal arcs of the south and
east are the youngest, with the “ negative belt ” surrounding
the East Indies (Fig. 211) still in process of development.
Asia thus appears to have grown by successive additions
to the Angaraland nucleus. The mountain ranges welded
425
Fig. 222
Tectonic map main trend-lines of successive orogenic belts
of Asia, showing the
and the intermontane plateaus and basins. The older and more stable regions
of Angaraland and India are indicated by horizontal shading
together the formerly isolated areas which are now the inter-
montane plateaus and basins. It follows that the latter repre-
sent the regions that furnished sediments to the geosynclines
out of which the mountain rims emerged. With the exception
of at least the southern half of Tibet they are not median areas,
like the Colorado plateau, but are essentially minor shields
or forelands, underlain by very ancient rocks such as the
granites and gneisses of Mongolia. The depressions of their
warped and faulted surfaces are, of course, veneered with
426
ASIATIC PLATEAUS AND BASINS
desert or lake sediments derived from the erosion of their

mountainous frames.
Tibetis the loftiest of the world’s plateaus, with an average
height of 16,000 feet. Nevertheless the southern part lay

beneath the Tethys as recently as Cretaceous times. The
northern part was already land during the Jurassic. During
the Tertiary orogenic cycle the whole region was wedged up,
as indicated diagramatically by Fig. 199. The Altyn Tagh
and Kwen Lun, although folded long before the Himalayas,
were then thrust upwards and outwards towards the Asiatic
foreland. Tibet is thus a sort of composite median area.
Beyond the steep scarps of the Kwen Lun lies the much lower
Tsaidam basin (8,000 feet), a desolate waste of stony flats and
salt marshes in which the Hwang Ho has its source. In
the east and south-east of Tibet several ranges rise above
the plateau and swing round towards the south, forming
the almost impenetrable barrier of the Great Snow Moun-
tains. Here the Salween, Mekong, and Yangtze Rivers,
which rise on the plateau, flow through closely parallel
gorges which are amongst the grimmest and least accessible
in the world.
To the east of the mountain barrier the country descends
by giant faulted steps into the Red Basin of Szechuan (1,500-3,000
feet), so called because of the prevalence of brick-red sandstones
which were deposited in Tertiary and Jurassic lakes. With its
humid climate, productive soil, and mineral resources, this
region is the most prosperous province of China. The capital,
Chungking, is situated in the south, on the banks of the
Yangtze.
North of Tibet the Altyn Tagh drops steeply into the dry
and wind-swept Tarim Basin (2,000-6,000 feet). The fault
scarps face inwards, as also do the precipitous walls of the
Tien Shan on the north. The basin is floored with bare gravel
and belts of sand dunes. The sluggish and intermittent
drainage terminates in the shifting salt marshes of Lop Nor.
Formerly the lower parts of the basin were occupied by a vast
lake, the shore terraces of which, originally flat, have recently
been tilted so that the western end is now 1,000 feet above the
427
eastern. A subsidiary basin to the north-east encloses the

depression of Turfan, 720 feet below sea level.
The Tarim Basin is practically open towards the east, rising
by low steps into the Gobi Desert of the Mongolian Plateau
(3,000-5,000 feet). This vast and largely desert tableland is
extraordinarily flat and monotonous. The low ranges of
block mountains which traverse it have all been worn down
to low relief. On the east the plateau drops abruptly to the
Manchurian plains along the steep fault scarp of the Great
Khingan.
To the north, Mongolia is bordered by the mountains
around Angaraland, within which lies the rift valley occupied
by Lake Baikal. The lake floor descends 3,939 feet from a
water level of 462 feet, indicating that the bottom of the rift
valley is at least 3,477 feet below sea level. Lake Baikal is
the world’s second deepest lake, and it occupies the deepest
known continental inbreak (Fig. 225).
The Plateaus and Basins of Africa
Africa, to which Arabia must be added from a structural

point of view, is a vast continental shield, margined with
folded mountains only in the extreme north and south. Apart

from limited marine invasions across the coastal plains and
more extensive incursions of the sea across Abyssinia and parts
of the Sahara, Africa has been a land area since the Pre-
Cambrian. Throughout this long period the movements of the
shield have been persistently epeirogenic, giving a structural
pattern of broad basins separated by irregular swells which
rise towards the east to form a coalescing series of plateaus,
the latter being traversed by a spectacular system of rift valleys
(Fig. 223). The plateaus and swells have been intermittently
uplifted and denuded, with the result that they now consist
of old rocks which were formerly very deep seated. The basins
have been the receptacle of thick deposits of continental sedi-
ments representing the material eroded from the uplifted
tracts.
428
Fig. 223
Map showing in a generalized way the tectonic basins, plateaus, swells, and
rift valleys of Africa
Some of the basins are arid regions of internal drainage.

The Chad basin one of the most remarkable. Lake Chad,
is
fed by the Shari River from the swell to the south, is a shallow
expanse of swamps and open water with no visible outlet.
Yet the water is not stagnant, and although evaporation from
its surface is high it does not become brackish. Despite appear-
ances, Chad is not a terminal lake, for it drains underground
and feeds the oases of the Borku lowlands 450 miles to the
north-east. El Juf, a vast desert depression north of Timbuktu,
is one of the most awesome and least known parts of the Sahara.
On the other side of the Equator lies the Kalahari basin, partly
429
grassy steppes and partly desert, with internal drainage in the

north towards the brackish swamps of “ Lake ” Ngami.
In striking contrast is the equatorial Congo basin, copiously
watered by the great river system after which it is named. The
whole region is underlain by thick continental sediments of
late Carboniferous to Jurassic age (collectively known in Africa
as the Karroo system). These everywhere dip gently inwards
from the surrounding swells, as a result of warping which
occurred in Tertiary times. The basin as a whole is slightly
tilted towards the south-east. Where the Congo meets the
swell on the coastal side it forms Stanley Pool, whence it escapes
across the obstruction to the Atlantic by way of a series of
cataracts. The basin of the Anglo-Egyptian Sudan is less easily
traversed by the Nile. In the southern part of the basin the
White Nile follows a sluggish and tortuous course through a
wide expanse of papyrus swamps and lagoons. Thick floating
accumulations of vegetable remains, known as the Sudd
(Plate 75a), obstruct the interlacing channels. These hot
reedy swamps are the relic of a former lake, the extent of which
is marked by vast spreads of sediment. The level rose until
the waters overflowed through a notch in the northern rim,
to become the vigorous Nile of the six cataracts between
Khartoum and Aswan.
To the east, from the head of the Red Sea to the Cape,
Africa has been uplifted into a series of plateaus interrupted
by occasional minor downwarps. The basin of the Karroo
in the south — the type area of the Karroo system — rises across
the High Veld and the Basuto Highlands to the basalt-capped
crest of theDrakensberg escarpment. The “ Great Escarp-
ment ” borders the Interior Plateau of South Africa right
round to the west, but nowhere developed
else is it so strikingly
as in the precipitous walls of theNatal Drakensberg, in places
6,000 feet high (Plate 87a). It should be noted, however, that
this escarpment is an erosion feature (page 179), and is not a
fault scarp. To the north is the much older basin of the Rand
goldfield, with Johannesburg on its northern rim, where the
gold-bearing strata dip southwards and are accessible to
mining until they sink to depths where the rocks are too hot
43°
EASTERN PLATEAUS OF AFRICA
to be worked. North of Pretoria is another ancient basin
the great Bushveld igneous complex, illustrated in Fig. 42.
Then comes Limpopo River, succeeded on
the depression of the
the north by the Rhodesian plateaus, drained by the Zambesi,
and deeply trenched on the north-east by the Nyasa rift valley.
Farther north is the oval-shaped Central Plateau of Tan-
ENTEBBE/ "'LAKE VICTORIA-

'
3720
Fig. 224
Map of Lake Kyoga, Uganda, to illustrate the effect of back-tilting of the plateau
east of the Western Rift Valley
ganyika, Kenya, and Uganda. This region was reduced to a

peneplain in Miocene times, but it has since been uplifted some
4,000 to 5,000 feet. Even the Central Plateau exemplifies the
African habit of basin and swell. Lake Victoria occupies a
recent crustal sag (Fig. 229). The lake itself has gently
shelving shores, and nowhere exceeds 270 feet in depth,
despiteits enormous area. The bordering swells are deeply
trenched by arcuate series of rift valleys. A significant feature
of the plateau is the characteristic rise of the surface towards

the edges of the rifts. The upwarping on the west is so recent
that rivers which drained westwards during the Pleistocene
are reversed, so that they flow inwards and drain either
now
into Lake Victoria or into the Victoria Nile (Fig. 224). The
curious shape of Lake Kyoga and its swampy margins, filled
with papyrus and sudd, is a clear indication that the Kafu
River and its tributaries have been ponded back towards a
subsidiary dimple north of Lake Victoria.
The Central Plateau is divided from the Abyssinian High-
lands by the relatively depressed and block-faulted tract of
plateau lavas and volcanoes in which the rift valley of Lake
Rudolf is sunk. The lavas continue into Abyssinia, where they
surmount Mesozoic marine sediments, all lying horizontally,
though now uplifted through many thousands of feet. The
Abyssinian rift valleys branch from Lake Rudolf and emerge
into the desert plains of Afar, a region which is structurally
part of the Red Sea depression, though cut off from the sea
itself by a long volcanic barrier (Fig. 227).
The African Rift Valleys

The term rift valley was introduced by Gregory for the
“ Great Rift Valley ” of East Africa, which he was the first
to recognize as a tectonic feature due to faulting. Gregory
defined the term to mean a long strip of country let down
—
between normal faults or a parallel series of step faults as if —
a fractured arch had been pulled apart by tension so that the
Fio. 225
Sections through rift valley lakes to illustrate local depression of the floor
below sea level. Each section is 50 miles across
43 2
RIFT VALLEYS
Fig. 226
Map of the Rift Valley system of Africa from the Zambesi to Abyssinia,
showing the southern (Nyasa) section and its bifurcation into the Eastern and
Western Rifts
keystone dropped in en bloc or in strips. The floors of some

rift valleys (Fig. 225) have obviously subsided, but in many
cases it is equally clear that they have merely lagged behind
the surface of the adjoining plateaus in the course of a general

uplift. Moreover, there is growing evidence that the orthodox
view which associates the inbreaks with tension is far from
being of general application. The term rift valley is therefore
used here without implications as to the mode of origin.
433
Fig. 227
Map of the Abyssinian section of the Eastern Rift Valley of Africa
The rift valleys do not constitute one long continuous

trough with a curving branch to the west, but nevertheless
they are all parts of a single system which extends from Syria
to the Zambesi, a distance of about 3,000 miles. From south
to north the system falls naturally into the following parts
(Figs. 226-28) (a) the Nyasa section and its bifurcations
:
;
(b) the Western Rift from Lake Tanganyika to Lake Albert ;
(c) the Eastern or Gregory Rift east of Lake Victoria (d) Lake ;
Rudolf and the Abyssinian section and (e) the Aqaba-Dead

;
Sea—Jordan valley section (Plate 85a), with a branch up the

Gulf of Suez. Between ( d) and (e) the Red Sea intervenes.
.
434
—
VOLCANIC ACTIVITY AND RIFT VALLEYS
Some of the scarps are old and deeply dissected by erosion,

but the most recent ones are steep and sharply defined. In the
Eastern Rift lavas preceded and intervened between the suc-
cessive movements and consequently the older flows, which
cover immense areas in
Kenya, are themselves
faulted. Most of the
rift faults follow direc-
from N.N.E.
tions not far
or N.N.W. Trends ap-
proximating to N. and S.
are less common, and
are most characteristic in
the extreme north and
south.
The association of
volcanic activity with the
rift valleys is too plain to
be missed, but it is tan-
talisingly In
variable.
some have
sections lavas
been erupted on an im-
mense scale, in some
only sporadically, but in
others not at all. It is
Map ° f the rift a ’ le >' s and associated faults
curious that some of the J
north ot the Red bea
deepest troughs, like that
of Lake Tanganyika, show no signs of vulcanism. In two of
the volcanic regions the alignment of great cones e.g. north
of Lake Kivu across the Western Rift (Plate 87b), and from
Kilimanjaro (Plate 45) and Meru (Plate 88) to the Giant
—
Craters in the Eastern Rift suggests that they lie over E.— W.
fissures. But no generalization as to distribution is possible.
The volcanic mountains of Kilimanjaro and Kenya rise from
the plateau well outside the Eastern Rift, while Elgon and its
neighbours are situated well on the inner side not far from
Lake Kyoga.
There is a remarkable and significant uniformity in the
435
' —
EARTH MOVEMENTS PLATEAUS AND RIFT VALLEYS
widths of the rift valleys, as the following measurements (in
kilometres) indicate :
Lake Albert
Lake Tanganyika (north)
.... 35-45
50
Dead Sea 35
Gulf of Aqaba 50
.
Lake Tanganyika (south) 40 Lake Rudolf 55

Rukwa 55-7 0 Lake Nyasa . 40-60
Rift valleys in other continents also have similar widths e.g.

the Midland Valley of Scotland (Fig. 180), 70-80 the Rhine
;
riftvalley (Fig. 218), 35-45 ; and Lake Baikal (Fig. 225),

30-70. The Red Sea, with a width of 200-300 km., is obviously
a depression of a different order.
Another about the African rift valleys is
significant fact
that the opposing walls and plateaus are not generally of the
same height (see Figs. 225 and 229). In some cases, indeed,
the depression is of the trap-door type, with a high wall on one
N.W.
Lake Lake or
Albert Lake Victoria . Natron
2030 ""'3720
1 *2060'
Sea®lufl > Level
Fig. 229
Section across Lake Victoria and the Western and Eastern Rifts
side and little or no faulting on the other. The floor levels

are also highly variable. The Western Rift block varies from
2,150 feet below sea level in the Tanganyika trough to 5,000
feetabove sea level south of Lake Kivu, dropping again to less
than 2,000 feet in Lake Albert. Lake Kivu occupies a high
level valleywhich would drain to the north, and so to the Nile,
but for the fact that the volcanic field of Birunga forms a
gigantic lava barrier which has dammed back its waters
(Fig. 226). The lake overflows to the south where the river
has cut a deep gorge through an older and lower volcanic
barrier. It thus drains into Lake Tanganyika, which in turn
overflows to feed one of the headwaters of the Congo. An
astonishing and unique feature of the Western Rift is the
towering horst of Ruwenzori, which rises from within the
encircling rift valley to snow-clad peaks up to 16,794 feet—
more than 10,000 feet above the general level of the plateau.
43 6
[L.X.A.
(B) Salt deposits of Lake Magadi
EASTERN RIFT VALLEY, KENYA COLONY

PLATE 87
'
.S. African Air l'oice
(A) Aerial view of the Great Escarpment and Interior Plateau, Drakensberg,
Natal
[C. van Rut c

Muhavura Mgaliinga Sabinyo Mikeuo
(B) Extinct volcanoes of Bufurnbiro, looking south-west from the plateau,
Western Rift Valley. Uganda
RUWENZORI
Fig. 230
Map of part of the Western Rift Valley, showing the volcanic areas (black) and
the horst of Ruwenzori, a section across which is inset
Ruwenzori by far the highest non-volcanic mountain in

is
Africa. an upwedged block of ancient metamorphic rocks

It is
like those of the plateau. Towards Lake Albert the block

narrows into a long “ nose ” flanked by fault scarps. Here
recent uplift is proved by the occurrence of a raised terrace of
alluvium, which was originally part of the Semliki valley floor.
(396)
43 7 29
The Eastern Rift a region of internal drainage with

is
several independent basins within it due to irregularities of the
floor level (Plate 86a). The latter varies from about 3,000
feet in the south to 2,000 at Lakes Natron and Magadi, 6,000
at Lake Naivasha, and 3,000 at Lake Baringo in the north.
Lakes Natron and Magadi, and the glaring w'hite salt-encrusted
plains around them (Plate 86b), contain vast reserves of soda
(Na 2 C0 3 ) and other chemical deposits which are exploited
commercially on a considerable scale. The salts are derived
from the soda-rich volcanic rocks of the district, partly from
the evaporation of the surface waters w'hich drain into the two
depressions, but mainly from hot springs. In 1917 the neigh-
bouring volcano of Oldonyo l’Engai actually erupted soda-
rich vapours as well as lavas and ashes. A pall of grey volcanic
ash permeated with soda settled over a large area, and with
the first rains the water holes became fouled with the bitter
salts. Many herds of cattle died through drinking from the
contaminated pools. The lakes are merely evaporating pans
in which the soluble volcanic salts are concentrated.
The Origin of the Rift Valleys

More than a century ago the Rhine rift valley was com-
pared with the fallen keystone of an arch. Thus arose the
traditional tension hypothesis, according to which a long
w edge-shaped block, tapering downwards, sank between
r
normal boundary faults as the sides were pulled apart. The

analogy is a misleading one, however, because a keystone has
an empty space into which it can drop until the inward-sloping
sides again find lateral support. A rift-valley w'edge could not
sink unless its weight displaced molten material in depth.
Volcanoes would then break out along the cracks. It has
long been thought this sort of process may have operated on
an incidental scale, but that it could not be an essential part
of the mechanism, since, as we have seen, some of the deepest
troughs show no signs of vulcanism. However, it has recently
been shown by geological work in the Eastern Rift that in
438
THE COMPRESSION HYPOTHESIS
several areas volcanic activity ceased altogether while the rift
movements were actively in progress. Such interruptions
strongly suggest that the magmatic channels through the crust
were then tightly closed by the movements, not widened and
further opened as they would necessarily have been according
to the tension hypothesis. Moreover, lateral tension fails
to account for the fact that the boundary faults are commonly
about 35-50 km. apart. The uplift of Ruwenzori is also
inconsistent with the tension hypothesis.
These do not arise on the alternative hypothesis
difficulties
—proposed by Wayland for the Lake Albert and Ruwenzori
RIFT VALLEY
Fig. 2.31
Diagrammatic section across Lake Albert by E. J. Wayland to illustrate his
hypothesis of the origin of rift valleys by compression
sections and by Bailey Dead Sea section accord-

Willis for the —
ing to which rift valleys are produced by deep-seated com-
pression. The boundary faults are then regarded as steep
upthrusts and the rift blocks as wedges (widening in depth)
held down by pressure from the upriding sides (Fig. 231). It
is not to be expected that the thrust faults should be generally
visible. As the plateau block rides up a high-angle thrust

plane, forward edge is necessarily left unsupported. Long
its
strips of the overhanging sides slump down, and the visible

walls thus appear to be normal faults, often arranged in suc-
cessive steps. In a few places, where ravines cut through the
fault scarps at right angles, subsidiary thrust faults have been
439
—
observed beneath the edge of the plateau. Such evidence is

not conclusive, however, as these thrusts may have been there
long before the rift valleys were formed. More convincing
evidence of the operation of compression is provided by the
folding of the sediments on the floor of the Lake Albert rift
valley (as indicated in Fig. 231).
The results of a gravity survey by Bullard in 1933-34
have thrown much new light on the origin of rift valleys.
The plateaus are found to be very nearly in isostatic
equilibrium, but over the rift valleys of Lakes Albert and
Tanganyika, Rukwa and Magadi, the observed values of
gravity, are abnormally low (Fig. 232). Like the “negative
Fig. 232
Sections across four of the African rift valleys with gravity anomalies (deter-
mined by E. C. Bullard) plotted below. A marked deficiency is clearly shown
beneath each of the rifts
belts ” of the East and West

Indies (pages 404 and 406),
these rift must be underlain by an excess of light
valleys
material. This important discovery means that if the rift
blocks were free to move under gravity as they would be —
if the boundary laults were normal faults due to
tension
they would rise. It follows that the rift blocks cannot have
fallen in under gravity, like a keystone. They must be
held or forced down by the plateau blocks on each side, a
state of affairs implying sideways compression.
If the rigid crust of the plateau region compressed
is later-
ally, it must buckle and then (Fig. 233), if the pressure be
first
sufficient, crack at a place A. When fracture occurs, one side

is thrust over the other, which is
pushed down. After the wall
has reached a height of a few thousand feet, a
second major
440
DYNAMICS OF RIFT FAULTING
crack occurs, accompanied by thrusting. This takes place at

B, where the curvature of the buckled crust is greatest, at a
distance x from the first break. Bullard shows that the distance
x, the width of the resulting rift valley, is determined by the
thickness and elasticity of the crust and the density (say, 3-3)
of the substratum. If the thickness is 20 km., x=39 km. If
the thickness is 40 km., *=65 km. The hypothesis thus satis-
factorily matches the leading characteristics of rift valleys :
their widths and the heights of their walls the rise of the ;
plateau towards the rifts and the occurrence of lop-sided

;
and single-walled “ rifts.” An inter-rift block mountain, like

Ruwenzori, represents the rare case where the second fracture
(or a later one) occurs at a place such as C. The block AC
e ¥ A; Crust of strong crystalline
t
£e
|
: rocks
Stage II
s tage III
C —
_ ^.Rlft
i tun
Single-walled Rlft
Valliy
Fig. 233
Diagrams to illustrate the compression theory of the origin of rift valleys.
(After E. C. Bullard)
is then wedged up, just as a ship caught in the ice of Arctic

seas is heaved up by the ice-pressure on its flanks.
The problem of the origin of rift valleys
is not yet, however,
completely solved. Many competent observers state positively

that some of the most clean-cut fault scarps, like the N.E. wall
of Rukwa (Fig. 226), are normal faults on far too big a scale
to be gravity slumps incidental to underlying thrusts. The
part played by volcanic activity is still far from being under-
stood. The pattern of the rift valleys, their relation to the
Lake Victoria sag, and above all, their relation to the Alpine
orogenic belt in the north have still to be elucidated. The rift
valleys of Africa, one of the major tectonic features of the
world, provide a magnificent field for future exploration and
research.
44 *

W. W. Atwood
The Physiographic Provinces of North America. Ginn, Boston, 1940.
N. M. Fenneman
Physiography of Western United States. McGraw-Hill, New York, 1931.
C. P. Berkey and F. K. Morris

Geology of Mongolia. G. P. Putnam’s Sons, New York, 1927.
G. B. Cressey
China’s Geographic Foundations. McGraw-Hill, New York, 1934.
J. S. Lee
Geology of China. Allen and Unwin (Murby), London, 1939.
J. W. Gregory
The Rift Valleys and Geology of East Africa. Seeley, Service and Co.,
London, 1921.
B. Willis
East African Plateaus and Rift Valleys. Carnegie Institution of Wash-
ington (Publication No. 470), 1936.
E. C. Bullard
Gravity Measurements in East Africa. Philosophical Transactions of the
Royal Society of London (A), Vol. CCXXXV., pp. 445-531, 1936.
442
Chapter XX
VOLCANIC ACTIVITY
General Aspects
As briefly indicated on page 29, volcanic activity includes all

the phenomena associated with the escape at the earth’s surface
of magmatic materials from the depths. An active volcano is
may be the orifice of a pipe or fissure, through
a vent, which
which the magmatic products are continuously or inter-
mittently discharged. It should be recalled that magma is
not merely molten rock : it differs essentially from the latter
in being more or less heavily charged with gases and volatile
constituents, just as soda water differs from ordinary water.
While the magma is confined under a sufficiently high pressure,
its gaseous constituents remain dissolved, but as the surface is
approached and the overhead pressure is reduced, the gases
begin to be liberated, either freely or explosively, according to
local circumstances. The molten material emitted at the
surface is thus relatively impoverished in gas, and is called
lava to distinguish it from the original magma.
In addition to the eruption of hot gases and molten lavas
from volcanoes, vast quantities of fragmental materials are
often produced by the explosion of rapidly liberated gases.
These materials, collectively known as pyroclasts, may them-
selves consist of molten or consolidating lava, ranging from the
finest comminuted particles to masses of scoriae and volcanic
bombs of considerable size, or they may be fragments of older
rocks (including the lavas and pyroclasts of earlier eruptions),
ranging from dust to large ejected blocks, torn from the walls
of the feeding channel or from obstructions in the vent. The
great clouds of gases, vapours, and pyroclasts that are the most
conspicuous feature of explosive eruptions may be luminous
or dark, according as the fragmental material is incandescent
or not. These “ fiery ” and “ smoky ” appearances, together
443
VOLCANIC ACTIVITY
with the glare reflected from the glowing lavas beneath, were
responsible for the formerly popular idea that volcanoes are
“ burning mountains.” Apparently supporting this delusion,
the pyroclastic materials that drop from the volcanic clouds
often resemble cinders and ashes, by which terms, indeed,
they are still commonly described. Actual burning, however,
is confined to the almost imperceptible flames of certain gases
Fig. 234
Diagrammatic section through a composite volcano
The main cone (a) isbuilt of lavas and tuffs fed from the conduit ( b ) and
braced by dykes (c). Formation of an explosion crater ( d ) is followed by growth
of an eruptive cone (r) fed from the conduit ( h ). Some of the later dykes (f)
serve as the feeders of lateral parasitic cones ( g ). Marine deposits interstratified
with lavas and tuffs are indicated by (m)
( From James Geikie, Mountain s : Their Origin Growth, and Decay,
, by permission of
Messrs. Oliver and Boyd)
such as hydrogen and plays only a secondary part in volcanic

activity.
When eruptions take place through a vertical chimney,
the orifice is widened into a crater with flaring sides by out-
ward explosion and inward slumping. By the accumulation

of volcanic products around the vent a conical or dome-
shaped mountain is gradually built up. Volcanoes with the
familiar cone-and-crater structure (Fig. 234) are said to be
of the central type, because the activity is centralized about a
pipe-like conduit. Within the walls of their truncated summits
444
VARIED HABITS OF VOLCANOES
some volcanoes have depressions which resemble

gigantic
greatly enlarged craters.Such a depression may be due to
engulfment of the former superstructure or, less commonly,
to the volcano having blown off its own head, and is called
a caldera. The diameter of a caldera is many times greater
than that of the eruptive vent and it is this contrast in size
which distinguishes it from a crater (Plates 88 and 94). Nearly
all the volcanoes of the present day are of the central type,
but at certain periods in the geological past and occasionally

during historical times volcanoes of the fissure type (Fig. 235)
have poured out prodigious volumes of lava from long fissures
in the crust, and have smothered the surrounding country in
sheets of a far greater order of size than the comparatively
limited flows of central volcanoes.
Fig. 235
Part of the Laki fissure, Iceland, show ing conelets formed towards the end of the
fissure eruption of 1783 (After Helland)
A few volcanoes remain continuously in eruption, but in

most cases activity is intermittent, and sometimes there are
long intervals of repose, during which all signs of activity either
cease or are restricted to exhalations of steam and other
vapours from vents called fumaroles, which at a later stage may
pass into geysers or hot springs (see p. 138).
As a volcano
becomes extinct it passes through similar
waning stages.
Some volcanoes have even been thought to have become
extinct, a disastrous recrudescence of activity proved
until
otherwise. A striking example
is provided by the first recorded
eruption of Vesuvius. At the beginning of the Christian era

the volcano had slumbered so long that no tradition of its
prehistoric eruptions survived. For centuries woods and
vineyards had clothed the fertile slopes of the cone and the
rugged walls surrounding the grassy floor of the summit
caldera were festooned with wild vines. In the year 79 a.d.,
445
—
VOLCANIC ACTIVITY
however, after a period of premonitory earthquakes, the

catastrophic eruption occurred which overwhelmed Pompeii
—
a cultured city of 30,000 inhabitants in an appalling blast
of white-hot ashes, and obliterated Herculaneum in torrents
of hot mud. It was formerly thought that the seaward part
of the caldera was then blown clean away, leaving a half-
encircling collar, known as Monte Somma, within which the
active cone of the modern Vesuvius has since developed
(Plate 92a). However, recent investigations have shown that
most of the missing part of the “ collar ” was blown away
during earlier, prehistoric eruptions.
Those who live far from volcanic districts sometimes
express surprise that men and women should settle where
their crops may be destroyed by vapours and ashes and their
fields and vineyards blotted out by streams of lava. But
volcanoes are not altogether unfriendly, nor are they uniquely
so. Tornadoes, earthquakes, and floods may be equally
devastating. doom that overtook
Calamities like the horrible
Pompeii in 79 and 1902 (p. 469) are fortunately
St. Pierre in
infrequent. All the volcanoes of the world fail to compete
—
with motor transport as a menace to life to say nothing of war.
And compensating for the risk there is the irresistible attraction
of the fertile soils for which volcanic districts are renowned.
The decomposition products of the lavas, rich in plant foods,
are carried down the rain-washed slopes to the plains beyond,
and intermittent showers of ash, if not too heavy, rejuvenate
the soil and add to its bulk. This double aspect of volcanic
—
activity destructive and beneficent —
has been recognized
from the earliest times of which legendary stories remain.
Volcanic Gases
Steam is by far the commonest of the volcanic gases.

Locally the steam may be partly or even wholly derived from
ground-waters and crater lakes, but when full allowance is
made for these superficial sources there is still ample evidence
that the steam liberated in most eruptions is largely of magmatic
VOLCANIC GASES
origin. Dangerous mud flows may be caused when the

torrential rainswhich descend from clouds of condensed water
vapour are accompanied by showers of ashes or sweep loose
deposits down the steep volcanic slopes. Gases have been
collected from cracks in small blisters of lava formed over
gas fountains on the edge of the lava-lake of Kilauea. By
this technique contamination by air is avoided as completely
as is humanly possible. Besides steam (60 to 90 per cent.)
the gases are found to consist, in order of abundance, of carbon
dioxide, nitrogen, and sulphur dioxide, and smaller proportions
of hydrogen, carbon monoxide, sulphur, and chlorine. Similar
gases are liberated elsewhere from active lavas and fumaroles,
together with various related compounds, such as sulphuretted
hydrogen, hydrochloric and other acids, and volatile chlorides
of iron, potassium, and other metals. Incrustations of sulphur
and chlorides, as well as of rarer compounds, often in great
variety, are deposited on cool surfaces.
An explosive eruption is only the final manifestation of the
propulsive power exerted by volcanic gases. Because the
density of magmareduced by the gases held in solution and
is
is still by the separation of gas bubbles, the

further lowered
magma is enabled to ascend to much higher levels than would
otherwise be possible. Another important effect of gases is
that of increasing the mobility and prolonging the active life
of magmas and lavas. A lava, while still retaining part of the
original gas content, may continue temperature
to flow until the
c
is down to 600 But when the same lava has
or 700° C.
crystallized and nearly all its gas has been expelled, the tem-
perature necessary to soften it again is found to be several
hundred degrees higher. It follows that loss of gases involves
rapid consolidation. The chemical some of the
reactions of
gases amongst themselves and with oxygen (whether from
magmatic ferric oxide or the air) may locally generate heat
at volcanic centres and so help to maintain high temperatures
and even for a time to increase them. Striking evidence of
such heat generation has been described from Kilauea (p. 463).
The source of magmatic gases is not know'n, but since at
least some magmas have formed in the crust from rocks poor
447
VOLCANIC ACTIVITY
in gaseous constituents, it is probable that additional gases
have been contributed from still deeper sources. The ascent

of highly energized gases from the primordial sun-born matter
of the earth is held by some vulcanologists to be the fundamental
cause of volcanic activity.
During geological time volcanic action has probably added
considerably to the water in the oceans and to the nitrogen
content of the air. Of the immense quantities of carbon dioxide
supplied to the atmosphere nearly all has been abstracted by
weathering processes and by plant life, and is now represented
in the rocks by limestones and coal and related deposits. The
chlorine locked up in rock salt and in the dissolved salt of sea
water is also very largely of volcanic origin.
Lavas
The temperature of freshly erupted lava is rarely much

above the melting-point, and according to the composition and
gas content it may range from 600° to 1200° C., the basic
lavas, like basalt, being generally the hottest. The mobility
of molten lava depends on the same factors. Silica-rich
lavas are usually stiff and viscous and congeal as thick tongues
before they have travelled far, whereas basic lavas tend to flow
freely for long distances, even down gentle slopes, before they
come to rest. The speed of a lava stream depends on the
mobility and slope, and may, quite locally, reach 50 miles
an hour. But such speeds are very rarely attained even 10
;
miles an hour is unusual and often the movement is sluggish.

In recent years, when approaching flows have threatened
villages on the slopes of Mauna Loa and Etna, the danger has
been averted by bombing from aeroplanes, whereby the flows
have been constrained to follow new and less menacing courses.
The surfaces of newly consolidated lava flows are commonly
of two contrasted types, described in English as block and ropy
lavas, but known technically by their Hawaiian names, aa
(ah-ah) and pahoehoe respectively. Block lava forms over
partly crystallized flows from which the gases escape in sudden
448
STRUCTURAL TYPES OF LAVA
bursts. During the advance the congealing crust breaks into
a wild assemblage of rough, jagged, scoriaceous blocks. Ropy
lava begins at a higher temperature, minute bubbles of gas
escape tranquilly and the flow congeals with a smooth skin
which wrinkles into ropy and corded forms like those assumed
by flowing pitch (Plate 89a and Fig. 236). It sometimes
happens that after the upper surface and edges of a flow of
this kind have solidified, the last of the molten lava drains away,
Pahoehoe lava, Halemaumau, Kilauea
leaving an empty tunnel. Some of the lava caves of Iceland

are famous for the shining black icicles of glass which adorn
their roofs.
When lava of the ropy type flows over the sea floor, or
otherwise beneath a chilling cover of water, it consolidates with
a structure like that of a jumbled heap of pillows and is then
appropriately described as pillow lava (Plate 89b). By the
time each emerging tongue of lava has swollen to about the
size of a pillow the rapidly congealed skin prevents further
growth. New tongues which then exude through cracks in the
449
VOLCANIC ACTIVITY
glassy crust similarly swell into pillows, and so the process

continues. The structure is a common one in the submarine
lavas associated with the geosynclinal sediments of former
periods, and has been seen actively developing in modern flows
that reached the sea floor.
Columnar structure (Plate 3) develops within the interior
of thick masses of lava which have come to rest and have
consolidated under stagnant conditions. It is especially
characteristic of very fine grained plateau basalts which are
relatively free from vesicles.
Pyroclasts
The fragmental materialsblown into the air shower down
at various distances from the focus of eruption according to
[L. Hatches
Fig. 237
Volcanic bombs in ash, Vesuvius
and the heights to which they are hurled. The

their sizes
coarserfragments, including bombs, blocks of scoria and
pumice and blocks of older rocks, fall back near the crater
450
PYROCLASTS
rim and roll down the inner or outer slopes (Fig. 237), forming
deposits of agglomerate or volcanic breccia, the latter term implying
that the blocks consist largely of country rocks from the
foundations of the volcano. Volcanic bombs represent clots
of lava which solidified, at least externally, before reaching
the ground. Some of them have globular, spheroidal, or
spindle-shaped forms due to rapid rotation during flight ;
others, of less regular shape because they were stiff from the
[-4 . Lacroix
Fic. 238
Volcanic bombs. (A) Basalt, Pay de la Vache, Auvergne (g natural size).

(B) Bread-crust bomb oi dacite, Mt. Pelee, 1902 ( J-natural size). (C) Basalt,
Patagonia (§ natural size)
start, have gaping cracks and are described as bread-crust bombs

(Fig. 238).
Smaller fragments, about the size of peas or walnuts, are
(little stones) according to their structure.
called cinders or lapilli
Still finer materials are referred to as ash. These fall mainly
on the slopes and form deposits which, when more or less
indurated, are known as tuffs. Sometimes showers of augite,
felspar, and other crystals that were already present in the lava
(before its comminution) fall from the volcanic clouds and
contribute to the tuffs. The finest particles, down to dust
size, and including shards and splinters of glass, often travel
far beyond the cone before they descend. When such material
is hurled to great heights and caught up by the wind it may
VOLCANIC ACTIVITY
be carried for immense distances. Microscopic volcanic dust

from the catastrophic eruption of Krakatao in 1883 encircled
the world, and its dispersion through the atmosphere was
responsible for the vividly coloured sunsets that were seen
during the following months.
Cones and other Volcanic Structures
The structural forms which result from volcanic activity

depend on the quantities, proportions, and characters of the
lavas and pyroclasts erupted. Some volcanoes have a domi-
nantly effusive habit, lava being the chief product ; a few are
wholly explosive but in most cases eruptions of these kinds
;
either alternate or take place simultaneously.
Classification of Volcanic Forms
Explosive Mixed Effusive

1. Explosion vents 3. Composite cones 4. Lava domes of interna]
(maars) of pyroclasts and growth
2 Ash and cinder lavas 5. Lava domes of external
cones growth (shield volcanoes)
6. Lava plateaus
Explosion vents are mere perforations of the crust,

1.
marked at the surface by small craters, each surrounded by a

low ring of pyroclasts in which fragments of the country rocks
are naturally most abundant. In certain recently extinct
volcanic districts, such as the Eifel (Fig. 239), Swabia (east
of the Black Forest), and the western rift of Africa near
Ruwenzori (Fig. 230), groups of these “ embryonic volcanoes ”
form a characteristic landscape, often diversified by lakes
which occupy some of the crater basins.
2. When a sufficient supply of fragmental material is

furnished an ash or cinder cone is built up. The profile is
determined by the angle of rest of the loose material that
showers down around the vent. Fine ash comes to rest at
angles of 30° to 35°, while nearer the summit the coarser
fragments may stand at 40° or more. In 1538 an eruption
of ash and pumice suddenly broke out in the country west
452
COMPOSITE CONES
of Naples and in a single explosive outburst, lasting only a few

days, Monte Nuovo, 430 feet high, was constructed. In 1937
a similar new cone -was built up at Rabaul in New Britain.
3. Izalco, west of San Salvador in Central America, is an
example of a volcano which began its life as an ash cone in
1770, and has since grown by almost continuous activity into a
typical composite cone standing well over 3,000 feet above the
surrounding country. All the larger volcanoes around the
Pacific and most of those elsewhere are of composite structure.
Fic. 23il
Laacher See, Eifel
and with few exceptions they have histories extending over

many thousands of years. Composite cones (Fig. 234) are
accumulations of successive layers of well stratified tuffs,
alternating irregularly with tongue-like lava flows. The lavas
are generally andesite or other types that can flow with
moderate ease. They may escape through breaches in the
crater wall, or through radial cracks which feed parasitic
craters— often arranged in linear series —
on the flanks. As
magma solidifies in the fissures, dykes are formed which help
(3o«>
453 30
)
VOLCANIC ACTIVITY
to strengthen the growing edifice. Etna, the largest volcano

in Europe, has hundreds of these secondary vents, and although
the summit crater remains a focus of explosive eruptions it is
no longer occupied by molten lava.
As already mentioned, a volcano may suffer an eruption
of such catastrophic violence that a vast caldera is formed, as
if the volcano had blown off its head. If a caldera did in
fact originate in this way, fragments of the missing material
would form the greater part of the pyroclasts representing the
eruption. When, as is usually the case, such fragments are
rare, the only alternative explanation is that the vanished
part of the cone must have foundered out of sight (Fig. 240).
Fig. 240
Sequence of events in the evolution of a caldera of subsidence
(1) Mild explosions (2) more violent explosions (3) culminating explosions
and cracking of roof (4) collapse of the cone into the magma chamber (5)
growth of new- eruptive cones on the caldera floor ( After R. IF. van Bemmelen
and H. Williams
Most of the calderas of composite volcanoes appear to have

resulted from the wholesale engulfment of the former super-
structure into the space previously occupied by magma, which
was rapidly discharged in explosive eruptions of paroxysmal
violence. The eruptive vent commonly resumes activity
sooner or later and builds up a new cone on the floor of the
caldera.
Crater Lake, Oregon (Plate 94), occupies a huge caldera
about 6 miles in diameter, formed by the collapse of what
was once a lofty composite cone. The cone must have been
there not long ago, for it supported glaciers that have left
abundant evidence of their former existence on the outer
454
CALDERAS
slopes of the caldera walls. Moraines and glacio-fluvial

deposits alternate with beds of pyroclasts ;
and U-shaped
valleys can be followed up to the rim, where they are abruptly
truncated by the walls. It is estimated that the ancestral
cone must have been about 12,000 feet high, and that since
it was sheathed in ice something like 17 cubic miles of the
original structure have disappeared. Howel Williams has

shown that the catastrophic eruptions that made room for
the great engulfment are by vast spreads of
represented
pumice and scoria, the total volume of which is about 12
cubic miles. Fragments of the old cone itself account for no
more than 2 cubic miles. The missing balance may well be
finely pulverized material that was hurled high into the air
and carried far away by the wind, as happened at Krakatao
in 18S3. Evidently the frothy eruptible contents of the
magma chamber were suddenly expelled by a swift succession
of paroxysmal explosions. The conical roof, shattered and
unsupported, then foundered into the depths. At some later
time a small cone, the summit of which rises above the lake
as Wizard Island, was built up by subsequent activity.
—
Even greater calderas sometimes referred to as “ super-
calderas ” —
occur in Japan and Sumatra. One of the largest
of these giant cauldrons, now partly occupied by Lake Toba,
is situated at the crest of the Barisan Highlands in N.W.
Sumatra. has an area of about 700 square miles and a
It
volume of about 300 cubic miles. During the late Pleistocene
there were paroxysmal eruptions of rhyolite-tuffs, followed later
by flows of rhyolite. The tuffs are thickly distributed over
7,000 square miles of Sumatra and they have been traced
into Malaya, where their thickness is still 5 to 20 feet. The
total volume of the tuffs is of the same order as that of the
cauldron itself, which came
into existence as a result of the
titanic outbursts. thought that an upward-expanding
It is
granite batholith arched up its roof when it had nearly reached
the surface and that the gas-laden magma near the top
eventually blew itself out through fissures in the roof. What
was left of the roof then subsided into the eviscerated head of
the batholith.
455
!
VOLCANIC ACTIVITY
4. The shapes of volcanic structures built wholly or

dominantly of lava flows depend on the fluidity of the lava
concerned. The more silica-rich lavas, such as rhyolite, dacite,
and trachyte and the corresponding obsidians, are often so
highly viscous that they cannot flow far from the vent. Steep-
sided and sometimes even bulbous domes are then constructed
immediately over the pipe. Because of the obstruction
further growth takes place mainly by additions from within,
the outer layers being cracked and pushed aside by the internal
expansion. Sarcoui (Fig. 241) and some of the other puys
of the Auvergne, and the “ mamelons ” of Reunion (north
of Madagascar), are notable examples of domes of internal
... . —— ~ —
4-\
v -
\~
a* v
[L. Hawkes
Fig. 241
Dome of Sarcoui, Auvergne
growth. Similar domes form in the craters of certain com-

posite volcanoes by the slow upheaval of stiff lava occupying
the conduit. As cracks develop, obelisks and spines of lava
are squeezed through the top and may reach heights of
hundreds of feet before they are demolished by gas expansion
within them or by explosion from below. A dome of this
kind rose in the crater of Mont Pelee during the disastrous
eruption of 1902, and by its tenacious obstruction of the vent
led to one of the most dangerous types of explosive activity
ever known (p. 462). Later, the plug of the conduit was forced
bodily upwards, through the dome, thus forming the celebrated
“ spine ”
of Mont Pelee (Plate 93b).
5. Highly fluid basaltic lavas,from which gases escape so
easily that explosive activity becomes subordinate, spread out
as thin sheets for great distances. By the accumulation of
456
SHIELD VOLCANOES
successive flows in various directions a wide-spreading dome

with gentle slopes, rarely exceeding 6° or 8°, is constructed.
The classic examples of these shield volcanoes are those of the
Hawaiian Islands others occur in the Samoa group and in
;
Iceland. Hawaii (Fig. 242), the largest of the island chain

to which it belongs, has been built up from the sea floor by
the coalescence of several shield volcanoes. Mauna Loa is the
Fic. 242
Map of Hawaii and its recent lava flows
highest, rising 30,000 feet (nearly 14,000 feet above sea level)
from a broad base 70 miles in diameter. Kilauea lies on its
flanks, 20 miles from the summit and 4,000 feet above the sea.
The “ craters ” are calderas of subsidence containing deep
pits often occupied by swirling lakes of lava, open to the sky.
These sometimes overflow, but at other times the lava drains
away through deep fissures, to emerge lower down on the slopes,
sometimes below sea level. When an underlying magma
457
VOLCANIC ACTIVITY
chamber is thus temporarily emptied, the roof is left with
diminished support. Caldera development then takes place

by foundering within arcuate faults, and the pits themselves
become enlarged by the caving in of slices of the walls.
6. In many volcanic belts the central volcanoes have a well-

marked linear distribution, suggesting that they are fed from
long deep-seated fissures. Owing to varying degrees of ob-
struction, the magma rises to very different levels along a
fissure and finally breaks through at isolated points where the
overhead resistance was least or where gas fluxing through
the rocks was most effective in opening a passage. The vents
thus localised afterwards tend to persist. In regions subjected
to powerful crustal tension, however, deeply penetrating
fissures may provide uninterrupted channels for the swift
ascent of enormous volumes of basaltic magma. Flowing
nearly as freely as water, the lava pours out through long rents
and floods the surrounding country, forming sheets that come
to rest with almost horizontal surfaces. Individual sheets may
average only 20 to 100 feet in thickness, but by repeated
eruptions from swarms of fissures vast basaltic plateaus thou-
sands of feet thick have been accumulated in the past, covering
areas up to half a million square miles.
The largest area of plateau basalts in the British Isles is
that of Antrim (Plate 17a and Fig. 243), but this is only a small
part of the far greater Brito-Arctic region (Fig. 252), extensive
areas of which were flooded with basalts during early Tertiary
times. Although considerable tracts have foundered and are
now beneath the sea, 60,000 square miles still remain in
Antrim, the Inner Hebrides, the Faroes, Iceland, and East
and West Greenland. Shield and other types of volcanoes
arose later at certain localities, including Iceland, where
activity still persists. In our own islands, especially in Mull,
Ardnamurchan, and Skye we see only the basal
(Plate 18b),
wrecks of the ancient volcanoes, worn down to their roots by
millions of years of denudation.
Plateau basalts covering areas of 200,000 square miles or
more occur in the Columbia and Snake River region of the
458
.
PLATEAU BASALTS
Fig. 243
Map of north-eastern Ireland showing the Tertiary plateau basalts (black), the
Tertiary intrusive centres ot Mourne, Slieve Gullion, and Carlingford (dotted),
and the Caledonian Newry Complex (broken vertical shading)
north-western United States (Miocene to Recent), the Deccan

of India (early Tertiary), and the Parana region of South
America Other vast areas which were flooded with
(Jurassic).
basalts in Jurassic or Tertiary times occur in Mongolia and
Siberia, in Arabia and Syria, in many parts of Africa e.g
(
Abyssinia, around the Victoria Falls, and along the Drakens-
berg behind Natal), and in parts of Australia (see Fig. 252 ).
Where the flows thinned out and denudation has since removed
them, the underlying rocks are seen to be penetrated by swarms
459
VOLCANIC ACTIVITY
of dykes representing the feeding channels of the vanished

flows. In many other regions, notably in the Karroo of South
Africa, similar basaltic magmas failed to reach the surface, but
riddled the stratified rocks just below with innumerable dykes
and sills. Altogether, more than a million cubic miles of basalt
have been transferred from the depths during the last 150
million years or so.
During historical times Iceland has been one of the few
regions in which fissure eruptions have been witnessed. The
greatest basalt flood of modern times broke out at Laki during
the summer of 1783. From a fissure 20 miles long torrents of
gleaming lava, amounting in all to three cubic miles, over-
whelmed 218 square miles of country, and sent long fiery arms
down the valleys beyond. As the activity diminished in
intensity, obstructions choked the long rent, gases accumulated
instead of effervescing freely, and small cones were formed at
intervals at points from which the waning lava continued to
exude (Fig. 235).
Types of Central Eruptions
Eruptions vary widely in character according to the pressure

and quantity of gas, and the nature of the lava from and through
which it is released. Several well-defined phases have been
recognised, of which the following are the chief (Fig. 244) :
Hawaiian Type —
Effusion of mobile lava is dominant and
.
gas is liberated more or less quietly. From the surface of a

lava lake jets and fountains of incandescent spray may be
thrown up by the rapid emission of spurting gases (Plate 90 b).
When caught by a strong wind the blebs of molten lava are
drawn out into long glassy threads known as Pele's hair, Pele
being the Hawaiian goddess of fire.
—
When less mobile lava is exposed to the
Strombolian Type .
air ina crater, the pent-up gases escape more spasmodically,
with moderate explosions which may be rhythmic or nearly
continuous. Clots of lava, often incandescent, are blown out,
to form bombs or lumps of scoria, while in phases of more
460
TYPES OF ERUPTIONS
NSSORE or ICELANDIC HAWAIIAN TYPE
PUNIAN
TYPE 'W
t, *1
Diagrams to illustrate the chief types of volcanic eruptions
intense activity the magmatic explosions may give rise to

luminous clouds. Stromboli, one of the Lipari Islands north
of Sicily, normally behaves in this way, its minor eruptions
recurring at short intervals ranging from a few minutes to an
hour or so.
Vulcanian Type (named after Vulcano, also in the Lipari

—
group). The lava is more viscous and pasty, and quickly
crusts over between eruptions. Gases accumulate and gather
strength beneath the congealed cover and blow off at longer
intervals, with correspondingly greater violence. The resulting
volcanic clouds are dark, and characteristically assume a
convoluted or “ cauliflower ” shape as they ascend and expand.
The major eruptions of many volcanoes begin with a vulcanian
phase whenever an obstructed vent has first to be cleared out ;
and they may end in the same way, when the waning activity
is just sufficient to throw out material which has avalanched
into the vent from the unstable crater walls.
Vesuvian Type . —
This is a paroxysmal extension of the
Vulcanian and Strombolian types, the new and specific feature
being the extremely violent expulsion of magma which has
become highly charged with gas during a long interval of
superficial quiescence or mild activity. In consequence of the
preliminary removal of the contents of the pipe down to a con-
461
VOLCANIC ACTIVITY
siderable depth —often as a result of the escape of lava through

lateral fissures and vents —the overhead pressure on the under-
lying magma is relieved. The magma then bursts into an
explosive froth and expels itself as vast luminous clouds of
“ cauliflower ” form. These ascend to great heights, and from
them showers of ashes are widely distributed (Plate 2).
Plinian Type .
—The most violent Vesuvian eruptions some-
times culminate in a stupendous blast of uprushing gas, which
rises to a height of several miles, and there spreads out into an
expanding cloud of globular masses of gas and vapour. This
phase was first observed by Pliny during the catastrophic erup-
tion of Vesuvius of a.d. 79.
Pelean Type . —Here the limit of high viscosity and ex-

plosiveness reached. Upward escape is prevented by the
is
growth of an obstructive dome above the conduit (page 468).

Intermittent spurts of tightly compressed magma force a
passage through lateral cracks, and each of these sweeps
down the slopes as an intensely hot avalanche of self-explosive
fragments of lava, lubricated by constantly expanding gases
and vapours. These downward-rolling explosive blasts,
one of which wiped out St. Pierre in 1902, are commonly
referred to as nue.es ardentes (Plate 93a). They may be either
dark or incandescent, and are not always “ glowing ” as the
French term suggests. Perret, who has observed them at close
quarters, defines a nu'ee ardente as an “ avalanche of an ex-
ceedingly dense mass of hot, highly gas-charged and constantly
gas-emitting fragmental lava, much of it finely divided,
extraordinarily mobile, and practically frictionless, because
each particle is separated from its neighbours by a cushion of
compressed gas.”
Kilauea
Kilauea is a low basaltic dome no more than a slight —

—
undulation on the eastern flanks of Mauna Loa (Fig. 242),
with a steep-walled caldera, about three miles in diameter,
sunk in its flat summit. The active vent in the caldera floor
462
HALEMAUMAU
is the deep fire-pit of Halemaumau (Plate 90), in which the
lava column alternately rises and falls, occasionally disappear-
ing out of sight, sometimes crusting over, but generally forming
a lake of surging lava, which presents a magnificent spectacle
at night as dazzling pools and streaks break through the duller
ruddy surface and burst into spraying fountains. On account
of its easy accessibility and relative freedom from danger,
Kilauea has become the most closely investigated of all
volcanoes.
In 1912, during a period when the pit was full to the brim,
Day and Shepherd measured the temperature of the lava
and found that in the course of 23 days it rose from 1070° C.
to 1185° C. Over the same period a steady increase in the rate
of gas discharge was observed, culminating at the time of
highest temperature in the maximum development of lava
fountains, of which over 1,100 were seen playing simultaneously
over the surface. As the level of the lake had not varied, it was
concluded that the increase of temperature was due to heat
generated by chemical reactions, involving the more active
of the uprising gases. By collecting and analysing the gases
it was found that reactions of the kind inferred must have
been inevitable. The potent source of heat thus revealed was
however, mainly superficial, for later measurements showed
that the temperature at the surface was about 100° C. higher
than that of the lava at a depth of 20 feet.
For several years the rise and fall of the lake did not exceed
some 700 feet. In 1924 the lava withdrew to a much greater
depth, clearly as a result of drainage through subterranean
fissures. Earthquakes occurred 30 miles to the east and be-
yond, and finally, out on the sea floor along the same
line of disturbance, the lava found an exit. The walls of
Halemaumau, no longer sustained by lava, avalanched into
the pit while it was emptying, thus enlarging it at the
top and choking it at the bottom. Ground-water seeped into
the debris, passed into high-pressure steam, and so removed
the successive obstructions in a series of violent explosions
(Plate91a and Fig. 245). Such explosive activity probably
occurs in Hawaiian volcanoes only on the rare occasions
463
VOLCANIC ACTIVITY
when the underlying magma falls far below the level of the
ground-water.
the pit was thoroughly cleared out, it was found that
When
the avalanches and explosions of a few weeks had increased
its surface dimensions from 800x500 feet to 3,400x3,000 feet.
[//. T. Stearns, U.S. Geol. Survev
Fig. 245
Halemaumau, May 22, 1924. Photo taken

Explosive eruption from the pit of
from the Hawaiian Volcano Observatory on the edge of the caldera of Kilauea
The floor of the great cauldron, 1,300 feet below the edge, was
seen to consist of solid, well crystallised rock, steaming vigor-
ously, but showing no sign of a central conduit for the lava.
Only small feeding channels and gas vents could be detected
in the walls, and when the lava began to return it first broke
464
VESUVIUS
through a scree of volcanic debris as a brilliant fountain that

spurted to a height of 175 feet. Halemaumau is evidently
a collecting chamber fed by a number of relatively small
channels leading back into the inaccessible regions from which
the magma ascends. In 1929 fountains 200 feet high were
observed, and for a time the lava rose in the pit at about
5 feet per hour. So far, however, in the present cycle of
activity the fluctuating level of the lake has not yet succeeded
in reaching the rim of the now greatly enlarged pit.
Vesuvius
During the sixteen centuries that followed the eruption of

a.d. 79 Vesuvius broke into violent eruption only ten times,
each outbreak being followed by a prolonged period of quies-
cence. With the eruption of 1631, after 130 years of repose,
the volcano assumed its modern habit of continual, but
cyclic activity, the cycles being marked by definite cres-
cendoes leading to outbursts of paroxysmal intensity at inter-
vals which so far have varied from 11 to 40 years. The
last two of these major eruptions occurred in 1872 (Plate 2)
and 1906 (see footnote on p. 468).
Throughout the 1906 eruption, Frank Perret, most courage-
ous of vulcanologists, remained at the observatory on the
western slopes, and his intimate record of the sequence of
events has become one of the classics of geological literature.
In 1905, when the eruption was already threatening, the cone
fissured high up on the north-west slope (A, Fig. 246), and thus
led to an emission of lava which temporarily relieved the tense
conditions in the crater. On 4th April 1906, the cone was
fractured at much lower levels (B-G, Fig. 246), and torrents
of gas-saturated lava gushed out. The effect of the relief of
pressure beneath the crater was a terrific intensification of
roaring explosions which kept the whole mountain in a con-
tinuous state of powerful vibration. White-hot lava, torn into
fragments, was projected for miles into the overhanging ash
cloud, like a fiery effervescing geyser. Intense electric dis-
465
—
VOLCANIC ACTIVITY
Fig. 246
Vesuvius and Monte Somma showing the points of emission of lava in 1905 (A)
and 1906 (B— G)
charges added to the lurid glare. At one stage a formidable

jet of projectiles shot over Monte Somma, and brought disaster
to the towns on the plains beyond.
—
Four days after the paroxysm the Vesuvian phase
began, it culminated in a mighty uprush of gases— the Plinian
—
phase (Fig. 244) which continued for the greater part of a
day, blasting out the throat of the chimney, tearing away
the upper portions of the cone, and reaching a height of 8
miles before spreading out. The level from which the gases
were escaping must by now have descended into material
too solid to be expelled from the volcano. An enormous
volume of this medium must, however, have been drained to
feed the prodigious blast. As the gas pressure gradually
be-
came less overwhelming, the walls of the widened crater began
VESUVIUS SINCE 1906
Fig. 247
Diagrammatic representation of fourteen years’ growth of the active conelet of
Vesuvius after the eruption of 1906
to fall in, creating temporary obstructions and thus leading to
powerful explosions and the generation of dark Vulcanian

clouds heavily charged with ash. This concluding stage of the
eruption persisted intermittently for several days, during which
the intensity gradually declined. By 22nd April the eruption
was over.
During the next few years the deep funnel of the crater was
gradually filled up by avalanches from the walls. The floor
of debris first came into view in 1909. Apart from the emission
of vapours, the volcano remained in repose until 1913, when
incandescent lava perforated the floor, and began to build up
a new eruptive conelet (Fig. 247). Since then Vesuvius has
been continuously active, long periods of quiet effervescence
of the lava within the crater of the conelet alternating with
occasional outflows which, on the whole, have tended to
become more voluminous as the years have passed (Fig. 248).
Judging from the past behaviour of the volcano it would appear
467
VOLCANIC ACTIVITY
Fig. 248
Sketch-maps showing the lava flows from Vesuvius in three recent years
that the crescendo of activity is approaching the climax of

another major eruption.*
Mont Pel£f
Before the tragic catastrophe of 1902, Mont Pelee (Marti-

nique, in the volcanic arc of the Antilles) had long been dor-
mant, the only previous activity known to the inhabitants
having been quite moderate Vulcanian eruptions in 1792 and
1851. In the early spring of 1902 Vulcanian explosions again
broke through, but this time the open crater so formed was
soon filled up and sealed over from within by the growth
* Since this sentence was written the threatened eruption has actually
occurred (March 1944),
468
MONT PELEE
of a dome of extremely viscous lava. The deeper, highly
explosive magma then began to escape at intervals through
lateral cracks, each outburst forming a detached globular mass
which rapidly expanded into a nuee ardente (Plate 93a), sweep-
ing down to the sea with the violence of a hurricane. Most of
these were directed towards the west, but on 8th May an
exceptionally powerful nuee unexpectedly burst out towards
the south. St. Pierre, the capital of Martinique, lay across the
track, and within a few minutes the whole city and its 30,000
inhabitants were utterly annihilated by the irresistible, asphyxi-
ating blast (Plate 93b). In the harbour ships turned turtle
and sank in the boiling sea. After this unparalleled disaster
activity continued for many months, but all the subsequent
nue'es followed their earlier track.
In October the plug of viscous lava in the chimney began
to be forced bodily upwards, and a gigantic steaming column
rose high above the crater-dome. Presently half of it broke
away along a slanting crack from the summit, the part that
remained thus assuming the shape of a spine. The latter
reached a height of about 900 feet above the dome in the course
of seven months, but by July in the following year it had
crumbled to pieces as a result of gas expansion within it, gas
fluxing around its base, and ordinary weathering.
Only two days before St. Pierre was wiped out a similar
devastating eruption occurred in the island of St. Vincent to
the south. After a series of violent earthquakes the crater lake
of the Souffriere volcano boiled and overflowed, and a black
nuee, heavily charged with incandescent masses of lava, de-
scended from the summit and destroyed everything in its path.
Here the death roll amounted to about 1,000.
From 1929 to 1932 Mont Pelee again became active, the
preliminary symptoms being earth tremors and rumblings
and increased emission of gases and vapours from fumaroles.
Soon after the eruptions had started Pcrret took up his quarters
on the volcano, and kept it under watchful observation. He
—
was soon able to reassure the population most of whom had
fled after the first nuee appeared —
and the normal business of
St. Pierre was gradually resumed. The eruptions followed a
469 31
VOLCANIC ACTIVITY
Fig. 249
View of the growing dome of Mt. Pelee in 1930, shouing spines in various stages
of protrusion and coilapse
{Drawn Jr om a telephoto taken from St. Pierre by F. A. Ferret)
remarkably rhythmic course, periods of nuee discharge alter-

nating with periods of dome formation with very little over-
lapping. Innumerable spines of viscous lava were squeezed
through the dome during its phases of growth (Fig. 249), but
within a short time, never more than a few days, each of them
collapsed into fragments which slithered down the ashy slopes.
The nuees all descended to the sea along a valley well to the
north of St. Pierre.
Krakatao
For two centuries before its impressive awakening in 1883
(Figs. 250 and 251), the old volcanic wreck of Krakatao (in
the Sunda Straits, between Java and Sumatra) had been
dormant. In May of that year the vent of Perboewatan
became active, Vulcanian explosions being followed by erup-
tions of moderate Vesuvian type. During the next few
weeks many new vents were opened around Danan, until by
August at least a dozen Vesuvian eruptions were in progress,
and steadily increasing in violence. The climax was reached
during the last week of August. On the 26th formidable de-
tonations were heard every ten minutes. Dense volcanic clouds
reached a height of 17 miles and ashes, transformed into stifling
mud by the incessant rain, fell over Batavia, which was plunged
into thick darkness, relieved only by vivid flashes of lightning.
470
KRAKATAO
Fig. 250
Block diagrams illustrating five stages in the history of Krakatao
{After B. G. Escher)
I Original andesitic cone of Krakatao

II After explosive evisceration, probably accompanied by collapse of the
superstructure, a great caldera was formed, rimmed by three small islands
IIIGrowth of Rakata, a basaltic cone
IV Krakatao before 1883, after two later andesitic cones had coalesced with
Rakata
V After the 1883 eruptions. Later eruptions have since built up the island
of Anak Krakatao within the caldera
On the morning of the 27th came four stupendous explosions,

the greatest of which was heard 3,000 miles away in Australia,
and a vast glowing cloud of incandescent pumice and ashes
rose 50 miles into the air. Although Krakatao was unin-
habited, the catastrophe did not forgo its toll of life. Enormous
47i
—
VOLCANIC ACTIVITY
SE
NW Rakata
f ( *el
Se. ;v .
NW Rakata SE
(A) Profile of Krakatao before 1883, with a dotted indication of the original
cone, of which relics a remain on Verlatan and near the base of Rakata ;
b represents the lavas, etc., of the later cones (stages III and IV of Fig. 250),
and c the gas-charged magma basin responsible for the 1883 eruptions.
(B) Profile of Krakatao after the eruptions of 1883. In this section c mainly
represents the materials that collapsed into the eviscerated magma chamber
sea waves, one of them 120 feet high, swept over the low coasts
of Java and Sumatra and 36,000 people were drowned.
When Krakatao again became visible, it was found that
two-thirds of the island had disappeared. Subsequent survey
showed that a deep submarine hollow had taken the place of
eight square miles of land. It was originally thought that the
greater part of the island —amounting to about 4 cubic miles
had been blown away by the colossal explosions. When the
surrounding deposits of tuff came to be examined, however, it
was found that they contain less than 5 per cent, of material
representing the vanished rocks. All the rest, consisting of
glassy ash and pumice, is a product of the magma that was
responsible for the eruption. Thus it was, for the most part,
not the rocks of the volcanic cover, but the contents of the
underlying magmatic reservoir that were blown away. The
cover then subsided, leaving a vast island-rimmed submarine
caldera, 4 by 4-5 miles across, at the surface.
After remaining dormant for 44 years, a new active vent
(Plate 91b) broke through the caldera floor late in 1927 and
472
DISTRIBUTION OF VOLCANOES
continued to erupt at intervals until 1933, building up a cinder

cone known as Anak Krakatao ( Anak “ child of”). It is of ,
interest to notice that the material erupted in 1883 was dacite,

averaging 65 per cent, of silica, whereas the bombs and lapilli
from the new vent were basalt, averaging 52 per cent, of silica.
The Distribution of Volcanoes
Nearly five hundred volcanoes are known to be active or

to have been in eruption during historical times, but besides
these many thousands of extinct cones and craters are still so
perfectly preserved that they must equally be taken into
account in considering the distribution of recent vulcanism.
Reference to Fig. 252 shows that a high proportion of the
late Tertiary to still active volcanoes are situated in or near
the Circum-Pacific and Alpine-Himalayan belts of the latest
orogenic cycle. Another noteworthy association with profound
crustal dislocations is illustrated by the past and present
vulcanism within and adjoining the African rift valleys and
their continuations north of the Red Sea. Other scattered
lines and groups of volcanoes occur in the Pacific, Atlantic,
and Indian Oceans.
Two-thirds of the active volcanoes and an immense number
not long extinct are distributed around the borderlands and
island festoons of the Pacific. The accompanying list indicates
the active occurrences by numbers (excluding merely parasitic
South-Eastern Continuation
of Alpine-Himalayan Belt Circum-Pacific Girdle
Aleutian Is. and Alaskan Peninsula (35)

Kamchatka (9) W. Canada (-)
Kurile Is. (13) W. United States (1)
Sumatra (11) Japan (33) Mexico (9)
Java (19) Philippines (98) Guatemala (14)
Lesser Sunda Is. (15) S.E. New Guinea and Nicaragua (7)
Moluccas (3) islands to N. (15) Costa Rica (5)
Solomon Is. (2) Lesser Antilles (9)
New Hebrides (7) Northern Andes (11)
Tonga (6) Central Andes (9)
Kermadec (1) Southern Andes (22)
New Zealand (4) Southern Antilles (2)
473
VOLCANIC ACTIVITY
latest
Mongolia Arabia,
the
,
and
to
(India)
Syria
:
belonging
are
Deccan
Abyssinia,
map)
basalts
the
Brito-Arctic,
on
Kimberley,
plateau
:
(named
and Tertiary
basalts
Lower
extinct)
Victoria-Queensland,
active)
plateau
long
Tasmania.
(still
the
many
of
River,
Iceland
ages Zambesi,
(including
Snake
The
and
Drakensberg,
volcanoes activity.
of Columbia
of Basin,
cycle
A
ran
distribution
Pa Patagonia,
geological
Siberia,
the :
: Recent
showing Jurassic
to
Miocene
Early
Map
DISTRIBUTION OF VOLCANOES
vents). Figures are also given for the highly active belt of the
Dutch East Indies, where the far eastern continuation of the
Alpine-Himalayan orogenic belt faces the Indian Ocean. In
two places the Pacific girdle of fire ” encroaches on the
Atlantic : in the volcanic loop of the Lesser Antilles, and in
the similar loop of the Southern Antilles, which links Patagonia
with Grahamland. The association of volcanoes with coasts
of Pacific type is clearly marked. In striking contrast, coasts
of Atlantic type are relatively free, and along vast stretches
entirely free, from volcanic activity.
A linear or arcuate arrangement of vents along the orogenic
belts is highly characteristic, and is well illustrated in the Dutch
East Indies (Fig. 211). Here and in similar arcs elsewhere
(Fig. 213) the outermost zones, where earthquakes are frequent
and mountain building is still in progress, show no signs of
vulcanism. The volcanoes, whether active or extinct, are con-
fined to parallel zones situated a hundred miles or more on
the concave side of the zones of present-day tectonic activity.
The vents are commonly strung out along lines of fracture in
belts that were folded and uplifted during an earlier phase
of the orogenic cycle now drawing to its close.
—
Along the Alpine-Himalayan orogenic belt apart from the
—
Dutch East Indies volcanoes are distributed more sporadi-
cally. The volcanic zone can be traced from Madeira and
the Canary Islands through the Mediterranean region ( e.g .
Vesuvius, Lipari Islands, Etna, and the JEgean volcanoes) to

the Caucasus (Elburz), Armenia (Ararat), Persia (Demavend),
and Baluchistan.Beyond the Himalayas the line continues
through Yunnan and Burma to the Andamans and the little
and so to the Dutch East Indies. The
islands to the east,
spacing is very uneven and there are long gaps, especially
along the Himalayas and the Alps. It has been suggested
that in such regions, where the crust has been intensely com-
pressed by severe overfolding and overthrusting, the structure
blocks the passageways which might otherwise have con-
tinued up to the surface, and so favours the injection of magma
along the thrust planes rather than across them.
However this may be, it is clear that many of the European
475
VOLCANIC ACTIVITY
volcanoes of the present cycle broke through the median

areas
around the
within the branches of the orogenic belt {e.g.
Tyrrhenian Sea along the Mediterranean coasts of Spain
;
and North Africa ;

and in the Hungarian basin), and also
through the forelands outside {e.g. the volcanic districts of the
Auvergne, the Eifel, and Bohemia). Further east, Tibet seems
to be free from volcanoes, possibly because it stands too high ;
only a few hot springs emerge. Active cones have been re-
corded from less lofty situations, however, near the rim of the
Tarim basin in the very heart of Asia, and others are known
in Mongolia and Manchuria. A noteworthy example of a
Circum-Pacific median area with many recent vents and cones
is the Colorado plateau (page 422).
Reference to the vulcanism associated with the African

rift valleys has already been made on pp. 435 to 437. The only
active volcano in West Africa is Cameroon Mountain, which
forms part of a long volcanic chain that extends as a string
of islands far into the Gulf of Guinea. Well-preserved cones
and craters have recently been observed in the deserts of
northern Africa, and there are many older volcanic piles
among the highlands of the Sahara.
Several of the volcanic islands of the eastern Atlantic have
already been mentioned, and to these must be added the Cape
Verde group. Well within the ocean the islands that rise from
the mid-Atlantic swell {e.g. the Azores, St. Helena, and
Ascension) are all volcanic. Most of these are now extinct,
but where the swell meets the ridge from Britain to Greenland
there occurs one of the world’s greatest lava fields —
Iceland,
with over a score of active volcanoes. Further to the north-
west, just before the swell dies out, the most northerly of active
volcanoes forms the island of Jan Mayen.
Besides Kerguelen, in the far south, the active volcanoes
of the Indian ocean belong to the island groups of Comoro,
Mauritius, and Reunion, all near Madagascar, which itself
has many cones not long extinct. The scattered oceanic
islands around Antarctica are also volcanic. On the fringe
of Antarctica there are volcanoes bordering the Ross Sea,
including the active cones of Erebus and Terror. The Pacific
476
SIGNIFICANCE OF DISTRIBUTION
has several groups and linear series of volcanic islands, while

the innumerable coral islands point to the former existence
of many more which have been worn down to submarine
platforms, and are now capped with atolls. North of the
Equator is the magnificently developed Hawaiian chain
(p. 457). The Galapagos Islands lie across the Equator off
the South American coast. South of the Equator there are
the volcanic islands of Juan Fernandez (active) and Easter
Island (extinct) in the east, and of Samoa in the west. Those
to the west and south-west of Samoa ( e.g Fiji, Tonga, and
Kermadec) belong to an outlying arc of the Circum-Pacific
belt.
In the early days of geology, while the interiors of the
continents still remained largely unexplored, it was not un-
naturally assumed that all volcanoes occur near the sea. This
apparent “ rule ” suggested the erroneous idea that sea water
infiltrated into the magmatic reservoirs and so caused eruptions
by being converted into steam. Such accidental steam
explosions do occur, of course, but they are merely a super-
ficial consequence of volcanic activity already there, and
they are in no way to be regarded as its cause. The old
generalization can no longer be justified, as the facts of
distribution already outlined prove beyond dispute. More-
over, the distribution of the Jurassic and Tertiary plateau
basalts (Fig. 252) shows that in the past widespread con-
tinental regions, most of which lay far from the seas and
oceans of the time, were deluged with lavas on a prodigious
scale.
Volcanic activity everywhere takes full advantage of pro-
found fracturing of the crust, and since the greatest crustal
deformations occur in the orogenic belts surrounding the
continents (Figs. 209 and 210), it is hardly surprising that
there should now be many volcanoes near the oceans. What
is a matter for surprise is that there are not more volcanoes
associated with the fractures responsible for coasts of Adantic

types. Obviously fracturing is not a sufficient condition for
vulcanism. The essential condition is that there should be
magma in depth to be tapped or to force its way up through
477
VOLCANIC ACTIVITY
openings of its own making. The real problem of vulcanism

is that of the origin of magmas, and to this subject we
now
turn our attention.
Speculations on the Causes of Vulcanism
Penetration of the earth’s crust by bore-holes and mines

shows that the temperature increases with depth. The rate
of increase, or, in other words, the temperature gradient, varies
considerably from place to place. Away from active volcanic
centres the average gradient is about 30° C. per km., but in
Ontario and the Transvaal it falls to as litde as 9° or 10°
per km. If the downward rate of increase continued uni-
formly, temperatures of fusion, say about 1050° C., would be
—
Granitic I Layer Fig. 253
It km. * 2-7 Diagram to illustrate the ascent of
basaltic magma through a fissure, as
a result of crustal pressure (P) acting

35 on a molten mass near the base of the
Basaltic Layer of the
basaltic layer. The densities
24-km. 30 rocks of the granitic and basaltic layers
are taken as 2-7 and 3 0 respectively
]P ]P_y l_4<_
BASALTIC ‘MAGMA
reached in “ average ” regions at a depth of about 35 km.
For this reason it was formerly thought that underlying the
solid crust there was a world-wide layer of magma. As a
result of high pressure such magma would be extremely stiff,
and probably more like a glass than a mobile fluid, but never-
theless it would be capable of flow, and it would at once
become eruptible if it were tapped by a fissure leading up
to the surface.
If the lower part of the basaltic layer were actually in
a state of fusion, it is easy to see that the magma could be
driven to the surface, and even to the summits of lofty volcanoes,
by the head of pressure due to the weight of the overlying rocks.
Taking the average crust illustrated in Fig. 253 as an example,
the pressure exerted by the rocks on a thin layer of basaltic
478
ASCENT OF BASALTIC MAGMA
magma would be proportional to 11 X 2-7 (= 29-7 due to

the granitic layer) plus 24 x 3 ( = 72 due to the basaltic
layer), making a total of 101-7. Let h be the limiting height
to which the magma could be forced through a fissure feeding
into the pipe of a volcanic cone. Since the density of molten
basalt is about 2-7, we have
2-lk = 101-7 ;
whence h = 37-67 km. = 35 + 2-67 km.
Thus, not only would the magma reach the surface, but it
could build up a cone and rise within the crater to a height
of 2-67 km. or 9,750 feet. This is comparable with the highest
of continental basaltic volcanoes. Etna, with a crater rim
reaching an elevation of 10,739 feet, has probably reached
the limit, for most of its eruptions are now confined to vents
that have opened on the slopes below the rim. It should, of
course, not be overlooked that the gases and vapours which are
present in natural magma lower its specific gravity (as com-
pared with that of artificially fused basalt) and so help its ascent,
,
especially in the vicinity of the surface, where the concentrat-

ing gases may eventually burst out with explosive violence.
The existence of active basaltic volcanoes and of vast
spreads of plateau basalts proves that, locally and at certain
times, adequate supplies of basaltic magma must be available
in the depths. The old idea that there is a world-wide eruptible
layer of basalt can, however, no longer be sustained. As we
shall see presently, the downward increase of temperature falls
off in depth, and the normally at a level far
basaltic layer is
above the depth at which it could exist in a molten state.

Moreover, if there were such a molten layer, a fissure eruption
of plateau basalt, once started, should continue without
interruption until all the eruptible material had either reached
the surface or formed dykes and sills on the way up. Actually,
we find that each of the great basaltic plateaus is made up
of a large number of separate lava flows, and that there was
a long interval of quiescence after each flow, during which its
top was weathered to soil or laterite, sometimes to a depth
of many feet. It follows that the available supply of magma
at any one time was limited and rapidly exhausted, and that
479
VOLCANIC ACTIVITY
the source must have been replenished again and again.

Theoretically, such replenishment might be brought about
either by localized fusion of the crystalline rocks of the basaltic
layer, or by the ascent of fresh supplies of basaltic magma
from still greater depths.
In the light of our present knowledge the problem of the
origin of basaltic magma has become extremely difficult, and,
strictly speaking, it cannot be regarded as solved at all, except
in a speculative way. The main trouble arises from the
distribution of temperature in depth. In 1906 Lord Rayleigh
detected the presence of minute quantities of the radioactive
elements in common rocks from all parts of the world (p. 409).
It was then realized that the rocks of the earth’s crust, including
basaltic lavas, contain within themselves an unfailing source
of heat. At first it was thought that the heat generated by
radioactive disintegration would be sufficient to ensure
temperatures of fusion near the base of the basaltic layer.
Volcanic activity was then interpreted as the means by which
the molten material escaped to the surface, so allowing the
steadily accumulating heat to be periodically dissipated.
Even if the temperature at the base of the basaltic layer
normally fell a little short of that required to bring about
fusion, it might locally reach the fusion point in places where
the crust was depressed and thickened, e.g. beneath geosynclines

blanketed by heavy sedimentation, and still more in the deep
mountain ranges, where the radioactive crust
roots underlying
becomes abnormally thick. However, it is no longer possible
to account so easily for vulcanism. More recent investigations
have shown that the crustal layers are not so thick
(a) and —
therefore the basal temperatures are not so high —
as was then
supposed ;
(b) that the actual flow of heat through the crust
in non-volcanic regions is far less than it would be if fusion
temperatures were reached anywhere near the base of the

basaltic layer ; and (c ) that in the roots of orogenic belts
the time required for temperatures to rise to the fusion point
would be well over a hundred million years, which means
that the igneous activity contemporaneous with mountain
building must have some other cause.
480
HEAT FLOW THROUGH THE CRUST
In assessing the downward increase of temperature we have
to take into consideration not only the temperature gradient
observed near the surface, but also the thermal conductivity
of the crustal rocks and the quantity of heat generated within
those rocks by the radioactive elements. The following
simplified treatment of this problem gives an approximate
c\ ;
H H/Hj
(
(Heat generated in GRANITIC LAYER)
1
L> H4 fHeat generated in BASALTIC LAYER)
Hs LOWER LAYER
Fig. 254
Diagram to illustrate the heat flow through the earth’s continental crust (away
from volcanic areas), with data for calculating the downward distribution of
temperature (see text)
Upper or Granitic Layer
T0 = temperature of ground at surface (about 10° C.)
T =4
temperature at base of granitic layer (to be calculated)
d = thickness of granitic layer (11 km. = 11 x 10 s cm.)
k = thermal conductivity of granitic rocks (-0068 c.g.s. units)
Hj = heat flow escaping at surface (1-3 x 10 6 cals/sec. /sq. cm.)
H2 = heat flow from the radioactive elements in the granitic layer (-33 X 10' *
c.g.s. units)
H =3 heat flow entering from beneath (= H 4
—H a
= -97 x 10 * c.g.s. units)
Intermediate or Basaltic Layer
T2 = temperature at base of basaltic layer (to be calculated)

D = thickness of basaltic layer (24 km. = 24 X 10 5 cm.)
K = thermal conductivity of basaltic rocks (-004 c.g.s. units)
H4 = heat flow from the radioactive elements in the basaltic layer ('65 x 10'*
c.g.s. units)
H =5 heat flow entering from beneath (= H 3 — H, = -32 x 10'* c.g.s. units)
soludon, using the data summarized under Fig. 254. It

should be explained that the heat flow at the surface is given
by the product of the temperature gradient by the thermal
conductivity of the rocks through which the heat flows, both
of which factors can be accurately measured. It is found
that the heat flow in different places varies much less than
the corresponding temperature gradients. In Britain, for
example, it ranges between 2-07 x 1 0~ 6 calories per second
481
—
VOLCANIC ACTIVITY
per cm. near Glasgow (where the influence of the Mull

sq.
swarm of dykes makes itself felt cf. Fig. 38), and 1-1, in the
same units, around London (where there was no Tertiary
igneous activity, and probably none for many hundreds of
millions of years). In South Africa the average heat flow is
IT 6, and in Michigan about 1-0. The average heat flow in
non- volcanic regions appears to be about 1-3 in the units
stated.
Theheat flow Hj from the granitic layer is the sum of 2 H ,
which the heat generated by the radioactive elements in

is
the layer, and 3 H

which is the heat that enters from below.
,
The average heat flow is (H 4 -f 3 )/2 and the average H ;
temperature gradient is (T t — T0 )/d.

Hence, we have
H,fH 3 _ k( T, -T 0)
2 d
whence t
Tl
1
T + d/Hi
= T
r i ,
T H3 \
0
From the data listed under Fig. 261,
T =
11 x 10 5 /1-3 + -33
=
x
10° + •0068
X 10-« 195° C.
\ 2
Similarly, the heat flow H

3 from the basaltic layer is the
sum of 4 H
which ,
is generated within the rocks of the basaltic
layer, and H 5 which is the heat that enters from below. The
,
average heat flow is (H 3 + H 5 )/2 ; and the average temperature

gradient is (T 2 — Tj)/D.
As before, Sl± H = HIl ~ T ’-

i)
2 D
whence, T =T +
2 3 5. f
= 195° +
24 x 10 5 /-97 + -32\
X 10-6 585° C.
004 [ 2 )
The curve drawn through these values for T 4 and T2 in

Fig. 255 gives the approximate distribution of temperature
482
DOWNWARD INCREASE OF TEMPERATURE
through the average crust. be seen that throughout

It will
the granitic and basaltic layers the temperatures are every-

where far below those required for magma formation. The
gap amounts to hundreds of degrees. The data are naturally
subject to variation from place to place, but, so far as we
know at present, no permissible variations in the data adopted
make it possible to bridge the gap. Yet the gap must some-
times be bridged to account for the local existence of basaltic
magma, and still more for the local generation of granitic
magma at no great depth beneath the surface. Evidently
_ ? Peridotite.?. . . .
Tempers Cure r
Diagram showing the approximate downward increase of temperature in a non-

volcanic continental region, together with the temperature required at each depth
for the generation of magma from the rocks of the crustal layers
enormous quantities of heat somehow appear in the orogenic

belts, and also beneath those parts of the more stable crust
where plateau basalts have flooded the surface. How is it
”
that these areas have become the temporary “ hot spots
of the earth, while the rest of the crust was cooling ? It is
obvious that the continental crust does normally cool, for
almost every part of it has been the site of igneous activity
at some time or other during the earth’s history. Throughout
geological time the hot spots have shifted about from one set
of localities to another, always accompanying the orogenic
belts of the time, but also appearing more or less at random at
various places in the intervening tracts.
From these considerations it appears that the local heating
483
VOLCANIC ACTIVITY
up of the crust is to be ascribed to processes operating in the

underlying substratum. And here we inevitably enter the
uncertain realm of speculation where there are various hypo-
thetical possibilities to be explored, but few guiding facts.
One possibility is that magma generation may be a consequence
of sub-crustal convection currents, and this is perhaps the
most promising suggestion available at the moment.
It conceivable that the base of the basaltic layer may
is
become fused by heat carried up by hot ascending currents

(cf. Fig. 215), but if such a process were at all important
it
would probably give rise to far more widespread conditions

of crustal fusion than we have any right to assume. It is also
conceivable that beneath the orogenic belts, where opposing
currents meet and turn down, part of the basaltic layer may
be dragged into the currents, as illustrated in Fig. 262 and
described on p. 507. Such basaltic material would be intensely
metamorphosed, and would ultimately become fused, wholly
or selectively. The resulting magma, because of its relatively
low specific gravity, would sooner or later ascend again. Part
of it might be squeezed back into the higher levels of the
orogenic belt, but much of it would probably reach the crust
far from its original source, and so become responsible for
basaltic eruptions on the continents and ocean floors. It
should be noticed that although Fig. 262 is drawn to illustrate
the case where magma ascends from beneath the ocean basins,
the mechanism portrayed is just as likely to bring basaltic
magma to continental areas.
The most conspicuously hot regions of the crust are un-
doubtedly the orogenic belts, and this may seem at first sight
to be inconsistent with the hypothesis that these are also the
regionswhere the convection currents begin to descend. Sink-
ing currents cannot be so hot as rising currents. But there
are good reasons for supposing that the crust overlying the
zone where two opposing currents approach and turn down
must inevitably become abnormally hot.
In the first place, the down-dragging of the crust and of
the upper crystalline part of the substratum, during the
development of a geosyncline and subsequently during root
484
PLATE !4
N.Y.
History
lake
the
Nat.
above
Mus.
m.
(yi
rising
Island
Wizard
of
cone
the
with
caldeia,
vast
Oregon.
Lake,
Crater
POSSIBLE SOURCES OF MAGMATIC HEAT
formation, must involve intense shearing and friction, and
therefore the liberation of a great deal of heat. The observa-
tion that certain thin bands of mvlonite (p. 80) show obvious
signs of having been fused proves that the heat due to friction
isby no means negligible. Although it is doubtful whether
such heat can be produced sufficiently rapidly to promote
fusionon any considerable scale by itself, it must nevertheless
be taken into account as an important contributory factor.
In the second place, when rocks are intensely sheared
certain constituents are liberated in a mobile state, so that
they become free to migrate towards places of lower pressure.
This process must operate within the mountain roots, and it
may become even more important in rocks dragged down
into the underlying currents, until it merges into true fusion,
as a result of the steadily increasing temperature. If basaltic
rocks are drawn into the descending currents they will grad-
ually be heated up by their surroundings, and blebs and streaks
of magmawill be generated as the latent heat of fusion is
supplied. Such magma, and indeed all the mobile constituents,
whether liberated by shearing or heat, or both, will be squeezed
back towards the crust, just as water is squeezed out of a wet
blanket as it is passed through the rollers of a mangle. The
descending currents must operate like a mangle towards all
the materials within them down to very great depths, and
consequently there should be an upward streaming of intensely
"heated mobile materials, rich in gases and volatile emanations,
all migrating towards regions of lower pressure — that is,
towards the surface.
Eventually most of these materials will pass into the
orogenic belt, partly as magmas, and partly as a procession
of highly energized and chemically active emanations. As
the latter saturate the rocks (rocks already heated by friction)
the cycle of metamorphism summarized in Fig. 21 reaches
its culmination and new magmas are formed within the crust
itself. From the testimony of the rocks now exposed in the
heart of ancient orogenic belts, it has been inferred that granite
magma can actually be generated in this way (p. 65). A
hypothetical source for the granite-making emanations has
(39«)
485 32
VOLCANIC ACTIVITY
now been traced, but it must be clearly understood that the

process envisaged in the above discussion is at present no
more than a plausible hypothesis designed to account for a
few of the more obdurate facts concerned with the puzzling
behaviour of our planet.

G. A. Cotton
Volcanoes as Landscape Forms. Whitcombe and Tombs, New Zealand,
1944 -
G. W. Tyrrell
Volcanoes. Butterworth, London, 1931.
K. Sapper, I. Friedlaender, and T. A. Jaggar
Physics of the Earth — /. : Volcanology. National Research Council,
Bulletin 77, Washington, 1931.
C. H. Hitchcock
Hawaii and its Volcanoes. Hawaiian Gazette Co., Honolulu, 1911.
F. A. Perret
The Vesuvius Eruption of 1906. Carnegie Institution of Washington
(Publication No. 339I, 1924.
The Eruption of Mt. Pele'e, 1929-1932. Carnegie Institution of
Washington (Publication No. 458), 1935.
A. L. Day and E. T. Allen
The Volcanic and Hot Springs of Lassen Peak. Carnegie Insti-
Activity
tution of Washington (Publication No. 360), 1925.
A. G. MacGregor
The Royal Society Expedition to Montserrat, BAI T. Philosophical
Transactions of the Royal Society of London (Bj, Yol. CCXXIX.,
pp. 1-90, 1938.
H. W ILLIAMS
Calderas and their Origin. University of California Press, 1941.
Crater Lake : The Story of its Origin. U
niversity of California Press
and Cambridge University Press, 1941.
E. M. Anderson
The Loss of Heat from the Earth's Crust in Britain. Proceedings of the
Royal Society of Edinburgh, Vol. LX., Part II., No. 16, 1940.
L. B. Slichter
Cooling of the Earth. Bulletin of the Geological Society of America, Vol.
LIL, pp. 561-600, 1941.
486
Chapter XXI
CONTINENTAL DRIFT
Continental and Oceanic Relationships
The continents are essentially thin slabs of sial, distributed to

form a northern pair, known together as Laurasia, and a
more scattered southern group, collectively referred to as
Gondwanaland. The outer peripheries of the members of
each group are defined by the orogenic belts of the last great
tectonic revolution (Figs. 209 and 210), and the coast-lines,
generally backed by mountains, are of Pacific type (p. 401).
The inner margins of the members of each group, against the
Arctic, Atlantic, and Indian oceans, are fractured and in many
places downfaulted towards the sea, and the coast-lines are of
Atlantic type (p. 400). Across the floors of these intervening
oceans the sial layer, where present at all, is patchy in distri-
bution and very much thinner than in the bordering continents.
Much of the Pacific floor, however, lacks a layer of sial alto-
gether. The major structural units of the face of the earth
thus fall naturally into the following pattern :
la The continents of the Laurasia group

lb The intervening North Atlantic and Arctic oceans
2a The continents of the Gondwanaland group
2b The intervening South Atlantic and Indian oceans
3 The oceanic basin of the Pacific, everywhere outside
1 and 2 (see Fig. 208)
To what extent these primary units and their arrangement

have been stable or otherwise during geological time is one
of the fundamental problems of geology. For nearly a century
this question has been vigorously debated as one line of evidence
after another has been discovered and followed up. Before
then the widespread occurrence of marine sediments over the
lands suggested that the continents could sink to oceanic
487
CONTINENTAL DRIFT
depths and the ocean floors rise to become dry land. It was
gradually recognized, however, that these deposits merely
prove flooding of the lands by shallow seas they do not
;
demonstrate interchange of continent and ocean. For this

reason, amongst others, Dana expressed the view in 1846 that
continents and oceans have never changed places, and that
the general framework of the earth has remained essentially
stable. Nevertheless, Edward Forbes, tackling the subject
from the biological side in the same year, found it impos-
sible to explain how animals and plants had migrated
from one continent to another unless some parts of the
oceans had formerly been land. Thus began the long con-
troversy regarding the permanence of the continents and
ocean basins.
Support for permanency is found in the fact that deep-sea
deposits like those now forming on the ocean floors are confined
to one or two marginal islands (p. 320) and are consistently
absent from the strata now exposed on the continents proper.
Moreover, from the standpoint of isostasy it is very difficult
to picture a process which could bring about widespread
changes of level amounting to two miles or more. The explana-
tion given for the subsidence of geosynclines and the uplift
of orogenic belts can hardly be applied to areas of continental
extent. But since the continents themselves are vast complexes
of orogenic belts of different ages, it is obvious that they must
have been profoundly modified during geological time, and
that a good deal of variation in extent and position must there-
fore be conceded.
The extreme advocates of permanency have also had to
yield ground in face of the evidence that certain regions that
were undoubtedly land long ago are now parts of the Atlantic
and Indian oceans. In Britain the sediments of the Torridonian
and some of those of the Old Red Sandstone and Carboniferous
were derived from a land that lay to the north and w est of r
Scotland. On the other side of the Atlantic the Appalachian

geosyncline was largely filled with sediment from the south
east. The gold-bearing conglomerates of the Gold Coast were
carried there by a great river that drained a land lying to the
488
)
CONTINENTAL DISPLACEMENTS
south. In each of these cases the site of the ancient land is
now open ocean.

What, then, has happened to these vanished lands ?
Theoretically there are three possibilities :
(a) They may have subsided bodily to great depths, while

retaining their original positions on the earth’s surface. This
is the apparently obvious answer, but a serious isostatic
it raises
difficulty. If a tabular iceberg split into two, the separated
bergs might slowly drift apart, but neither could sink. This
analogy has a value in introducing the idea of continental
drift as an alternative to continental sinking.
(
Bodily horizontal displacement may have
b) occurred.
If this happened there would be no subsidence. Labrador
might be the land that formerly lay adjacent to Scotland.
The gold-bearing tracts of the Guianas and Brazil might be
the source of the gold deposits of the Gold Coast.
(c More
probably, however, the crustal layers, including
the sial, may have been
stretched out horizontally between
the displaced continents, the sial thereby becoming thin and
patchy. In this case the resulting isostatic. readjustment would
involve sinking. The known structure of the Atlantic and
Indian ocean floors is consistent with this explanation.
For many years it was naturally assumed without question

that if interchanges between continent and ocean had to be
postulated the movements involved could not be other than
vertical. The
suggestion that there might have been lateral
displacements of the continental masses on a gigantic scale
is generally ascribed to F. B. Taylor in America (1908) and
to Alfred Wegener in Germany (1910). For several years
these pioneers developed their unorthodox hypotheses quite
independently. Actually, however, the same idea had occurred
to Antonio Snider more than fifty years before. In a book
with the optimistic title La Creation et ses Mysteres devoiles
(Paris, 1858) he published the two maps here reproduced as
Fig. 256. Snider’s reconstruction of Carboniferous geography
was intended to explain the fact that most of the fossil plants
489
CONTINENTAL DRIFT
Maps published by A. Snider in 1858 to illustrate his conception of continental

drift. The left-hand map represents the supposed coalescence of the continents
in late Carboniferous times
preserved in the Coal Measures of Europe are identical with

those of the North American Coal Measures. Although the
two diagrams reappeared in J. H. Pepper’s highly entertaining
Playbook of Metals (London, 1801), the idea they embodied was
evidently regarded as too fantastic and outrageous to be
worthy of attention. Not unnaturally it soon became com-
pletely forgotten.
Taylor’s Hypothesis of Continental Drift
It was not until \\ egener published his

famous book on
the subject in 1915 that the possibility of continental drift
began to receive serious attention. But Taylor must be given
credit for making an independent and slightly earlier start
in this precarious field. His immediate object was to account
for the distribution of mountain ranges. He pictured the
original Laurasia as being a continuous sheet of sial and
supposed it to have spread outwards
towards the Equator,
more or less radially from the Polar regions, much as a conti-
nental ice sheet would ao. herever the resistance was least
the crust flowed out in lobes, raising up mountainous loops and
490
)
taylor’s hypothesis
arcs in front(cf. Fig. 209). Such movements, of course, would

be impossible without complementary stretching and splitting
in the rear. And, indeed, there is ample evidence of down-
faulting and disruption in the coastal lands and islands of the
Arctic and North Atlantic, and especially in the highly
fractured region between Greenland and Canada, the map
of which looks like a jig-saw puzzle with the separate bits
dragged apart. In the Southern Hemisphere the originally
continuous Gondwanaland similarly spread out, breaking up
into immense rafts which also migrated towards the Equator
and raised up mountains in front (Fig. 210). The basins of
the South Atlantic and Indian oceans are interpreted as the
stretched and broken regions left behind or between these
drifting continents.
For two reasons Taylor’s hypothesis received scant atten-
tion. As we have already seen (p. 384), a certain amount of
lateral continental movement is implied by the structures of
orogenic belts, but it seemed to be unnecessarily extravagant
to invoke thousands of miles of horizontal displacement when
—
from twenty to forty— rarely more would suffice. Secondly,
Taylor’s attempt to explain the alleged movements was quite
unacceptable. He postulated that the moon first became the
earth’s satellite during the Cretaceous, and that at the time of
its close approach and capture it was very much nearer to the
earth than it is to-day. The resulting tidal forces were supposed

to be sufficiently powerful not only to alter the rate of the
earth’s rotation, but also to drag the continents away from the
poles.
Apart from the improbability that the earth was without
a moon before the Cretaceous, there are two fatal objections
to this hypothesis :
(a) If the late Cretaceous and Tertiary mountain building

is to be correlated with the supposed capture of the moon,
then we are obviously left with no explanation for all the
earlier orogenic cycles.
If the tidal force applied to the earth by the newly

(b
captured moon had been sufficient to displace continents and

49i
CONTINENTAL DRIFT
raisemountains on the scale required, then, as Jeffreys has

shown, the friction involved would have acted like a gigantic
brake and the earth’s rotation would have been brought to a
standstill within a year.
Taylor’s “ explanation ” is completely untenable, but from

the criticisms one very important conclusion may be drawn.
The fact that the earth continues to rotate shows that neither
tidal friction nor any other force applied from outside the
earth can be responsible for mountain building or for conti-
nental drift, if it occurs. We
have already found a cause for
mountain building inside the earth, and if a cause for conti-
nental drift be also required, it too must be looked for within
the earth.
Wegener’s Hypothesis of Continental Drift
Wegener’s highly complex conception of the evolution of

the continents is graphically illustrated by his own strange,
but now familiar, maps (Fig. 257). His picture of the world
in Carboniferous times is strikingly similar to Snider’s, except
that India and Antarctica are tucked in between Australia
and Africa, with the horn of South America forming an outer
wrapping. For this compressed combination of Laurasia and
Gondwanaland he proposed the name Pangaa. In one im-
portant respect, however, neither Wegener’s Carboniferous
map nor Snider’s gives adequate expression to its author’s ideas.
Snider urged that the forests of the Coal Measures were tropical,
and that in consequence Europe and North America must
have been near the Equator. This implies that South Africa
must have been near the South Pole. Conversely, Wegener
inferred that the Carboniferous South Pole occupied a position
just off the present South African coast (Fig. 260). His scheme
thus involves not only continental drift, but also extensive
wandering of the poles.
The present distribution of the continents is regarded as a
result of fragmentation by rifting, followed by a drifting apart
of the individual masses. The southern continents began to
unfold during the Mesozoic era by being dragged away from
492
CONTINENTAL DRIFT
wherever the South Pole happened to be at any given time

during the progress of the outward movements. Somewhat
later North America began to break loose and to drift away to
the west, Greenland being the last to go. The Atlantic is the
immense gap left astern, filled up to the appropriate level by
sima from below. A peculiarity of Wegener’s interpretation
is his insistence that the opening of the North Atlantic was
accomplished almost entirely during the Pleistocene. By the

time the poles had reached their present positions Antarctica
found itself stranded over the South Pole Africa lay athwart
;
the Equator India had been tightly wedged into Asia, where
;
its originally northern part now lies buried under the high
plateau of Tibet and Australia had advanced far into the

;
Pacific, by-passing the Banda arc and coming to rest against

its eastern end.
The drift away from the
of the continents poles was
dramatically described by Wegener as the Polflucht — the Flight
from the Poles. He ascribed it to the gravitational attraction
exerted by the earth’s equatorial bulge. The force is a real
one but, unfortunately for Wegener, it would require a force
many millions of times more powerful than this to drag the
continents from their moorings.
Wegener also postulated a general drift towards the west.
As the Americas moved westwards against the resistance of the
Pacific floor, their prows were crumpled up into great mountain
ranges. Between the two immense rafts a trail of fragments
lagged behind and formed the islands of the West Indies.
The stretched-out isthmus connecting South America and
Antarctica similarly lagged behind, forming the horns of the
two continents and shedding bits of sial that now remain as
the island loop of the Southern Antilles. The alleged effects
of the westerly drift of Asia are less happily conceived. The
great oceanic deeps are supposed to represent gaping fissures
torn in the Pacific floor and not yet fully healed, while the
island festoons are strips of sial that partially lost their attach-
ment to the mainland.
i he westward movements are ascribed to the differential
attractions of the moon and the sun on the continents. Tidal
494
Wegener’s “ explanations ” unacceptable
brake on the rotating earth, and as the effect
friction acts like a
on protuberances is greater than than that on lower levels of
the crust the continents tend to lag behind. If they did lag
behind, they would appear to drift to the west. But here
again the force invoked is hopelessly inadequate to overcome
the enormous resistance that opposes actual movement. The
tidal force barely affects the earth’s rotation, and is actually
ten thousand million times too small to move continents and
raise up mountains.
In support of his presentation of the case for continental
drift Wegener marshalled an imposing collection of facts and
opinions. Some of his evidence was undeniably cogent, but
so much of his advocacy was based on speculation and special
pleading that it raised a storm of adverse criticism. Most
geologists, moreover, were reluctant to admit the possibility of
continental drift, because no recognized natural process seemed
to have the remotest chance of bringing it about. Polar
wandering, the “ flight from the poles,” and the westerly tidal
drift have all been discarded as operative factors. Neverthe-
less, the really important point is not so much to disprove
Wegener’s particular views as to decide from the relevant

evidence whether or not continental drift is a genuine variety
of earth movement. Explanations may safely be left until we
know with, greater confidence what it is that needs to be
explained. Let us, then, turn to the evidence with an unbiased
mind.
The chief criteria for continental drift are based on the
following considerations :
(a) If two continents, now far apart, were originally united,

it should be possible to detect the fact by the recognition
of certain features that were shared in common by the
separated lands, e.g. orogenic belts of which the broken
ends can be naturally joined up other details of geological
;
history as revealed in the sedimentary sequences and the ;
identity of the fossil remains of animals and plants (especially

of land and freshwater species) which could migrate freely
across the united continents but not across an intervening
ocean.
495
CONTINENTAL DRIFT
(
b) If the continents formerly occupied widely different
positions on the earth’s surface, then the distribution of climatic
zones, as inferred from geological evidence, should have
correspondingly changed.
The Opposing Lands of the Atlantic
The parallelism of the opposing shores of the Atlantic has

been a subject of discussion ever since Francis Bacon first drew
attention to it in 1620. To Wegener it suggested that the
Atlantic is an enormously widened rift with the sides still
matching “ as closely as the lines of a torn drawing would

correspond if the pieces were placed in juxtaposition.” The
fit, even of the broken edges, is far from being as perfect as
this, however. No significance can be attached to an argument

based on a geographical pattern that is little more than a
temporary accident. In Tertiary times the outlines of the
coasts were very different from those of to-day. Nor is parallel-
ism to be expected as a normal result of continental drift, for
it is mechanically impossible that the sial blocks could have
moved apart without a certain amount of rotation and a good

deal of marginal distortion due to stretching and faulting. We
know that there are patches of sial on the Atlantic floor, the
two most notable being the long S-shaped swell that traverses
the ocean floor from end to end, and the broad rise that links
Greenland to Britain by way of Iceland and the Faroes. If
all this intervening s>al were closed up again, until it became
a sheet of normal continental thickness, it would make a land

many hundreds of miles wide. Consequently, if we imagine
the Atlantic to be closed up, it is obvious that not only the
present shores but also the edges of the continental shelves
would be still separated by a distance of this order. Matching
of the geological correspondences will clearly be much less
precise tnan would be expected if the coast-lines dovetailed as
perfectly as on 'Wegener’s too closely fitting maps.
Nevertheless, the actual similarities are very remarkable.
As illustrated by Fig. 258, the transverse orogenic belts all
496
MATCHING THE OROGENIC BELTS
Fig. 258
Map to illustrate the tectonic correspondences on the opposing shores of the
Atlantic
appear to match surprisingly well. The westward convergence

of the Caledonian and Hercynian fronts towards Ireland is
continued in North America, where the fronts finally cross.
The “ fit ” is not altogether satisfactory, however, because the
times of most intense folding in the Appalachians are not the
same as in Europe. Nor is it quite certain that the two Cale-
donian fronts should be linked together as indicated. There
is another stretch of Caledonian front along
the eastern side
of Greenland, precisely where it should be to fill the gap
between the North-west Highlands and Spitzbergen. But it
should be remembered that the Iceland ridge stands badly in
497
CONTINENTAL DRIFT
the way of a former close-up, and it must not be overlooked

that the Greenland front may have been directly connected
with the loose end of Newfoundland. The British Caledonian
belt might then join up with a Caledonian belt known to
traverse the Sahara. This particular problem cannot be
settled until the geology of Greenland and the Sahara are
better known.
It is improbable that the Tertiary orogenic belts that
strike out into the Atlantic were ever adjacent, as they repre-
sent earth movements that occurred when continental drift
(assuming it for the moment) was already well advanced.
Farther south, however, there are converging Pre-Devonian
and Triassic orogenic belts in South America which can be
matched in South Africa. Here again, the crossing fore-
shadowed near the River Plate is accomplished behind Cape
Town (Plate 95 a). The distinguished South African geologist,
du Toit, has suggested that the Cape Folds are part of the same
orogenic belt as that of eastern Australia (Fig. 261).
For many years du Toit has been indefatigable in assem-
bling the evidence bearing on continental drift. In his well-
known book Our Wandering Continents he shows that a
striking series of correspondences can be recognized in the
sediments, fossds, climates, earth movements, and igneous
intrusions of the two sides of the Atlantic. Both had essentially
the same geological history during
Palaeozoic and early
Mesozoic times, and the combined evidence points very
persuasively to the high probability that they were then very
much closer together than now. Du Toit considers it possible
that the original distance between the present opposing shores
may have been as little as 250 miles. But this is a minimum
estimate and not perhaps the most probable.
The chief adverse argument is a palaeontological one.
Columns A and B of the adjoining table indicate the degree
of resemblance found between land animals that are free to
migrate from one part of a continuous continent to another.
C indicates the actual resemblance so far found between the
known fossil remains of South African and South American
Triassic reptiles. A possible reason for the obvious discrepancy
EVIDENCE : FOR AND AGAINST
A B G
Families 100 89 43
Genera . 82 64 8
Species . 65 26 0
A Percentage of recent Ohio mammals also occur-

ring in Nebraska, 500 miles away
B Percentage of recent French mammals also occur-
ring in northern China, 5,000 miles away
C Percentage of known South American Triassic
reptiles also found in the Triassic of South Africa,
now 4,750 miles apart
is that the proportion of individual land animals preserved as
so minute that the chances of finding fossils of the same
fossils is
genus in widely separated localities are slight, and of identical

species very remote. Negative evidence may be destroyed at
any moment by fresh discoveries, whereas genuine positive
evidence can never be explained away. And positive evidence
is by no means lacking. Near the top of the Carboniferous in
both South Africa and South America there is a thin band of
deltaic clay containing the bones of a small freshwater reptile
called Alesosaurus. The little animal has been found nowhere
else in the world. The region between the ancient deltas in
which he lived must have been drained by rivers, and it was
therefore occupied by land. The choice evidently lies between
accepting continental drift or postulating a giant land bridge
across what is now 4,000 miles of ocean. Here the late Car-
boniferous glaciation of Gondwanaland helps us in making a
decision.
The Climatic Zones of the Late Carboniferous
Beds of tillite (consolidated till or boulder clay), now known

tobe of late Carboniferous age, were first recognised in tropical
India in 1857, in South Australia in 1859, in South Africa in
1870, and in Brazil in 1888. As these amazing discoveries
were followed up it became unmistakably clear that Gond-
wanaland had been glaciated on a gigantic scale at a time when
Laurasia enjoyed mild or tropical climates.
The widespread Dwyka tillite of South Africa has been
499
CONTINENTAL DRIFT
partly obliterated by erosion, and is partly hidden by later

formations. But innumerable exposures still occur at intervals
from the Transvaal towards the Cape and from South-West
Africa to Natal. In many places it can be seen resting on a
glaciated floor, characteristically scored with striations.
Roches moutonnees, excavated rock basins, drumlins, and varved
clays have been discovered. The tillite itself contains grooved
and ice-faceted boulders and erratic blocks (Plate 95b), some
of which have been transported for hundreds of miles from the
north. In some localities two or three tillites are known, with
intervening interglacial deposits, showing that, as in the
Pleistocene ice ages, therewas more than one major advance
and retreat of the ice. The successive glaciations were not all
from the same centre, but migrated from west to east. The
associated deposits show that the glaciated region was one of
moderate relief, and for the most part low-lying. At the
margins the ice terminated in shallow water, marine, brackish,
or fresh, which followed up the ice as it retreated. No high
mountain range or plateau lay to the north, from which great
valley glaciers might have descended. The glaciation was the
work of a continental ice sheet that spread outwards under the
pressure of its own great thickness.
The ice came from
centres lying far to the north, and in the
latestof the glaciations from beyond Natal, outside the present
continent. Since it must have radiated not only towards the
south, but outwards in all directions, it follows that the Dwyka
tillites should be only part of a once continuous ring of such
deposits, surrounding the region of ice dispersal. Confirming
thisdeduction tillites of the same age have been found in the
north of Angola, in the eastern Congo, in Uganda, and in
Madagascar, and in the first three of these territories it has
been established that the ice moved from the south. As indi-
cated in Fig. 259, the ice sheet advanced beyond the Equator.
In India, far to the north of the Equator, similar evidence
has been found in Orissa and the Central Provinces, and still
farther north in the Punjab. The Himalayas did not then
exist, nor did any other mountain range
from which the ice
might have spread over the plains. Here again we see only a
5 00
CARBONIFEROUS GLACIATION
segment of a great ice sheet, a part in which the ice radiated

northwards, away from the present Equator. Four of the
Australian States, together with Tasmania, were also glaciated
from what is now the south. In addition to the equivalents
of the Dwyka tillites there is evidence in Australia of an earlier
glaciation about the middle of the Carboniferous, and of a
later one in the Permian, both of which were also experienced
in South America. The tillites of Dwyka age, however, are
those best represented in South America, where the ice ad-
vanced from what is now the Atlantic over parts of Brazil,
Uruguay, and Argentina, and the whole of the Falkland
Fic. 259
Map showing the distribution of the late - Carboniferous glaciation of Gond-
wanaland. The arrows indicate the directions of ice movement
Islands. Of all the fragments of Gondwanaland the only one

that has failed to furnish evidence of the late Carboniferous
glaciation Antarctica. As most of Antarctica is at present
is
shrouded by ice, this is not a matter for surprise.

Aglance at Fig. 259 shows that the glaciated lands now
occupy a considerable area of the tropics on both sides of the
Equator. With the continents in their present positions such
a distribution of ice-sheets is hopelessly inexplicable. The
suggestion that Gondwanaland rose from sea level to a plateau
so enormously high that itwas above the snow line is negatived
by ample evidence that it was nowhere very high. But whether
itwas or not, the tropics could not have been glaciated down
( 396) 501 33
CONTINENTAL DRIFT
to sea level without the development of still greater ice sheets
over the northern lands. The
only evidence of Carboniferous
glaciation in the north is found in Alaska, which has probably
never been far from the North Pole, and near Boston, in the
Appalachian orogenic belt, which at that time may well have
been a high mountain range. On the other hand the great
Carboniferous coal forests were flourishing from North America
Fig. 260
Map showing the distribution of the late-Carboniferous glaciation of Gond-
wanaland, with the continents reassembled, though not so closely, as interpreted
by Wegener
to China while Gondwanaland lay under ice. Moreover,

and bauxite that could only have formed in
deposits of laterite
a tropical climate are found in the Upper Carboniferous of
the United States (Kentucky and Ohio), Scotland (Ayr-
shire), Germany, Russia (south of the Moscow basin), and
China (Shantung). The inference that the equatorial zone
of the time is roughly indicated by this lateritic belt is irre-
sistible.
502
ATTEMPTED REASSEMBLY OF THE CONTINENTS
No amount of polar wandering, even if it could be admitted,
would give a distribution of climatic girdles around the globe
corresponding to the picture outlined above. Wherever the
South Pole is imagined to have been in order to account for
any one of the glaciated regions, it would still have been too
distant from the others to account for more than one of them.
The problem, indeed, remains an insoluble enigma, unless
the straightforward inference is accepted that all the continents
except Antarctica lay well to the south of their present positions,
Fig. 261
Map showing the distribution of the late-Carboniferous glaciation of Gond-
wanaland, with the continents reassembled according to du Toit’s interpretation
and that the southern continents were grouped together around

the South Pole. In attempting a circumpolar reassembly
the position to be allotted to Antarctica is necessarily un-
certain. Wegener places it between Australia and South
America (Fig. 2601 whereas du Toit, guided by meagre
;
stratigraphicai and tectonic clues, thinks it may have been

between Australia and Africa (Fig. 261). With either arrange-
ment the ice sheets all fall within an area comparable with that
glaciated in the Northern Hemisphere during the Pleistocene.
Moreover, as indicated in Fig. 260, the lateritic belt then comes
into line with the Equator of the time, and other known details
5°3
CONTINENTAL DRIFT
of the Carboniferous climatic girdles also fall consistently into

their appropriate places. The site of the Hawaiian Islands
would have been approximately over the North Pole at this
time. Consequently, no evidence of a North Polar ice cap is
to be expected. The nearest of the present land areas where
signs of glaciation might reasonably be looked for are Cali-
fornia and Alaska. The Carboniferous rocks of California
are marine sediments, where again no evidence could be ex-
pected. But in Alaska a late Carboniferous tillite occurs, just
where it ought to be.
The only serious argument advanced against the validity
of the above solution is that it merely exchanges one embarrass-
—
ing problem for another the difficulty of explaining how
continental drift on so stupendous a scale could have been
brought about. By itself this consideration might be a reason
for sitting on the fence, but the real antithesis is not so simple.
If one rejects continental drift and accepts the possibility that
Central Africa could have been glaciated while Britain had
a tropical climate, one must also admit the necessity for land
bridges, which have since subsided to oceanic depths. The
continental drift solution has the advantage that it reduces two
baffling problems to one, while at the same time it removes
many other less intractable difficulties.
Before leaving the subject of climates, it is of interest to
notice that South Africa was glaciated several times before the
Carboniferous. A
widely distributed tillite occurs in strata of
Lower Devonian age in the Cape Province. In the late Pre-
Cambrian, glaciation occurred on a scale comparable with that
of the late Carboniferous— the regions affected including the
Transvaal, Rhodesia, the Congo, Angola, and South-West
Africa. Still earlier in the Pre-Cambrian yet another tillite is
preserved in the Transvaal, and farther back still South-West
Africa provides evidence of what may be one of the earliest
glaciations known. It thus appears that for at least
1,000
million years the position of Africa relative to the
South Pole
did not significantly alter. Africa gives us no evidence of
having drifted from a situation far to the south until com-
paratively late in its geological history.
504
)
THE MACHINERY OF CONTINENTAL DRIFT
The Search for a Mechanism

It has been shown that in looking for a possible means of
“ engineering ” continental drift we must confine ourselves to
processes operating within the earth. To be appropriate, the
process must be capable (a) of disrupting the ancestral Gond-
wanaland into gigantic fragments, and of carrying the latter
radially outwards as indicated in Fig. 210 Africa and India
:
towards the Tethys Australasia, Antarctica, and South

;
America out into the Pacific ; ( b of disrupting Laurasia,

though much less drastically, and again with radially outward
movements towards the Tethys and the Pacific, as indicated
in Fig. 209. We have already seen that the peripheral orogenic
belts probably mark the regions where opposing systems of
sub-crustal currents came together and turned downwards.
The movements required to account for the mountain struc-
tures are in the same directions as those required for continental
drift, and it thus appears that the sub-crustal convection
currents discussed on pages 408 to 413 may provide the sort
of mechanism for which we are looking (Fig. 262).
To explain the peripheral orogenic belts three systems of
convection currents are called for (or three co-ordinated groups
of systems), with their ascending centres situated beneath
Gondwanaland, Laurasia, and the Pacific respectively. In-
cidentally, should be noticed that the coalescence of the usual
it
chaotic or small convective systems into three gigantic ones

involves a coincidence that can rarely have happened in the
earth’s history, and one that is just as likely to have come about
during the Mesozoic era as at any other time. The often-asked
question : How is it that Pangaea did not begin to break up
and unfold until Mesozoic time ? thus ceases to have any
significance. If continental drift could have been caused by
the gravitational forces invoked by Wegener, then it should
have occurred once and for all very early in the earth’s history,
since those forces have always been in operation. If convection
currents are necessary, continental drift may have accom-
panied all the greater paroxysms of mountain building in
former ages but, if so, it would usually have been on no more
505
CONTINENTAL DRIFT
than a limited scale. That there was a quite exceptional in-

tegration of effort in Mesozoic and Tertiary times is forcibly
suggested by eruptions of plateau basalts and building of
mountains on a scale for which it would be hard to find a
parallel in any earlier age.
There are, therefore, good reasons for supposing that at this
critical period of the earth’s history the convective circulations

became unusually powerful and well organised. Currents
”
Diagrams to a purely hypothetical mechanism for “ engineering
illustrate
continental drift. In A sub-crustal currents are in the early part of the con-
vection cycle (Stage I ot Fig. 215). In B the currents have become sufficiently
vigorous (Stage 2 of Fig. 215) to drag the two halves of the original continent
apart, with consequent mountain building in front where the currents are
descending, and ocean floor development on the site of the gap, where the
currents are ascending
flowing horizontally beneath the crust would inevitably carry

the continents along with them, provided that the enormous
frontal resistance could be overcome. The obstruction that
stands in the way of continental advance is the basaltic layer,
and obviously for advance to be possible the basaltic rocks
must be continuously moved out of the way. In other words,
they must founder into the depths, since there can be nowhere
else for them to go (Fig. 262).
Now this is precisely what would be most likely to happen
when two opposing currents come together and turn down-
SUPPOSED EFFECTS OF CONVECTION CURRENTS
wards beneath a cover of basaltic composition. The latter

then suffers intense compression, and like the sial in similar
circumstances it is eventually drawn in to form roots {cf. Figs.
215 and 216). On the ocean floor the expression of such a
down-turning of the basaltic layer would be an oceanic deep.
The great deeps bordering the island festoons of Asia and the
Australasian arc (Tonga and Kermadec) probably represent
the case where the sialic edge of a continent has turned down
to form the inner flanks of a root, while the oceanic floor con-
tributes the outer flanks.
It is not difficult to see that a purely basaltic root must have
a very different history from one composed of sial. The
density of sial is not significantly increased by compression.
Consequently, when a sialic root is no longer being forcibly
held down, it begins to rise in response to isostasy, heaving up
a mountain range as it does so. But when rocks like basalt or
gabbro (density 2-9 or 3-0) are subjected to intense dynamic
metamorphism they are transformed into schists and granulites
and finally into a highly compressed type of rock called eclogite,
the density of which is about 3-4. Since this change is known
to have happened to certain masses of basaltic rocks that have
been involved in the stresses of mountain building, it may safely
be inferred that basaltic roots would undergo a similar meta-
morphism into eclogite. Such roots could not, of course, exert
any buoyancy, and for this reason it is impossible that tectonic
mountains could ever arise from the ocean floor. On the con-
trary, a heavy root formed of eclogite would continue to
develop downwards until it merged into and became part of
the descending current, so gradually sinking out of the way,
and providing room for the crust on either side to be drawn
inwards by the horizontal currents beneath them (Fig. 262).
The eclogite that founders into the depths will gradually
be heated up as it shares in the convective circulation. By the
time it reaches the bottom of the substratum it will have begun
to fuse, so forming pockets of magma which, being of low den-
sity, must sooner or later rise to the top. Thus an adequate
source is provided for the unprecedented flows of plateau basalt
that broke through the continents during Jurassic and Tertiary
5° 7
CONTINENTAL DRIFT
times. Most of the basaltic magma, however, would naturally

rise with the ascending currents of the main convectional
systems until it reached the torn and outstretched crust of the
disruptive basins left behind the advancing continents or in
the heart of the Pacific. There it would escape through in-
numerable fissures, spreading out as sheet-like intrusions within
the crust, and as submarine lava flows over its surface. Thus,
in a general way, it is possible to understand how the gaps rent
in the crust come to be healed again and healed, moreover,
;
with exactly the right sort of material to restore the basaltic

layer. To sum up during large-scale convective circulation
:
the basaltic layer becomes a kind of endless travelling belt on

the top of which a continent can be carried along, until it
comes to rest (relative to the belt) when its advancing front
reaches the place where the belt turns downwards and dis-
appears into the earth.
To go beyond the above indication that a mechanism
for continental drift by no means inconceivable would at
is
present be unwise. Many

serious difficulties still remain
unsolved. In particular, it must not be overlooked that a
successful process must also provide for a general drift of the
crust over the interior a drift with a northerly component
:
on the African side sufficient to carry Africa over the Equator,

and Britain from the late Carboniferous tropics to its present
position. The northward push of Africa and India, of which
the Alpine system and the high plateau of Tibet are spectacular
witnesses, could not have been sufficient by itself to shove
Europe and Asia so far to the north. To achieve this the aid
of exceptionally powerful sub-Laurasian currents directed
towards the Pacific is required. The total northward com-
ponents might then overbalance the southward components,
and a general drift of the crust would be superimposed on the
normal radial directions of drift.
It must be clearly however, that purely speculative
realised,
ideas of this kind, specially invented to
match the requirements,
can have no scientific value until they acquire support from
independent evidence. The detailed complexity of convection
systems, and the endless variety of their interactions and
5°8
NECESSITY FOR CAUTION
kaleidoscopic transformations, are so incalculable that many

generations of work, and mathe-
geological, experimental,
matical, may well be necessary before the hypothesis can be
adequately tested. Meanwhile it would be futile to indulge
in the early expectation of an all-embracing theory which
would satisfactorily correlate all the varied phenomena for
which the earth’s internal behaviour is responsible. The
words of John Woodward, written in 1695 about ore deposits,
are equally applicable to-day in relation to continental drift
and convection currents “ Here,” he declared, “ is such a
:
vast variety of phenomena and these many of them so delusive,

that ’tis very hard to escape imposition and mistake.”

A. Wegener
The Origin of Continents and Oceans. Methuen, London, 1924.
A. L. du Toit
Our Wandering Continents. Oliver and Boyd, Edinburgh, 1937.
T. H. Holland
The Evolution of Continents : A Possible Reconciliation of Conflicting
Evidence. Proceedings of the Royal Society of Edinburgh, Vol. LXI.,
Part II., No. 13, 1941.
509
1 9
INDEX
Aa lava, 448 Alpine geosyncline, 383, Fig. 207
Aar massif, 396, Fig. 205, Pi. 51 — mountain system, 382-83, 386-87
Abrasion, glacial, 215 Fig. 200 dislocation of, 419-20
— wind, 254, 256-58 — ;
orogenesis, 108, 109, 390-401

Absorption of drainage areas, 1 77 — revolution, 397, Figs. 208-10
Abyssal deposits, 316-20 Alps, glaciation, 216, 245, 247-49,
— zone of deposition, 314, Fig. 171 Fig. 133
Abyssinian section of rift valley, 434, — terminal moraines, 229, Fig. 1
Fig. 227 — Western, 390-97, Figs. 203-7

1
Adobe, 269 Alum Pot, PI. 26

.Egean Archipelago, 420 Amazon drainage basin, 143
Aiolian deposits, 12 1. 259-65, Figs. Amethyst, 49
I37-4E Pls 5- 58-61
.
-
Ammonites, Fig. 50
Africa, Carboniferous glaciation, 499- Amorphous substances, 37
500, Figs. 259-61, PI. 95 Arnphibolcs, 41-42
— crossing of orogenic belts, 498, Amygdales, 49
Fig.
258, 95 PI. Anak Krakatao, 473, PI. 91
— escarpments, 192-93, 87 PI. Andes, uplift of, 419
— inselberg landscape, 274-75, Fig. Andesite, 5i 52, 453, Fig. 15
:
Anglo-Egyptian Sudan, 430, Fig. 223
— H4 Animals, geological work of, 22, 115
— Pleistocene
Pre-Carboniferous
246 glaciation,
504 glaciation, Anomalies of gravity, 404-6, Figs.
— rift vallevs,
432-41, 225-32, Figs. 21 -13, 440, Fig. 232
1
— 86-87pattern, 428-38, Anorthite, 39

Pis.
tectonic 233 Fig. Antarctica, 207, 246, Fig. 105
— volcanoes, 435-37, 476, 87-88 Pis. — Beardmore Glacier, 209
Agassiz, Lake, 243 — continental drift, 501, 503, Figs.
Agassiz, Louis Jean (1807-1873), on 260-61
Pleistocene glaciation, 216 — volcanoes, 476
Agate, 49 Antecedent drainage, 200-1
Age of Earth, 26, 105, 109 Anthracite, 336, 338-39, Fig. 178
Age of rocks, absolute, by radioactive Anticlines, 72, Figs. 25, 28, Pis. 12, 80
minerals, 103, 105, Figs. 51-52 — as oil traps, 349, Fig. 18
— relative, by fossils, 99, 101 Anticlinorium, 74, Fig. 26
Agglomerate, 451 Antipodal relation of land and sea, 18
Aiguilles Rouges massif, 386, 395, Antrim, dyke, PI. 17
Figs. 204-5 — Giant’s Causeway, Pis. 3, 14
Ailsa Craig, erratics from, 227 — plateau basalts, 458, Fig. 243
Alaska, Carboniferous glaciation, 502, —
— pot-hole, PI. 34
504 waterfalls, 158
— Chisana Glacier, PI. 5 Appalachian folding, 387, Fig. 202
— Malaspina Glacier, 209 — geosyncline, 379
— uplift, 27, 359 — zone, 159
fall
Albert, Lake, 434. 436-37, 439, Figs. Appalachians, 387-90, Figs. 201-2
226, 230-32 — crossing of orogenic belts, 387, 390,
Albite, 39 Figs. 201, 258
Aide, River, deflection of mouth, 299, — orogenic revolution, 109, 387
——
-
Fig. 161 rivers, 184
Alkali felspars, 39 thick strata, 379, 381
—
•
Alluvial fans, 201-2, Fig. 104, 271 superimposed drainage, 184

— deposits, 167, 201, 272, Pis. 37. 40 Aquifei , 132-34, Figs. 56-58
— terraces, 195-96? Pig* IOO > Pi* 43 Archaean, 104
Alluvium, 12 1, 167, 201, 272 Arches, natural, 135, Fig. 60 198,
—
;
Hwang Ho, Fig. 79 Fig. 102, PI. 44 ; 287, PI. 68
51 1
INDEX
Arcuate delta, 170, Fig. 134 Badland type of erosion, 268, PI. 63
Areas, earth, land, and sea, 11, Fig. 2 Bagnold, R. A., on desert sand-forms,
Arete, 220, Fig. r 14 262
Arid climate, weathering, 114 — on singing sands, 266
— cycle of erosion, 273-74 Baikal, Lake, 244, 428, Figs. 222, 225
Aristotle (384-322 b.c.), erroneous Bailey, E. B., on Appalachian struc-
views on circulation of water, 126 ture, 390
Arran, Tertiary dykes, 83 Baltic Shield, 384-85, Fig. 200
Artesian wells, 132, Fig. 56 Banda arc, 398, 403, Figs. 211-12
Arun River, Himalayas, 201 —• volcanoes, 473
Ash and cinder cones, 452-53 Banks, submarine, Fig. 176
Ash, of coal, 341 Bar, 299, Figs. 162-63
— volcanic, 444, 450, 470-1, Fig. 237 —connecting, 299
Asia, tectonic pattern, 425-28, Fig. 222 —offshore, 304-6, Fig. 167
Asphalt, 343-44, 351-53 Barbados, abyssal deposits, 320
Assam earthquake, 361 Barchan, 262, Figs. 138-39
Atlantic, bordering lands, 496-8, Fig. Barrier beach, 304-6, Fig. 167
— 322, 325, Figs. 172-73,
— 258
reefs,
borings through 175-76, Ph 73
— coastal type, 301, 400,319-20
floor,
475, 477, — spits, 299, Figs. 160-61
Fig. 165 Basalt, abundance 89 of,
— continental 494, 496-98,
drift, — columnar structure, 47, 76
— faulted256-58
Figs. 45°, Fig. 29, 14 Pis. 3,
margins, Figs. 209-10 — distribution. Fig. 252
— former land areas,400,488 — early views on origin, 48-49
— structure of 373, 496, Fig.
crust, — magma, suggested origin, 484-85,
194 507
— volcanoes, 476, 252 Fig.
.
— overlying Chalk, 17 PI.

Atmosphere, 8 — plateau,
— changes composition, 251
in — spheroidal458-60, Fig. 243
weathering, 118-19,
— circulation of water, Fig. 7 Pis. 24-25
Atolls, Figs. PI. Basaltic layer, 371-73, Figs. 193-94
— 322, 325-29,174, 176,175 73
origin, Fig. — heat flow and temperature, 481,
Atomic structure of crystals, Fig. 254
— relation cleavage, 40 36, 37
to —relation to continental drift, 506-
Attrition, river,
15 508, Fig. 262
— wind, 251, 258-59, Fig. 136 —relation to vulcanism, 478-85,
— marine, 286, 294 Figs. 253, 255
Aughrabies Falls, 159, Fig. 69 Base-level, 153, Fig. 65
Augite, 42, 52, Fig. 14 —arid regions, 256
Aureole of contact metamorphism, —local, 155-56, Figs. 66-67
62^63, Figs. 20, 45 Basin, London, Fig. 57
Australia, artesian basin, 133 Basin Ranges, 423, Figs. 219, 221
— Carboniferous glaciation, 499, 501, Basins, African, 428-32, Figs. 223, 229
Figs. 259-61 — artesian, 132-33, Fig. 57
— Great Barrier Reef, 322, Fig. 173, — Asiatic, 426-28, Fig. 222
PI. 73 — desert, 273
Axenstrasse, Lake Lucerne, recumbent —
— glaciated, 223-26, Fig. 118, PI. 54
folds, PI. 81 oceanic, 400
Axial planes of folds, 73, Fig. 25 — tectonic, 73, 414-15
—
-
in orogenic belts, 381-82
“
— wind excavated, 256, Fig. 135
Axis of fold, 73, Figs. 25, 27 Basket of eggs ” topography, Fig. 120
Batholiths, 89, Fig.
Bacteria, 8 — emplacement, 90-9245
— biochemical activities, 115, 31 I— — volcanic break-through, 455
3 r 3> 332, 347 Bathyal deposits,
317
— role in weathering, 1 ib
.
— zone of deposition, 314, Fig. 171
Badlands, 1 16, 146, Pis. 32, 63 Bauxite, 41, 1 18, 120
512
INDEX
Bauxite, Carboniferous, 502 Boulders, erratic, 227,Pis. 4, 55
Beach, 292 —weathering residuals, PI. 25
— barrier, 304, 306, Fig. 167 Brachiopods, Fig. 50
— Chesil, 300, Fig. 164, PI. 71 Braided streams, 167, 171, PI. 40
— drifting, 295-96, Fig. '5 8 . PI- 7° Brazil, Carboniferous glaciation, 499,
—
—
profile, 293, Fig. 157
—
F"g- 259
Sugar Loaf, Rio de Janeiro, PI. 65
raised, 27, 33, PI. 6
Beardmore Glacier, 209 —waterfalls, 159
Bed (stratum), 54 Bread-crust bombs, 451, Fig. 238
Bedding, 54, Figs. 23-24, Pis. 1, 8, Breakers, 285, PI. 66
42, 68 Breccia, 54, PI. 17
— plane, Fig. 23 — fault, 79
— traces in slate, Fig. 19 Bridal Veil Falls, PI. 52
Bedrock, 122 Bridges, natural, 135, Fig. 60 198, ;
Belemnite, Fig. 50 Fig. 102, PI. 44 287, PI. 68

;
Bell, H. S., on suspension currents, British Columbia, fjords, Fig. 117

310, Figs. 169-70 — river terraces. PI. 43
Belts, negative anomalies of gravity, Brito-Arctic basaltic region, 458, Fig.
Figs. 211-13 252
— 404-406,
orogenic, 377 Broads of East Anglia, 194-95, 299
Benthos, 314 Brown coal, 335
Bergello massif, PI. 83 Bryce Canyon, 422, PI. 32
Bergschrund, 211-12, Figs. 108-109 Bufumbira volcanoes, PI. 87
— relation to corrie growth, 219 Bullard, E. C. on rift valleys, 440-41,
Bilateral mountain systems, 382, Figs. Figs. 232-33
198-99 Buried forests, 28, 196, 301
.
Biosphere, 8 Burma, delta, PI.

75
Biotite, 41, 45, 51, 52 — wood, 141
fossil
Bird’s foot delta, 171, Fig.80 — oilfield,
352, 77 PI.
Birunga volcanoes, 436, Fig. 236, Pl.87 Burman arc, Figs. 211-12
Bitumen, 343-45, 352 — volcanoes,398,473403,
Bituminous coal, 335-38, Fig. 178 Bushveld, 87, 431, Fig. 42
Black Forest massif, 415, 419, Fig. 218, Butte, 182, Fig. 89
PI. 82
Black Sea, oil-bearing muds, 347 Cainozoic Era. 104
— recent inbreaks, 420 —igneous activity, 108
— salinity currents, 281, Fig. 146 Calcareous deposits, 314, 316-25
Block faulting, 415, Figs. 217, 221 Calcite, 43, 54, 58-59, 141-42
— lava 448-49 Calcium carbonate, 56, 59
— mountains, 415, Figs. 217-18,221, —deposition, 135-36, 138, 312
PI. 84 —replacement, 141
Blow-hole, 287, PI. 67 —solubility, 134, 316
Boghead coal, 336, 341 Calderas, 445, 454-55, Fig. 240
Bombs, volcanic, 450-51, Figs. 237-38 —Crater Lake. PI. 94
Bone beds, 313 —Krakatao, 472, Figs. 250-51
Bonneville, Lake, 425 —Mt. Meru, PI. 88
Bore, 280 —relation to cauldron subsidence, 90
Borrowdale, boulder clay, PI. 55 Caledonian orogenesis, 108-9, Fig. 27
— lacustrine flats, PI. 37 —orogenic belt, 99, 385-87, Fig. 200
— Volcanic Series, io8,38i,P1s.22,37 —igneous activity. Figs. 51-52
Boss, 90 —Lake District, 182, Figs. 91-92
Bottomset beds, 169, Fig. 78 — volcanic activity,
381
Bouguer, Pierre (1698-1758), on gravi- California, desert, 60, PI.
tational attraction of Andes, 16 — earthquakes and 5 faults, 360, 363,
Boulder clay, 226-27, PI- 55 Fig.
— earth pillars, 147, PI. 33 — lake 184 63
terrace, PI.
— Carboniferous, 499 — salt marsh, PI. 85
— valley in, PI. 35 Cambrian period, 104
513
INDEX
Canadian Falls (Niagara), 158, PI. 38 Changes of sea level, 107, PI. 6
Cannel coal, 336, 338, 341 — resulting from glaciation, 194, 238,
Canyon, Grand, 16 1, 199-200, Figs. 70, 249, 32 7, 4!7-. !8
103. Pis. 1, 84 Charnian orogenesis, 108-9
Canyons, submarine, 306-10, Fig. 168 Chemical weathering, 112, 116-25
Cape Ranges, South Africa, 498, Fig. — in desert, 269
258, Pis. 30, 95 Chernosem, 123-25
Carbonate rocks, 43, 55-57 Chert. 142
— metamorphic, 57-59 Chesil Beach, 300-1, Fig. 164, PI. 7 1
Carbon, compounds, 43 Chiltern Hills, 180, Figs. 48, 57
— in fuels,
331 China, 427, Fig. 222
Carbon dioxide, 43. 56, 59 — loess, 267-68, Fig. 79, PI. 63
— chemical weathering, 1 1 Chisana Glacier, Alaska, PI. 5
— climatic effects,
251 Chlorite, 43
— source of and oxsgcn,
fuels 311, Chronology, geological, io5,Figs. 51-52
330, Fig. 177 — ice recession, 240, Fig. 127
— volcanic,
259- 447-48 — interglacial stages, 247-49, Fig. 133
Carboniferous, coalfields, 335, 339-43, Cinder cone, 452-53
260-
Fig. 180 Circum-Pacific, earthquake zones, 365,
— 339, Fig. 179,
forests, 74 PI. Fig.
187
— fossils, Fig. PI. — orogenic belts, Fig. 208
50, 9
— glaciation, 250, 499-504, Figs. — volcanoes, 473, 109, 398,
252 Fig.
61, 95 PI. Circulation of meteoric water, 22-23,
— palaeogeography, 256-57, Figs. 127, Fig. 7
61 Circumference of Earth, 1
— sandstone, 8 PI. Cirque, 209
Carboniferous Limestone, crinoidal, Clarain, 337-38
PI. 9 Clarke, Frank Wigglesworth (1847-
— Denbigh, PI. 42 1931), on average composition of
— Inglebcrough, Fig. 47 rocks, 35
— Malham Cote, Pis. 8, 24 Clay, 54
— unconformable base, PI. 21 — porosity, 128-29
Caribbean arc, 399, 405, Fig. 213 — relation to laterite, 120, Fig. 53
Carlsbad Cavern, New Mexico, 135, Clay minerals, 41, 54, 60, 117
PI. 27 Cleavage, crystal, 40-42
Garrick Castle, recumbent fold. Fig. 27 — slaty,60, Fig. ig
Castle Rock, Edinburgh, 218, Fig. 12 1 — fan,75, Fig.26
Catchment area, 129. Fig. 56 Cliffs,development, 287
— Sahara ground-water, 1 33 —
Caucasus, glaciated rock basin, Fig. 18 1 — erosion,
recession
285
of,
291, Figs.
154-55,
Cauldron subsidence, 90 Pis.
66, 69
Caverns, limestone, 134-37, PI. 27 — scenery, 287, 67-69
Pis. 15,
Caves, lava, 449 Climate and cycle of erosion, 194
— sea, 287, PI. 67 —
— ulpifted, 27 — landforms, 122-24
soil tvpes,
145
Cementation of sediments, 54, 127, 141 — weathering, 113, 118-19

Central volcanic eruptions, 28, 444, Climatic changes, causes 249-51 of,
460-62, Fig. 244, PI. 2 — Carboniferous, 339, 499-504
Chad Basin, 429, Fig. 223 — future, 249
Chalk, 55 — Pleistocene, 247-49, Fig. 133
— cut by dyke, PI. 1 — Thames terraces, 196
— 141
flint in, Cloudbursts in desert, 271
— London Basin, 132, Fig. 57 Coal, 334-39. PI. 76
— Lulworth Cove, 70 PI. — seams, 339-41
— Needles,
. of Wight, Fig.
Isle 151 Coalfields, 335, 342-43, Fig. 180
— not an abvssal deposit, 320 Coal Measures, 102, 339, Fig. 180
— Sussex 6g
cliffs, PI. — faulted, PI. 15
Chalybite, 43 — folded, PI. 12
INDEX
Coastal, dunes, 260-61 Continents, 10

— plain. North America, Fig. 201 — origin unknown, 21
Coast erosion, 286-94, Fig. 155, Pis. — sialic structure, 19, Fig.
4
66-70 — tetrahedral Fig.
distribution, 18, 5
Coasts, Atlantic type, 301, 400, 473, Contraction hypothesis, 407-8
477. F 'g s 165', 258 — joints, 76, 77, Fig.
29
—
-
emergence, 277, 304-6 Convection current hypothesis, 408-

— Pacific type, 301, 401, Figs. 166, 413, Figs. 215-16, 262 ; 484-86,
208 5°5A), Fig. 262
— submergence, 277, 301-4, Fig. 165 Convection currents, 20, Fig. 6
Col, 175, Fig. 83 Cooling of Earth, 407-9
Cole, Grenville Arthur James (1859- Coral reefs, 321-29, Figs. 172-76,
1924I, on soil, 122 PI- 73
Colluvial deposits, 121 — sands, 57
Colorado, Desert, Pis. 5, 60, 63 — uplifted reefs, 403
— Grand Canyon, 161, 199-200, Corals, fossil, Fig. 50
— reef-building, 322-23, PI. 73
— Plateau, 199, 421-22,84
Figs. 70, 103, Pis. 1,
Figs. 201, Cordillera, 378
219 — North American, 420-25, Figs.
— River, 199, 422 201, 219
Columbia Plateau, 421, 201 Fig. Corrasion, fluvial, 150
— — marine, 286
— plateau
basalts, Fig.
252
River terraces, PI. 43 Corrie glaciers, 209
Columnar structure, 47, 76-78, 450, — tarns, 219
Fig. 29, Pis. 3, 14 Corries, 209, Figs. 108-9
Composite volcanic cones, 452-54, — confluent, Fig. 116
Fig. 234 — origin, 218, Fig. 113
Composition, peat and coal, Fig. 178 Corrosion, marine, 286
— rocks, 35, 38, Fig. 15 — river, 150
— sea water, 281 Cotswold Hills, 55, 180, Fig. 48
Compressional movements, isoclinal Country rocks, 71
folding, 73, Fig. 26 Cowal anticline, 74, Fig. 27
— slates, 61 Crag and Tail, 217-18, Fig. 112
— thrust faults, 80, 381-82, 385, Crater Lake, Oregon, 454~55> PI- 94
Figs. 198-99 Craters, 444, Figs. 234, 247, Pis. 88,
Concretions, 141 90-92
Concordant intrusions, 83, 84, Figs. 37, Crescentic dunes, 262-63, Figs. 138-139
41, 42 Crest of fold, 73, Fig. 25
Cones, alluvial, 201, Fig. 104 Cretaceous, coalfields, 335
— volcanic, 452754, Fig. 234 — fossils, Fig.50
Cone-sheets, 87, Fig. 43 —
-
oilfields, 346
Conglomerate, 54, PI. 21 — period, 104
Congo Basin, 143, 430, Fig. 223 — strata, Fig. 57
Connate water, 127 Crevasse infillings, 233, 235
Connemara marble, 59 Crevasses, 211-12, Figs. 107-9, PI- 47
Consequent streams, 144, 173, 179, Cross bedding, 95, Figs. 46, 96, 137,
Figs. 84, 85, 87 PI. 20
Contact aureoles, Figs. 20, 45 Cross-profile of river valleys, 161, Fig.
— metamorphism. 62 7°
Continental crustal layers, 371-72, — modified by glaciation, 220, Fig.
Figs. "4
— J93-94
deposits, 121 — modified bv rejuvenation, 195
— drift, 487-509 Crossing of orogenic belts, 387, 390,
— ice sheets,206-8, Fig. 105 497-98, Figs. 200-1. 258, PL 95
—
—
plattorm, to, Figs. 2, 171 Crust of Earth, 9, 65, Fig. 4
— composition, 35
shelf, 11, Figs. 2, 171
— slope, 10, 306-8, Figs. 2, 168, — layered structure, 371-73, Figs.
I7i I93'94
INDEX
Crystalline limestone, 58-60 Dendritic stream pattern, 174, Fig. 81

— schists, 63-64 Density, average of Earth, 1
— of rocks, 9
Crystals, 36, Figs. 10-11
Cuesta, 181-82, Fig. 89 Denudation, 24, 31, 201
Cuillin Hills, Skye, Pis. 18, 53 Deposition, 24, 31
Culbin Sands, 261, PI. 58 Deposits, aeolian (wind), 121, 254-55,
Culmination, tectonic, 392, Fig. 205 259-65, Figs. 135, I37-4 1 , pls -
Cumulose deposits, 12 58-60

Current bedding, 95, Figs. 46, 96, 137, — continental, I2t
PI. 20 — deltaic, 169, Fig. 78
Currents, convection, 20, Fig. 6 — glacial, 226-32, Figs. 118-21,
— 281,
salinity, 146 Fig. Pis. 55, 57, 66
— suspension, 308-10, Figs. 169-70 —
- glaciofluvial, 226, 232-35, 238,
— tidal, 2 80 Figs. 122, 124, PI. 57
-
Cycle, metamorphic, 65-66, Fig. 21 — ground-water, 135-36, 140-42,

— orogenic, 109, 401-3, 410-11, PL 27
— hot-spring,
Fig. 215 138, 28 PI.
Cycle of erosion, arid, 273-74 — loess, 255, 267-69, Figs. 79, 143,
— normal, 145, 185-89, Figs. 94-95 ; 63PI.
— marine, Figs. 153,
interruption of, 145, 155, 194 294-95, 3'5-27,
159-64, 171-74, 69-73 Pis.
Dacite, 51, 52, Fig. 15 — pelagic, 316-20, 72 PI.
Dalmatia, Pacific type of coast, 301, — 24, 151-52, 167, Fig. 77,
river,
Fig. 166 Pis. 35, 37, 40
Daly, R. A., on origin of coral reefs, Depressions, desert deflation, 256, 273,
326-29 Fig- 135
— on origin of submarine canyons, —
geosynclinal, 379-81
—
308-10 tectonic, 392, Fig. 205
Dana, James Dwight (1813-1895), on - rift-valley, 432, Fig. 225
geosynclines, 380 Desert, dust storms, 253, Fig. 134
— on permanence of continents, 488 —
oases, 256, Fig. 58, PI. 61
—
Darwin, Charles (1809-1882), on coral
— pavement,
rainstorms,
258
reefs, 322, 325, Fig. 175
Davis, William Morris (1850-1934), on — sands and 271dunes, 259, 262-67,
the cycle of erosion, 185 60
Pis. 5,
Day, A. L., study of Halemaumau, 463 — surfaces,
255, 61 PI.
Dead Sea, 434, 436, 439, Figs. 225, 228 —varnish, 270
De Geer, Baron Gerard Jacob (1859- Deserts, 255-59, 262-74, Figs. 135,
1943)) on varve-chronology, 238- 1 41, Pis. 5, 60-64
240, Fig. 127 Desmarest, Nicolas (1725-1805), on
Death Valley, California, Fig. 184 origin of columnar basalt, 47
— floor of, 423 Devonian period, 104
— salina, PI. 85 Diabase, 50, 52
— sand dunes, PI. 60 Diatom ooze, 318
Debris slides, 148 Diatoms, 312, 315-16
Deepening of valleys^ 152-53, 195 Dinaric Alps, 397, Fig. 203
— glacial, 220-6, Figs. 114, 116, Pis. Diopside, 42, 118
51-53 Diorite, 51-52, Fig. 15
Deep Sea, deposits. 316 Dip of strata, 71, Figs. 23-24
— platform, 10, Figs. 2, 171 —effect of soil creep, 147
—
- oozes, 312, 316-20, PI. 72 Dip slope. Figs. 88-89
Deeps, oceanic, 11, 12, Fig. 2 Discordant intrusions, 82
— views as to origin, 494, 507 Disintegration, atomic, 103
Deflation, 254, 256, Fig. 135 —weathering, 1 1 2- 1
Deforestation, 116 Dismal Swamp, Fig. 167
Delaware Water Gap, 184. Fig. 93 Dismembered rivers, 194
Deltaic deposits, 169, Fig. 78 Distributaries, 170, 201
Deltas, 169-72, Figs. 78-80, 135, PI. 75 Dolerite, 50, 52, 83, Pis. 17-18
516
1 1
INDEX
Dolerite, spheroidal weathering, 1 1 8, Eastern Alps, 391, 393, 397, Fig. 203
24-25
Pis. Eastern Rift Valley of Africa, 434-35,
Dolomite, 43, 57 438, Figs. 226-27, PI. 86
— metamorphism, 59 Echo sounding, 306-7
Domes, salt, 349-50, Fig. 181 Eclogite, 507
— tectonic, 73, 183, Fig. 92 Edinburgh, crag and tail, 218, Fig. 1 12
— volcanic, 456-57, 462, 469, Figs. Egyptian Desert, 256, Fig. 135
241-42, 249 — sand sea, 265, Fig. 141
Downthrow of fault, 79 — wadis, PI. 62
Drainage, basins, 143-44, 176-79 Elements, 1, 35
— internal, 143, 270, 414, 424, 438 — radioactive, 103, 409, 480-81
—
—
patterns, 173-75, Figs. 81, 85 Emanations, magmatic, 446-48, 485
superimposed, 182-84 Emergence, coasts of, 277, 304-6
Drakensberg, Natal, 192, 430, PI. 87 Emplacement of batholiths, 90-2
Dreikanter, 258, Fig. 136 End or terminal moraines, 214, 228,
Drift, continental, 487-509 Figs. 118-19, 122, PI. 3
— glacial and glaciofluvial, 226 Englacial moraine, 213
Drumlins, 230-32, Figs. 120-22 Entrenched meanders, 198, Figs. 10 1-
Dunes, coastal, 260-6 1, 293 102, PI. 44
— desert, 262-67, Figs. 138-41, Pis. Epeiric seas, 12
5>.6° Epeirogenic movements, 107, 378,
— migration and structure, Fig. 137
.
414-41
Durain, 337-38, pl 7 6 - Epicentre of earthquake, 364, Figs.
Durham, Fig. 101 185-86, 192
Dust storms, 253, 255, Fig. 131 Epicontinental seas, 12
Dutch East Indies, abyssal deposits, 320 Eras of geological time, 101, 106-7
— recent orogeny, 403-5, Figs. 21 1 — Erg, Sahara, 255
212 Erosion, 24
Du Toit, A. L., on continental drift, — arid 273-74
cycle,
49 8 > 503, Fig. 261 — badland, n6, Pis.
Dutton, Clarence Edward (1841-1912), — dykes, Fig. 36 146, 268, 30, 63
on isostasy, 15, 2 — 214-26
glacial,
Dwyka Tillite, 499-500, PI. 95 — marine, 286-95, 66-70 Pis.
Dykes, 82, Fig. 36, PI. 17 — normal cycle, 145, 185-89, Figs.
— columnar jointing, 78
— swarms of, 83, 108, Fig. 38 — 94-95146-47, 33
rain, PI.
— volcanic, 453-54, Fig. 234 — 150—51
river,
Erratics, glacial, 227, Pis. 4, 55, 82
Earth, age, 26, 105, 109 — in Dwyka Tillite, 500, PI. 95
— core, 374-75, Figs. 195-96
•
Eruptions, 28, 444-45, 460-62, Fig.

— crust, 9, 65, 371-73. Figs. 4, 193 245, Pis. 2, 90, 92-93
— density, — central, 28, 444, 460-62, Fig. 244,
— relief
1
of surface, 10-12, Fig. 2 PI. 2

— shape, 13 — fissure, 28, 445, 458, 479, Figs.
Earth movements, 26-28, 31 235. 244
— epeirogenic, 107, 378, 414-41 — geysers, 138, 140, PI. 29
— orogenic, 106, 401-3, 410-11, — submarine, 381, 450
„ Figs. 51-52, 215 Escarpments, 180, Figs. 48, 57, 88, 89,
Earth pillars, 147, PI. 31 Pis. 42, 87
Earthquakes, association with faults, — desert,
272
7 8 359. 424, Figs. 183-84 Escher von der Linth, Arnold (1807-
— ,
damage, 361, Pis. 78-79 1872), on Alpine structure, 390

— distribution, 315-17, Fig. 187 Eskers, 233-35, Figs. 122, 124
—
— geological effects, 362 Etna, 454
intensity, 363, Fig. 185 Euphrates, river, 167-68
— seismic waves, 368-72, Figs. 189- European ice sheet, 217, 245, Figs.
190, 195 127, 132
— volcanic, 358, 365, 367, 463, 469 — loess, 267, Fig. 143
( 396 )
517 34
INDEX
European orogenic belts, 384-87, Fig. Floods, control, 168
200 — deposits, 167
— orogenic revolutions, 109 — Hwang Ho, 171
— volcanoes, 476 — 168
Mississippi,
Eustatic changes of level, 418 — desert, 271
Everest, Sir George (1790-1866), on Flowage, ice, 209-1 223
gravitational attraction of Hima- — rock, 69-70, 74-75
1,
layas, 16 — isostatic readjustment, 32, Fig. 9

Everest, Mt., 27, 201 Fluviatile deposits, 121, Pis. 35, 40
Exfoliation, 114, PI. 25 Focus of earthquake, 364, Figs. 185-86
Expanded-foot glaciers, 209 Folding of strata, 27, 70-75, Figs. 25-
Extrusive rocks, 28, 50, 52 28, Pis. 7, 12, 13, 80-81
Foliation, 63
Face of the Earth, 17, 18, 487 Footwall, 79
Faceted pebbles, 258, Fig. 136 Foraminifera, 31 1, PI. 72
— spurs, 220, Fig. 1 14, Pis. 52-53 — deposits, 316-20
Falcon Clints, Teesdale, Fig. 18 Forbes, Edward (1815-1854), on land-
Fall Zone, Appalachian, 159 bridges, 488
Fans, alluvial, 201-2, Fig. 101 Forelands of orogenic belts, 382, Figs.
Fault, breccia, 79 198-99
— mountains, 415, Figs, 217, 221, Foreset beds, 169, Fig. 78
Pis. 84A, 85B Forests, Carboniferous, 339, Fig. 179,
— scarp, 79, 84B PI. 74PI.
— trough, 416, 217-18, 225
Figs. — submerged, 28, 196, 301
Faults, 70, 78 Formation, geological, 101
— association with earthquakes, 358- Forth, River, meander belt, Fig. 76
183-84
Figs. Fossils, 99, 101, Fig. 50
— 360,
block, 415, 217, 221
Figs. Fractures, 69, 70, 75-81
— on ground-water, 130
effects — conical, 87, Fig. 43
— normal, 79, Figs. Fragmental rocks, 53-54, 57
— reverse, 80, Fig. 3230-31, 15
PI.
Frazer River, B.C., alluvial terraces,
— 438-41,
rift-valley, Figs.
231, 233 PI. 43
— strike-slip,tear or transcurrent, Fringing reefs, 322, 325, Fig. 172
Fig. Frost, 22, 113, Pis. 4, 82
— 81, 80,35 Fig. 33, 16
thrust, PI. — corries, 218, Fig. 113
Felspars, 39, 53, 65 — screes, 113, Pis. 22-23
— decomposition, 40—41, 117 — strandflat, 289, Fig. 154
Ferromagnesian minerals, 41-42, 51-52 Fuels, 312, 330-56
decomposition, 118 Fumaroles, 138, 445
Fetch of wind, 282 Fusain, Fusinite, 337-38
Findhom River, alluvial terraces, PI. 43 Fusion, crustal and subcrustal, 65, 66,
Fingal’s Cave, 47, 76, PI. 3 Fig. 21, 479, 483, 485
Finland, esker. Fig. 124
— gneisses and migmatites, 64-65, Gabbro, 50, 52, PI. 18
PI. 10 Gaping Ghyll, 135, PI. 26
— granitization, PI. 1
— ice-moulded Gases, natural (petroleum), 343, 349,
shore, Fig. no 35i
— islands,
304 — .
volcanic, 443, 446-48

— lakes, 240 Geanticline, Fig. 207
Fissure eruptions, 28, 445, 458, 479, Generation of tides, 279, Fig. 144
Fig. 235, 244 Geo, 287, PI. 67
— springs, 128, Fig. 55c Geological processes, 23, 30-32
Fjords, 224-26, Fig. 1 1 —
Flint, 141, Fig. 151 — record 97-1056,
time, 25-26, 102-5, Fig. 52
Flood-plains, 167-68, 188, Fig. 77 Geology, branches and scope, 4-6
— initiation, 164, PI. 39 Geosyncline, 380
— meander belts, 166 —origin, 410, Fig. 215
Floods, causes, 148, 168-69 —part of orogenic cycle, 402
INDEX
Geosyncline, present-day example, 403 Grand Canyon of the Colorado, 161,
— thick sedimentation, 379 199-200, Figs. 70, 103, Pis. t, 84
Gersoppa Falls, 59 Granite, 38, 45
Geyserite, 138-39
1
— Alpine, 397, Pis. 50, 83

Geysers, 138-40, Fig. 61, PL 29 — batholiths, 86
Giant’s Causeway, 2, 47, 78, Pis. 3, 14 — Dartmoor, Figs. 8, 62
Gilbert, Grove Kari (1843-1918), on — Finland, 64, Pis. 10, 11
laccoliths, 86 — Land's End, PI. 15
Glacial, control theory of coral reefs, — Matopo Hills, PL 25
—
— 326-29
origin, 46, 65, 96, Pl. 1
deposits, 226-32,
121, Figs. 118- — Scilly Isles, PL 66
— Shap
— 121, 22655, 57, 66
drift,
Pis.
— Skye, PL
Fells, Fig. 12
18
— epochs, Fig. 133 — weathering, 1 19, PL 25
— erosion, 247-49, 320,
214-26, Figs. 1 10-17,
-
Granitic, layer of crust, 371-73, Fig.

Pis. 51-54 "93
— 49,241,
lakes, Figs. 118, 123, Pis. 54, — magma, g2
56 . Granitization, 65, 91, 378, 402, PL 11
Glaciation, 216 Granodiorite, 51-52, 89, Fig. 15
— Africa, 504 Granophyre, Fig. 13
— Carboniferous, 250, 499-504, Figs. Granulitic rocks, 64
259-61 Graptolites, Fig. 50
— interruption of cycle of erosion, Gravitation and shape of earth, 13, 15
"45 Gravity, anomalies of, 404-6, 440, Figs.
— Mauna Kea, 327 211-13,232
— Pleistocene, 245-49, Figs. 127, Great Basin, 414, 421, 423, Figs. 201,
132-33 219
— suggested causes, 249-51 Great Barrier Reef, 322, Fig. 173,
Glacier, flowage, 209-11 p l- 73
— 205
ice, Great Escarpment (South Africa), 193,
— mills,214 430, pl 87
— tables,214 Great Lakes of North America, 237,
Glaciers, 204-14, Fig. 106, Pis. 46-48, 241-44, Figs. 128-31
56 Great Rift Valley, 432, Figs. 226-28
— zones of fracture and flowage, 69, Great Ross Barrier, 207-8, Fig. 105
21 Great St. Bernard Nappe, 396, Fig.
— zones of supply and wastage, 216 201
Glaciofluvial deposits, 121, 226, 232-35, Great Salt Lake, 424-25
238, Figs. 122, 124, PI. 57 Great Whin Sill, 84-86, 103, 157, Figs.
Glaslyn, PI. 53 "8, 39, 40, PL 18
Glenariff, falls, 158 Greenland ice sheet, 206
— pot-hole, PI. 35 Gregory, John Walter (1864-1932), on
Glen Feshie, PI. 40 rift valleys, 432
Glen Roy, Parallel Roads of, 236-37, Griggs, D., on sub-crustal convection
Fig. 125, PI. 56 currents, 410-12, Fig. 216
Globigerina ooze, 55, 316-19, PI. 72 Grikes, 117, Pis. 8, 24
Gloup (blow hole), 287, PI. 67 Ground moraine, 213
Gneiss, 64, Fig. 21 Groundsvvell, 282-83
Gobi Desert, 428, Fig. 222 Ground-water, 8, 22, 126-42, Figs. 54-
Gondwanaland, 367, 399, 487, 491, 61, Pis. 26, 27
Figs. 208, 210 — effect on vulcanism, 463-64
— 505,
Carboniferous glaciation, 499-503, — in oilfields, 347-49, Fig. 181
Figs. 259-61 — limits wind deflation, 256
Gorges, 153, 158-61, 196, 200-1, Pis. Groynes, 296, Pl. 70
Guano, 313
„ 21 ’ 35
,
Grade, of rivers, 153-54, Figs. 65-67 Guayra Falls, 159
Grain of igneous rocks, 48, 50, 52 Gulf of Bothnia, uplift, 33, 238
Grampians, 191, Fig. 98 Gullying by rain, 146
5r9
INDEX
Hamaatite, 39, 54 Hwang Ho, Fig. 79

Halemaumau, 463-65, Fig. Pis. — delta, 170
90-9 1
245,
— floodplain, 168
Hall, Sir James of Dunglass (1761- Hydraulic action, rivers, 150
1832), on recrystallization of lime-
— sea, 286, 294
stone, Hydrocarbons, 343-44
59
Halley, Edmund (1653-1742), on Hydrosphere, 8
salinity of lakes and
seas, 25 Hypabyssal rocks, 50, 52
— on circulation of water, 126 Hyposographic curve, 10, F'ig. 2
Hammada (Sahara), 255
Hanging
—
glaciers, 209
valleys, 221, Fig. 114, Pis. 52-53, — 205 206,
Ice,
caps, PI. 45
— cave, 48 PI.
— 69 wall, 79 — flowage, 209-11, 223
— heaving, 147
Hardpan, 125
Hawaiian Islands, 457, Fig. 242 — tunnels, 214
— type of eruption, 460, 462-63, Fig. Ice ages, Carboniferous, 250, 499-502
— Pleistocene, 245-49, Fig. 133
244, PI. 90
Headward erosion, valleys, 153 — suggested causes, 249-51
— conies, 219 Ice
—
206-7,
sheets, 105, 132
Figs.
Heat, flow through crust, 481-82 on sea
effect 249, 327, 417
level, 12,
— volcanic, 447, 480-85 — 33, 237-38, 418,
isostatic effects,
Hercynian, igneous rocks, Figs. 51, 52 Fig. 126
— massifs, 386, 419, Figs. 203-5, 218, — Pleistocene, 245-46, Fig. 132
PI. 82 — Carboniferous, 94, 499, 502, Figs.
— orogenesis, 108-9, 386-87, 4 2 5> 259-61
Fig. 200 Icebergs, 206
Hex River and Mountains, South Iceland, bar, Fig. 163
Africa, Pis. 30, 95 — basalt and granophyre, Fig. 13,
High Calcareous Alps, 394, Figs. PI. 23
203-6, PI. 81 — fissure eruptions, 460
High Force, Teesdale, 86, 157 —
—
geysers, 138, Fig. 61
Himalayas, antecedent drainage, 200-1 ice-dammed marginal lake, PI. 56
— gorges, 153 — relation continental
to drift, 497-
— gravitational
— snowline, 204
.
attraction, 16
— 498 screes, PI.
23
— 201
uplift of peaks, 190, Igneous
—
activity,
28, 3:
Historical Geology, 6, 93-109, Figs. 108
British,
50-52 Igneous rocks, 38, 45, 49-52
Hogback, 182, Fig. 89, PI. 42 — ages, 102, 108, Fig. 51
Holderness, coastal erosion, 291, Fig. — defined, 29, 67
1 55
— modes of occurrence, 70, 82-90,
Hooks, 297, Fig. 159, PI. 71 Figs. 36-45, Pis. 17, 18
Horizon, distance of, 13 —
origin, 45-47, 66, Fig. 21
Hornblende, 42, 52, Fig. 15 Iguazu Falls, 159
— -schist, 64 Ilmenite, 39, Fig. 14
Homblendite, 52 Impermeable rocks, 128—29
Homfels, 63 Impervious rocks, 128-29
Horsts, 415, Figs. 217-18 Incised meanders, 197-98, Figs. 101-2,
Hot springs, 137-38 PI. 44
— deposits, 138, PI. 28 India, Carboniferous glaciation, 499-
Hudson Bay, 12 500, Figs. 259-61
— region, 152, 189 —cotton soils, 125
Humus, 1 15, 122, 332
- —
laterite, 120
Hutton, James ( 1 726- 1 797) on granite, Indus, floods, 169
—
,
46, 65 gorges, 200

— “ no vestige of a beginning,” 66 Ingleborough, Fig. 47
— key to past, 93 —
caves, 135
520
INDEX
Ingleborough, “ grikes,” 117, PI. 24 Jurassic fossils, Fig. 50
— swallow holes, PI. 26 — period, 108
Inselberg landscape, 114, 274-75, Fig. — plateau basalts, 459, Fig. 252
144, PI. 65 Juvenile water, 23, 127, 133, 138, 140
Intensities of earthquakes, 363, Fig. 185
Interglacial stages, 247-49, Fig. 133 Kaieteur Falls, 158
Interior lowlands, 180 Kalahari Basin, 429-30, Fig. 223
Interior of earth, 371-75, Figs. 193, —roaring sands, 267
195-96 Kalambo Falls, 192, Fig. 99
— North America, 388, Fig. 201 Karnes, 233-34
Intermediate layer of crust, 371, Fig. Karelian orogenesis, 109
•93 Karroo Basin and System, 430, Fig. 223
Intermittent, streams, 143 —dykes and sills, 460
— zone of saturation, 127-28 Karst region, 117, Fig. 165
Internal drainage, 143, 270, 414, 424, — topography, 135, Fig. 166, PI. 8b
438 Kettle holes, 233, Fig. 123
.
Intrusions, igneous, 50 Kilauea, 447, 457, 462-65, Figs. 242,

— modes of occurrence, 82-90 245, Pis. 89-91
Intrusive rocks, 29, 50-52, Figs. 36-45, Kilimanjaro, 435, Fig. 226, PI. 45
Pis. 17, 18 Killarnean orogenesis, 109
— ages of, 102, 108, Fig. 51 Kivu, Lake, 435-36, Fig. 226
Iron ores, 38, 39 Knick point, 195
— replacement of limestone, 60 Kopje, 182
Irrawaddi delta, PI. 75 Krakatao, 452, 470-73, Figs. 250-51,
-
— rate of denudation, 152 PI. 91
Islay, raised beaches, 27, PI. 6 Kuenen, Ph. H., on suspension currents,
Isoclinal folding, 73, Fig. 26, PI. 69 310
Isomorphism, 37, 39 Kvoga, Lake, 432, Figs. 224, 226
Isopiestic level, 15
Isoseismal lines, 364, Fig. 185 L (Long) waves, 369, 373, Figs. 189-
Isostasy, 15, 32-34, Figs. 3, 4 191, 194
— ice sheets, 327-28, Fig. 126 Laacher See, Fig. 239
— orogenic belts, 404-6, Figs. 211-13 Laccolith, 86, 422, Fig. 41
— relation to denudation, 189-90, Lacustrine deposits, 121, 156, Pis. 31,
Figs. 86-87 37. 63
— rift valleys,
440-41, Fig. 232 Lagoons, coastal, 306
— sedimentation, 379-80, Fig. 197 —coral-reef, 322-28, Figs. 174-75
— of glaciated
uplift areas, 33, 238, Lake Agassiz, 243
242, 418 —Albert, 436, Figs. 229-30, 232
Isostatic equilibrium, 32 —Algonquin, Fig. 130
— compensation, 15 —Baikal, 244, 428, Fig. 225
— readjustment, 32, Fig. 9 —Bonneville, 425
Izalco, volcano, 453 —Garda, 229, Fig. 119
—Kivu, 436
Japan, earthquakes, 359, 362-63, Figs. —Kyoga, 432, Figs. 224, 226
183, 186, Pis. 78-79 —Magadi, 438, Figs. 226, 232, PI. 86
Jeffreys, H., on tidal friction, 492 —Nakuru, PI. 86
Joints, 70, 75-76 —Natron, 438, Figs. 226, 229
•
— columnar basalts, 76-77, Fig. 29, —Tanganyika, 434-36, 440, Figs.

Pis. 3,
14 225-26, 232
— granites, 76, Fig. 62, PI. 15
— influence on
— Titicaca, 244
—
-
Toba,
erosion, 76, 214, 217,
258, 287, Figs. 62, hi, Pis. 15, — Victoria,455431-32, Figs. 226, 229
32, 64, 67-68 Lake basins, classification, 244-43
— sedimentary rocks, Pis. 8, 14, 24 Lake District, 61, 108, 182-84, Figs.
Jordan, meanders of River, PI. 85 91-92, Pis. 22, 37, 55
Jura, raised beaches, 27, PI. 6 Lakes. 240-45
Jura Mts., 393, Fig. 203, PI. 80 — elimination, 1 55-56, 333, Pis. 3 1, 37
521
INDEX
Lakes, North American, 240-44, Figs. Limonite, 37, 39, 49, 54, 313
128-31 Lisbon earthquake, 362
— ox-bow, 165, PI. 41 Lithosphere, 9
— temporary, 271 Littoral deposits, 313, 316, Fig. 171
— terraces, 236-37, 427, Fig. 125, Load of stream, 151-52
Pis. 56, 63 Loam, 122
— salt, 24-25, 271, 424, 438 Local base-level, 155-56, Figs. 66-67
Laki fissure eruption, Iceland, 460, Loch Coruisk, Skye, PI. 54
Fig-. 235 Lochs, 224
Lamination, 54 Loess, 255
Land and sea, 11, 17, 18, 28 —China, 267-69, 363, Fig. 79, PI. 63
Land bridges, 488, 499, 504 —
Europe, 267, Fig. 143
Landes, sand dunes, 261 London, alluvial terraces, Fig. 100
Land’s End, PI. 15 — artesian wells, 132
Landslides, 78, 148, 150, 169, 362-63, — future fate, 249
Figs. 63-64, PI. 35 London Basin, Fig. 57
Lapilli, 451 London Clay, 94
Laramide orogenesis, 397 Longitudinal, crevasses, 2 II
Lateral moraine, 213, PI. 5 — dunes, 263-64, Fig. 140
Laterite, 119-20, 124, Fig. 53 — profile of valley floor, 153-56, 222,
— Carboniferous, 502 Fig. 115
Laurasia, 367, 399, 487, 490, 505, Longshore currents, 295
208-9
Figs. Long waves, 369, 373, Figs. 189-91, 194
Laurentian orogenesis, 109 Lopolith, 86, Fig. 42
Lauterbrunnen valley, 222 Lower layer of crust, 371-72, Fig. 193
Lava, 28, 443 Lulworth Cove, 302-3, PI. 70
— cascade, PI. 89
— caves and tunnels, 449 Maars, 452, Fig. 239
— columnar, 47, 76-78, Pis. 3, 14 Magadi, Lake, 438, Figs. 226, 232,
— fountains, 460, 463, 465, 90 PI. PI. 86
— lake of Halemaumau, 457, 463, Magellan, Ferdinand (1470-1521), cir-
PI. 90 cumnavigation of globe, 12
— plateau, 458-60, Fig. 243 Magma, 28, 443
—
— pillow, 449, PI. 89 cause of ascent, 478-79, Fig. 253
— types, 448-50 — origin, 65, 92, 378, 480, 483-86,
Lead ratio of radioactive minerals, Fig. 21
103, 106 — temperature, 448, Fig. 255
Levees, 167-68, Fig. 77 Magmatic emanations, effects, 59, 60,
Life, biosphere, 8, 9 64-65, 92, 378, Fig. 2
— destructive work, 22, 112, 115 — suggested source, 485
— marine, 314-17
-
Magmatic stoping, 91
—
—
organic sediments, 25, 311-29 Magnesian limestone, 57
Magnetite, 39
— source122of
- soil,
330-47
fuels, Malaspina Glacier, 209, Fig. 106
Lignite, 335, Fig. 1 78 Malham Cove, Pis. 8, 24
Limb of fold, 72, Fig. 25 Mamelons, 456
Limestone, 25, 38, 55-57, 312, 314, Mammoth Cave, 135
Mammoth Hot
— caverns,24134, 55,
Pis. 9, Springs, 138, PI. 28
Fig. Pis. 26-27 Mantle of rock- waste. 12
— coral,
323 Marealbian orogenesis, ioq
— crinoidal, 55, PI.
9 Marble, 57-59
— 58-60
crystalline, Marginal, crevasses, 21 Fig.
—
—
jointing, Pis. 8,
24
—
1, 107,
oolitic, 56, Fig. 17 ice-dammed lakes, 235-38, PI. 56

—
—
129
porosity, Marine deposits, 294-95, 313-27, Figs.
shelly, 55, PI.
9 59“64 7 i_ 74, Pis. 69-73
I
53 i
— solution, 1 16, 134, Fig. 55/, Pis. ~ S1

pV’ cc
>.
’ 76,
>
i
286"95, Figs. 151-57,

24, 26
522
INDEX
Marine transgressions, 107, 417-18, Modes of occurrence, igneous rocks, 50,
420, 428 71, 82-90
— zones of sedimentation, 313, Fig. Moine thrust, Figs. 33, 180
171 Molasse, 393
Marl, 123 Molluscs, Fig. 50
Matopo Hills, PI. 25 Monadnocks, 189
Matterhorn, 220, 395, PI. 50 Mongolian Plateau, 426, 428,Fig. 222
Mature stage, arid cycle, 273 Mont Blanc massif, 386, Fig. 204,
— normal cycle, 145, 166, 188, PI. 82
Figs.
94-95 Mont Pelee, 462, 468-70, Fig. 249,
—- soil development, 123 PI- 93
Mauna Loa, 457, Fig. 242 Monte Nuovo, 453
Maviston Sandhills, Pis. 58-59 —
Rosa, 395 nappe, 396, Fig. 204
;
Meander belt, 166, Fig. 76 —

- Somma, 446, Figs. 246, 248,
Meanders, 164-66, Pis. 41, 85 PI. 92
— PL 44
incised, 197-98, Figs. 101-2, Monsoon climate, soils, 124
Medial moraine, 213, 46 PI. —
weathering, 118-19
Median areas, 383, Figs. 198-99 Monument Valley, Utah, incised
— volcanoes, 476 meanders, PI. 44
Mediterranean, advance into ffsgean —
sandstone pillars, 422, PI. 64
420 Monzonite, 52
— earthquake belt, 365, Fig. 187 Moon, alleged capture during Cretace-
— orogenic belts, 398-99, Figs. 200, ous, 491
—
203 separation from Earth, 19, 20
— salinity currents, 281, Fig. 146
-
—
tidal effects, 279, Fig. 145
Meinesz, F. A. Vening, gravity anoma- Moraines, 2 1 3, 2 1 4, 228, Figs. 1 1
8, 1
1 9,
lies and orogenesis,404-5 122, PI. 5, 46
Mercalli scale, modified, 363 Mountain, building, 106, 377-413
Mer de Glace, 209, PI. 46 — glaciers, 206, 209
Meres of Cheshire, 245 —
range, 379
Meru, Mt., 435, PI. 88 Mountains, fault-block, 415, Figs. 217-
Mesa, 182, Figs. 89, 90, 142 2 18, 221, PI. 84
Mesosaurus and continental drift, 499 —
folded, 377, 381-401, Figs. 202,
Mesozoic Era, 104-5 204, Pis. 80-81
Metamorphic aureoles, 63, Figs. 20, — roots, 13, 17, 378, 402, 404, Figs.
4, 212, 216
— 45 cycle, 21 Fig. Mount Everest, 12, 27, 201
— rocks, 65,
57-67 Mozambique, inselberg landscape, 1 14,
Metamorphism, 30-31, 58-67 274-75, Fig- 144
— on ground-waters, 127
effect Mud volcanoes, 351
Meteor Crater, Arizona, 422 Mudcracks, 76, PI. 19
Meteoric water, 23, 127, Fig. 7 Mudstone, 54
Mexican Plateau, 421, Fig. 201 Mull, cone-sheets, 87
Mica, 41, 45, 53, 63 — dyke swarm, 83, 482, Fig. 38
— -schist, 64 Muscovite, 41, 45, 1 16, Fig. 15
Midland Valley of Scotland, 108, 436, Mylonite, 80, 485, Fig. 34
Fig. 180
Migmatite, 65, Pis. 10, 1 Nabesna Glacier, Alaska, PI. 46
Mindel glaciation, 248 Nappes, 80-81, 387, 391-97, Figs. 28,
Mineraloids, 37 204-5
Minerals, 35-43, 51-52. Fi S- >5 Natural bridges, 135, Fig. 60 198, ;
Mineral springs, 138 Fig. 102, PI. 44 287, PI. 68

— veins, 142 Natural gas, 343, 351
;
Mississippi, black bottoms, 125 Near earthquakes, 371, Fig. 193

— delta, 171, Fig. 80 Needles, Isle of Wight, 289, Fig. 151
— floodplain, 168 Negative anomalies of gravity, 404-6,
— 168
levees, 440, Figs. 21 1— 13, 232
— meanders, 165 Nekton, 315
523
INDEX
Neritic deposits, 314, 316, PI. q Old Red Sandstone, coastal features,
N6v£, 204-5, Fig. 108 Fig. 152, Pis. 67-68
New
— Zealand, geysers, 138 — conglomerate, 2t PI.
White Terraces, PL 28 —
Newton, Sir Isaac (1642-1727), on — current
earth
bedding, 20
pillar, PI.
31
PI.
shape of earth, 13, 14 —

-
fish, Fig.
50
Niagara Falls, 157, Fig. 68, PI. 38 — mudcracks, 19 PI.
Nigeria, inselberg landscape, 274, PI. 65 Oldestknown rocks, 65
Nile, River, fed by ground water, 129, Oldham, Richard Dixon (1858-1936),
on earthquakes, 364
— delta, 170, Fig. 134 Oldham, Thomas (1817-1878), on oil
— sudd, 430, PI. 75 in Burma, 352
Nivation, 219 Oldonyo l’Engai (Doinyo Ngai), 438
Nodules, 141 Olivine, 42, 52
Normal, cycle of erosion, 185-89, Figs. Oolitic limestone, 56, Fig. 17
94-95 Oozes, deep sea, 312, 316-20, PI. 72
— fault, 79, Figs. 30-31, PI. 15 ~ lake, 333
North America, Appalachians, 387-90, Orange River, Aughrabies Falls, 159.
Fig. 202 Fig. 69
— badlands, 146, PI. 30 — dendritic tributaries, Fig. 81
— Cordillera, 420-25, Fig. 219 Ordovician fossils, Fig. 50
—
—
glaciation, 245, Fig. 132 — period, 108
Great Lakes, 237, 241-44, Figs. — volcanic activity, 381, Fig. 51
128-31 Ore deposits, 35, 142
— tectonic map. Fig. 201 Orogenic belts, 106, 377
North Platte River, PI. 40 — Alpine-Himalayan, 397-401, Figs.
North Queensferry Sill, Pis. 24-25 208-1
North-West Highlands, 386, Fig. 33 — Appalachian, 387-90, Figs. 201-2
Norway, erratics, 227 — Caribbean, Fig. 213
— fjords, 224-26, F'ig. 117 — Circum-Pacific, 109, 398, Fig. 208
— strandflat, 28g, Fig. 154 — dislocations, 419-20
Nude ardenle, 462, 469-70, Fig. 244, — Dutch East Indies, 403-51, Figs.
PI. 93 211-12
Nullipores, 323, 325 — structures, 381-84, Figs. 198-99
Nunataks, 206 — suggested origin, 505, Fig. 262
Orogenic cycles, 401-3, 410-11, Figs.
Oases, 133, 256, 429, Figs. 58, 135, 51.521,215
PI. 61 — revolutions, 106
Oblique slip fault, 79, Fig.
31 Orogenesis, 106, 378
— waves, 284, 295, Fig. 149, 158 — present-day, 403-6
Obsequent streams, 180 Orthoclase, 39, 52, 117, Figs. 12, 15
Obsidian, 50 Outwash plain, 232-33, Fig. 122
Ocean, currents, 281 Overflow channel, 235
— floor,
319 Overfolds, 27, 75, Pis. 13, 81
—
— platform, 10, Figs. 2, 171 Overthrust fold, 75, Fig. 28
salinity, 25, 280-81 Overthrusting, 80, 381-83, Figs. 33,
Oceanic deeps, S
11, 12, 494, 507, Fig. 2 .98-99, PI. 16
— crust,
373, Fig. 194 Ox-bow lakes, 165, PI. 41
Offshore bar, 292-93, 304-5 Oxides, 36, 39
Oil, mineral, 343-54 Oxygen, source of atmospheric, 311,
—
—
pools, 345-48, Fig. 18 330-31, Fig. 177
shales, 336, 345-46
Oilfields, Fig. 182 P (Primary) waves, 368-74, Figs. 189-
Old Age stage, arid cycle, 274
— normal cycle, 145, 188-89, Figs.
191, 193
Pacific coastal type, 30 x, 401, Figs.
94-95 166, 208
Old Faithful Geyser, PI. 29 — 82
floor,
Old Man of Hoy, 289, Fig. 152 — structure of crust, 373, Fig. 194
524
1
INDEX
Pacific volcanoes, 476, Fig. 252 Plain. Lombardy, 397, Fig. 203
Pahoehoe lava, 448, Fig. 236, PI. 89 — Swiss, 393, Fig. 203
Palaeozoic Era, 104-5 — tract of river, 152
— igneous activity and orogenesis, 108 Plankton, 315-17
Pangaea, 492, 505 Plant life, coal, 334-36, 339, Fig. 179,
Parallel Roads of Glen Roy, 236-37, PI. 76
Fig. 125, PI. 56 — destructive action, 115, PI. 4
Parana, plateau basalts, 159, Fig. 252 — geological work, 22, 1 15, 1 16, PI. 4
— waterfalls, 159 — marine, 315
Parasitic cones, 453, Fig. 234 — peat, 332-34
Pavement, desert, 258 — petroleum, 346
Peat, 12 1, 124, 196, 332-34, Fig. 178, — protective action, 115-16,253,261
PI. 74 — 122
soil,
Pebbles, effect on desert surface, 255, 265 Plateau basalts, 108, 458-60. Fig. 243,
— faceted, 258, Fig. 136 PI. 17
—
— glaciated, 227 distribution, Fig. 252
—
relation to basaltic layer, 479-80,
Pediment, 272, 274
Pelagic deposits, 315-20, PI. 72 4 8 4> 507
Pelean type of eruption, 462, 469, Plateau glaciers, 208
Fig. 244, PI. 93 Plateaus, 414
Pelee, Mont, 468-70, Fig. 249, PI. 93 —Africa, 428-32, Fig. 223
Pele’s hair, 460 —Asia, 425-28, Fig. 222
Peneplain, 144, 189, 274 — North America, 389, 420-25,
— uplifted, 191-93, 419, 431, Fig. 98 Figs. 201-2
Pennine Alps, 395-96, Figs. 203-4 Platform, shore-face, 289, Figs. 153-54
— source of nappes, Fig. 207 Playa, 424
Penrith Sandstone, 260 Playfair, John (1748-1819), on trans-
Perched blocks, 227, PI. 55 port of erratics, 227
Percolation, 127 Pleistocene, changes of sea level, 194.
Peridotite, 42, 52 238, 2 49> 327 4I7-i8
Periods, geological, tot, 104-5, I0 7> — ,
desert climates, 270-1

Fig- 5i — glaciations, 245-49, F>g s - I2 7>
Permanence of continents and oceans, 132-33
488 Plinian type of eruption, 462, 466, Fig.
Permeable rocks, 128 244’
Perrault, Pierre (1611-1680), on rain- Plucking by ice, 214
fall and stream flow, 126, 143 Plumb-line, deflection, 16
Perret, F., observations on Mont Pelee, Plutonic rocks, 49-52
462, 469 Po, River, delta, 1 70
— Vesuvius, 465 —flood plain 168
Persian Gulf, during ice age, 417-18 Podsol, 123-24
Peru, 14, 16 Polar flattening of earth, 14
— uplift, 27 Poles, alleged wandering, 250, 492,
Pervious rocks, 128-29 495. 503
Petrified wood, 141 Polftucht, 494-95
Petrol, sources of, 344-45 Polyps, coral, 322
Petroleum, 343-45 Porphyrins, 346
— concentration, 348-51, Fig. 181 Porphyrite, 52
— distribution, 346, 354, Fig. 182 Porphyritic texture, 45, 50, Fig. 12
— production, 354 Porphyry, 50, 52
Port Sudan, dust storm, Fig. 134
Phosphates, 313
Piedmont, alluvial plain, 20 Porosity of rocks, 128-29
— glaciers, 206, 209 —effect on oil migration, 348
Piggot, C. S., on sampling ocean floor, Portland stone, 25, 55-56
3'9 Pot-holes, 150-51, 2 1 4, PI. 34
Pillow lavas, 449, PI- 89 Pre-Alps, 393-94, Figs. 203-4
Pitch of fold, 73 Pre-Cambrian eras, 104-5, io 9
Plagioclase, 39-40, 52, 117, Figs. 14, 15 —orogenic revolutions. Fig. 52
525
INDEX
Primary waves, 368-74, Figs. 189-91, Recrystallization, ice, 210
193 —rocks, 58-59, 63
Processes, geological, 30-32 Recumbent folds, 75, Figs. 27, 28,
Profile, cross-, 153, 161, Figs. 70-71 Pis. 13, 81
— graded, 154, Fig. 65 Red Clay, 316-18, 320
— longitudinal, 153-56, 222, Fig. 115 Red Hills, Skye, PI. 18
— of equilibrium (shoreface), 291-93, Reefs, coral, 321-29
Figs. 156-57 Refraction, earthquake waves, Figs.
Protective effect of vegetation, 115-16. i93» J 95
253, 261 — sea waves, 284-85, 297, Figs. 149-
Pteropod ooze, 317, 319 .150. "59
Pteropods, 315 Regional metamorphism, 63
Pumice, 50 Rejuvenation, rivers, 195-200
— tuffs, 472 —
mountains, 377
Puys of Auvergne, 47, 456, Fig. 241 Relief, Earth’s surface, 10-12, Fig. 2
Pyro-bitumen, 344-45 —relation to isostasy, 15, Fig. 4
Pyroclasts, 71, 443, 450-52, 455, 462, Replacement, 141
472, Figs. 237-38 —-
granitization, 65, 91, PI. 11
Pyroxenes, 41, 42, 52 Reservoir rocks, 345, 348, Fig. 181
Pyroxenite, 52 Residual boulders, PI. 25
— deposits, 1 21
Qattara depression, 256, Fig. 135 — 1 9, 1
landforms, 144, 274, Figs. 62, 90,

Quartz, 36,51,52-54, u6. Figs. 10-12, 144, Pis. 64-65
16 Residues, weathering, 1
1
Quartzite, 64 Reverse fault, 80, Fig. 32
Quartz-diorite, 52, Fig. 15 Revived landscape, 195
Quartz-porphyry, 50, 52 Revolutions, orogenic, 106, 107, Figs.
Quaternary period, 104 5'-52
Queensland, artesian basin, 133 Rheingraben, 416, Fig. 218
— fossil wood, 141 Rhine rift valley, 416, 346, Figs. 203,
218
Radial drainage, 183-84, Figs. 91-92 — “ keystone ” hypothesis, 438
Radioactive elements, 103, 409, 480 Rhone Glacier, Pis. 47-48
— minerals, 106
103, Rhyodacite, 52, Fig. 15
— source of heat in earth, 481, Fig. Rhyolite, 50-52, 91, 158
254 — -tuffs, 455
Radiolarians, 312, 315, PI. 72 Rias, 301, Fig. 165
Radiolarian ooze, 317-18 Rift valleys, 416, 433
Rainbow Bridge, Utah, 198, Fig. 102, —African, 432-38, Figs. 225-32,
PI. 44 Pis. 85-86
Rainfall, disposal of, 22, Fig. 7 —
earthquakes, 367
— in deserts, 270 —
origin, 438-41, Figs. 232-33
Rain, geological work, 22, 146-47, —
volcanoes, 435-36, 438, PI. 87
PI- 33 Ring-dykes, 87-88, Fig. 44
Raised beaches, 27, 33, PI. 6 Rio Grande, meanders, PI. 41
Rank of coal, 335-36 Ripple marks, water, 96, PI. 20
Rapids, 156 —wind, 261, Pis. 5, 60
Rayleigh, Lord, on radioactivity of Riss glaciation, 248
rocks, 409, 480 River, base level, 153, Fig. 65
Recessional moraines, 228, 240, Fig. — bends, 161, Figs. 72-73
I2 7 — capture, 177-79, Figs. 84, 86-87
Recession, coast lines, 303, 306, PI. 66 — deltas,
169-72, Figs. 79, 80, 135
— escarpments, Fig. 88 — deposits, 24, 167, Fig. 77, Pis.
— European ice sheet, 238-40, Fig. 37. 40
35,
127 — erosion, 150-51, Pi. 1

— glacier snouts, 205 — meanders, 164-67, 197-98, Figs.
— waterfalls, 151, Figs. 67-68 75-76, 101-2, Pis. 41, 85
— watersheds, Fig. 82 — system, 144, 173
526
INDEX
River terraces, 195-96, Fig. 100, PI. Jean de Luz,
— tracts, 150
43 St.
St. Paul’s
folding, PI. 13
Cathedral, 25, 55
— transport, 144, 151-52 St. Pierre, annihilation of) 462,
469,
Rivers, antecedent, 200
— consequent, 144, 173
PI. 93
— dismembered, 194
Salinas, 271, 424, PI. 85
Salinity, currents, 281, Fig. 146
—
— grading of, 153, Figs. 65-67 —of sea water, 25, 280-81
insequent, 174
— obsequent, 180
Salt deposits, 25, 438, PI. 86
— domes, 349-50, Fig. 181
— rejuvenation of, 195-200 —
— subsequent, 144, 175 —
lakes, 24-25, 271, 424, 438
marsh, 303, 306, 427, PI. 85
Roches moutonnies, 217, Fig. Ill, PI. San Andreas fault, 359-60, Fig. 184
— Carboniferous, 500
49
San Francisco earthquake, 82, 360,
Rock avalanche, 148
— -flour, 215
363, Fig. 184
Sand banks, 163-64
Rock-forming minerals, 38-43
Rocks, composition, 35, 38
—dunes, 260-65, Figs. 137-41, Pis.
— disintegration, 113 —5» 60
— documents of history, —
drifts, 262
— 93
—ridges, 263-64, Fig. 140
— igneous,
metamorphic,
45, 49-52, 67
29, sheets, 263
— sedimentary, 24, 5357 30, Sandblast erosion, 257-59

Sandhills, Maviston, PI. a8-aq
Rock-waste, mantle of, 121 Sands, 24, 295
Rogers, Henry Darwin (1809-1866), —coral, 317
and Rogers, William Barton (1805- —Culbin, 261, PI. 58
1881), on Appalachians, 370 —millet seed, 259
Roman Wall, 86, PI. 18
Roots of mountains, 13, 17, 378, 402,
—singing, 266
Sandstone, 38, 53, Fig. 16
— 404, Figs. 4, 212, 216 —bedding, PI. 8
—
basaltic, 507, Fig. 262 —
jointing. Pis. 8, 14
— generation of magma in, 485 — pillars, PI.
— origin, 410-12, Figs. 215-16

outflow, 402
— porosity, 12964
— ripple mark, 96, Pis. 5, 20
Roots, plant, geological work, 23, 1 Sarcoui, Auvergne, Fig. 241
15,
PI.
4 Saturation zone of ground-water, 127,
Ropy lava, 448-49, Fig. 236, PI. 89 Fig.
Ross Barrier, 207-8, Fig. 105 54
Saussure, Horace Benedicte de (1740—
Rotomahama terraces, PI. 28 1
799 ', on roches moutonnies 217
Rounding of sand grains, 259
Rukwa, rift valley, 436, 440-41, Figs.
— ,
on former glaciation of Alps, 216
Schists, 63-64, Fig. 21, PI. 10
226, 232
Run-off, 127
—granitization, 65, PI. 11
Schistosity, 63-64
Ruwenzori, 436-37, 439, Figs. 226, Scoria, 450
230
— glaciation, 246
Screes, 113, 148, Pis.
— 22-23
— neighbouring volcanic vents, 452,
interglacial,
247
Scotland, chief faults, Fig, 180
Fig. 230 — former glaciation, 216
S (Secondary) waves, 368-74,
— igneous activity, 108
Figs. Sea, extent and depth, 11,17
189-91, 193
Sagami Bay earthquake, 359, 362-63,
— caves, 287, PI. 67
— changes of level, 12, 107, 194, 249,
Fig. 183
327, 417-18, PI. 6
Sahara, climate, 269
— — chffs, 187-89, Fig.
153, Pis. 66-70
oases, 133, 256, Figs. 58, 135, — composition, 281
— rocky61
PI.
surface, 255, PI. 61
— stacks, 287-89, Figs. 151-52, PI. 68
— sand dunes, 60
Seat earth, 340
— sandstone PI.
pillars, PI. 64
Secondary waves, 368-74, Figs. 189-91,
93
527
1 1 1
INDEX
Sedentary deposits, J2i Sierra Nevada, 421, 423, Figs. 184,

Sedimentary rocks, 24, 38, 53-57 201, 219
Sediments, cementation, 141
— Yosemite Valley, 221, Pis. 51-52
— continental, 121 Silica, 36, 39, 51
— marine, 295, 313, 316-20 Silicate minerals, 39-42
— varved, 238, 57 PI. Siliceous organic deposits, 314, 316-18
— sinter, 138, Fig. 61
Seepage, oil, 351
— water, 128, 130 Sills,83, Figs. 37, 39-40
Silurian, fossils, Fig. 50
Seif dunes, 263-64, Figs. 135, 140
Seismic areas, 365-67, Fig. 187
—
-
• isoclinally folded strata, PI. 69
— methods, 206, 353, 368 — period, 108
— waves, 368-72, Figs. 189, 193, 195 Sima, 9, 10, 13, 14, 41, Fig. 4
Simplon nappes, 395-96, Fig. 204
Seismograms, Figs. 189, 193
Seismograph, 358, 368, Fig. 188 Sinkholes, 134-35, PI. 26
Semi-arid regions, rain gashing, 146 Sinter, 138, Fig. 61
—
- rivers, 143 Skye, cone-sheets, 87
—
— soils,
24 1
— Loch Coruisk, PI. 54
— weathering, 14 1 Red Hills, PI. 18
Slate, 60-61, Fig. 19
Sequence of strata, 97-98, Figs. 47-48
Seracs, 2 1 Slickensides, 79
Sericite, 41 Slip fault, 79, Fig. 31
Serpentine, 42, 59 Slip-face of dunes, 261, 263
Shadow zone, cast by Earth’s core, 374, Slip-off slopes, 162, PI. 41
Figs. 195-96 Slumping, 148, Fig. 63
Shale, 38 —
submarine, 362
—
lamination, 54 Smith, William (1769-1839), founder
— metamorphism, 61, 63-64 of stratigraphy, 99, 101
— 129
porosity,
316, rig.
Snider, Antonio, on continental
489, Fig. 256
drift,
Shallow water deposits, 313,

Snow, bridges, 21
171
Shantung Peninsula, 170-71, Fig. 79 — change into ice, 204-5
Shape of Earth, 13, 14 — 204, 219
fields,
Shap Fells granite, 45, 90, Fig. 12 — line,

204, 246, Fig. 133
— age, 102-3 — 219
rotting,
— erratics, 227 Snowdonia, corries, 220,
—
slates, 60
PI. 54
Sheet-floods, desert, 271
Sheet structure of granite, 76 — volcanic rocks, 381
Shelf, continental, 1 Snout of glacier, 205, Pis. 5, 48
—seas, 12 Soil, 23,
— 121-25
Shelly deposits, 57, 293, 314, Ph 9 climatic types, 122-24
Shenandoah \ alley, 184 — creep, 147-48
Shield areas, Asia, 426 — erosion, 255
—Baltic, 384, 401, Fig. 200 — frozen, 204
—Canadian, 387, 402, Fig. 201 — profile,
Solifluclion, 148
123-25
Shield volcanoes, 457-58, Fig. 242
Shingle, 163-64 Solution of limestone, 116, 134, Fig.
Shore-face terrace, 291 55 f, Pis. 24, 26
Shore features, 278 Sorby, Henry Clifton (1826-1908),
— profile of equilibrium,
Figs. 156-57
291-93, investigation of rocks by thin
sections, 48
Shorelines, emergence, 27, 33, 277, —on slaty cleavage, 61
304-6,6 PI. Souffriere, eruption of, 469
—
submergence, 277, 301-4, Fig. 165 Source beds (oil), 348
Space problem, batholiths, 90-92
Shortening of crust by orogenesis, 80,
384-85, 390 — ring-dykes, 89
Sial, 9-10, 12, 14, 41, Fig. 4 Specific gravity, 9
—
concentration in continents, 19, 20 Spheroidal weathering, 118-19, Pis.
Siderite, 43 24-25
528
INDEX
Spines. Mont Pel6e, 456, 469-70, Fig. Subsoil, 121

249. PI- 93 Substratum, 9, 373
Spits, 297 — convective circulation, 408-13,
— hooked, Fig. 159, PI. 7 1
—
484-85, 505-9, Figs. 215-16, 262
Spitsbergen, corries, Fig. 113 fiouage, 32, Fig. 9
— glaciation, 208 Sudd, 430, 432, PI. 75
Suess, Eduard (1831-1914), on epeiro-
Spores in coal, 332, 336-38, PI. 76
Springs, 129-30, Fig. 55 genic movements, 416
— Sun cracks, 76, 94, PI. 19
— gas and 351-52
oil,
hot,
137-38, 28 PI. Superficial deposits, 121
— submarine, 309 Superimposed drainage, 182-84, Fig.91
Figs.
Spring tides, 280 Suspension currents, 308-10,
Spurn Head, 298, Fig. 155 169-70
Stacks, 287-89, Figs. 151-52, PI. 68 Svecofennian orogenesis, 109
Staffa, 47, 76, PI. 3 Swallow holes, 134-35, PI-
Stair Hole, Dorset, 302, PI. 7 Swamps. 124, 128, 333. Fig. 167
Stalactites, 136, PI. 27 — deltaic, 334, PI. 75
Stalagmites, 136, PI. 27 — flood-plain, 168
Steam coal, 338 Swarms, dyke, 83, Fig. 38
Stocks, 89-90 Swells, 414, 428, Fig. 223
Stone-rivers, 148 Swire Deep, 12
Stoping, magmatic, gi Swiss Plain, 393, Figs. 203, 206
Strandflat, 289-90, Fig. 154 Syenite, 52
Strata, 54 Symmetrical fold, 73, Fig. 28
— sequence, 97 Syncline, 72, Fig. 25, PI. 1
— thickness, 105, 379 Synclinorium, 74, Fig. 26

— upside down, 75, 95-96, Figs. Systems, geological, 101
27, 46
Stratification, 54, 71, Figs. 23-24, Pis. Table Mountain, 182
42, 68
1, 8, Tachylyte, 50, 83
Streams, see Rivers Tanganyika, Lake, 434-36, 440, Figs.
Striae, striations, 215-17, PI. 49 225-25, 232
Strike, 71-72, Figs. 23-24 —
Plateau, 192, 431, Figs. 99, 223
Strombolian type of eruption, 460-61, —
Rift Valley, 434-36, Figs. 225-26
Fig. 244 Tarim Basin, 427-28, Fig. 222
Subcrustal convection currents, 408-13 Taylor, F. B., on continental drift,
— continental drift, 505-9, Fig. 262 489-92
— orogenesis, 409-13, Figs. 215-16, Tear faults, 81, 360, Figs. 35, 184
262 Tectonic earthquakes, 358, 365-67
— vulcanism, 484-86 —
revolutions, 106, 358
Subglacial, moraine, 213 Temperature, changes (weathering),
—
-
streams and eskers, 234 1 13-14, PI. 23
Submarine earthquakes, 361 —

gradient, 478, Fig. 255
—
— orogenic belts, 403-6, Figs. 21 1-13 of lavas, 447-48
Submerged, banks, Fig. 176 Tension, faulti, 79
— forests, 28, 196 —
dykes. 83
Submergence, coasts of, 277, 292, 301, —hypothesis of rift valleys, 432,
Fig. 165 438-39
Subsequent streams, 144, 175, figs. — joints, 76-77, Fig. 29
84-86 Terminal moraines, 214, 228, Figs.
Subsidence, Black Sea, 420 1 18-19. I22 > PI- 3
— coastal regions, 194, 277 Terraces, lake, 236-37, 425, Fig. 125,
— continents, 489
—
Pis. 56, 63
— coral 326,
reefs, 175 Fig.
—
river, 164, 195-96, Fig. 100, PI. 43
— 171-72
deltas,
— tilted, 427
— geosynclines, 380-81, 402
— mountains, 419-20 — • wave-built, 289, 291, Figs. 153, 156
wave-cut, 289-90, Figs. 153-54*
— rift valley floors,
417, Fig. 225 Pis. 6, 12, 69
529
INDEX
Terracettes, 148 Transgressive intrusions, 82-83, 88-89,
'Terra rossa, 1 1 7, 124 Figs. 36, 38, 43-45
Terrigenous deposits, 314, 316 Transport of sediment, 23
Tertiary basalts, 458-60, Figs. 243, —
glacial, 212-13, 226-27, Pk. 5> 46
—
252, Pis. 3, 14, 17 marine, 280, 292-95, 295-300
— — 271
—
coalfields,
335
intrusions, Figs. 38, 243, Pis. 17-18 — wind, 253, 151-52,
river, 144,
255, 267, Fig. 134
83 Transverse crevasses, 21 1, Fig. 108
— 346
oilfields, Traps for oil, structural, 348-50,
— period, 104-5, Fig. 51 Fig. 1 81
Travertine, 138
Tethys, 383, 399, 427, Fig. 207
Tetrahedral hypothesis, 18, 19, Fig. 5 Triassic ammonite, Fig. 50
Texture, basalt, 48, Fig. 14 —
—
period, 104-5, Fig- 5 1
— igneous rocks, 50 reptiles and continental drift,
— porphyritic, 45, 50, Fig. 12 .
498-99
Thames, floods, 169 Tributaries, 144, 146
— terraces, 196, Fig. 100 —patterns and types, 173-75, t8o
Thermal conductivity, 481 Trilobites, Fig. 50
— contraction hypothesis, 407-8 Trinidad asphalt lake, 352
— gradient, 478,
— metamorphism, 225 Fig. Tropical soils, 124
—
weathering, 113, 118-20
83,62, 20 Fig.
Thrust 16
faulting, 75, Figs. 28, 33, PI. Trough end, 223, Fig. 116
— rift valley,
439-41, Figs.
231, 233 Trow Ghyll, Fig. 58
Tiber, delta growth, 1 70 Truncated spurs, 220, Fig. 114, PI. 51
Tibet, 200-1, 427, 476, Figs. 199, Tsunamis, 362
222 Tuffs, volcanic, 451
Tidal, currents, 280 Tugela Falls, 192-93
— forces and continental drift, 491, Tundras, 124
494-95 Tyne, North, river capture by, 179,
Tidenham Bend, PI. 41 Fig. 87
Tides, 279, Fig. 145 Types of coal, 335-38
Tien Shan, 427, Fig. 222 Tyrol, earth pillars, 147, PI. 33
Till, 226
Tillite, 499 U-shaped valleys, 218, 220, Figs. 114,
Tilt blocks, 415, Fig. 221 116, PI. 51
Tilted shore lines, 27 Uinta Mts., 422-23, Fig. 220
— strata, 71, 97, Figs. 23-24, 48 Ultrametamorphism, 67, 92
— terraces, 427 Unconformity, 98-99, Figs. 47, 49,
Time-distance curves, Fig. 191 21
PI.
Time, geological, 25, 101, 104-5 — dating of orogenesis, 107
Timor, uplifted coral reefs, 326, 329 — on ground-water,
effect 131, Fig. 55
Toba, Lake, 455 Undercutting, by 162rivers,
Tobel Drun ravine, PI. 31 — sand 258
blast,
Tombolo, 299 — waves, 287, Fig. 153
Topography, inselberg, 114, 274-75, Underground waters, 126, 362
Fig. 144, PI. 65 Underthrusting, 383, Figs. 198-99
— karst, 135, Fig. 166, PI. 8 b Undertow
— current, 286
— mature, 187-88, 273, PI. 30 transport by, 293-94
— old 188-89, 274
age, Unilateral mountain ranges, 382
— young, 186, 273, Pis. 1, 36 Upper layer of crust, 371-73, Fig. 193
Uplift, coral reefs, 325-26, 329
Top-set beds, 169, Fig.
78
Torrent tract, 152 —Colorado Plateau, 421-22
Tors, granite, 76, Figs. 8, 62 — isostatic, 189-90, 238,
33, Figs.
Trachyandesite, 52 96-97
Trachyte, 52 — mountains,
Transcurrent fault, Fig. 35 — peneplains, 201,
—
377, 402-3, 410
191-92, 419, 431
Transgressions, marine, 107, 417-18, rejuvenation of
420, 428
— Uinta Mts., 422-23, 195-201
rivers,
220 Fig.
53 °
INDEX
Uprush of breakers, 285 Volcanic domes, 456-57, 462, 469,

Upthrow of fault, 79 Figs. 241-42, 249
Ural Mts., 386, Fig. 200 — earthquakes, 358, 365, 367, 463,
Uraninites, as geological timekeepers,
103, 106 — eruptions, 28, 444-45, 460-62
Ur of the Chaldees, 167 — 446-48
gases,
— glass,
49
— lake basins, 452, 454~55, pl 94
V-shaped valleys, 153, 161, fig. 71 -
Val Guif, U-shaped valley, PI. 51 — products, 443, 446-61

— rocks, 49-50,
Valley, glaciers, 206, 209, Pis. 5, 46 52, 102, Fig. 51
— tract of rivers, 152 Volcanoes, 28, 443
— Colorado Plateau, 422
Valleys, development, 143-44, '5 a ~53>
160-GG, PI. 31 — distribution, 473-77, Fig. 252
— drowned, 194, 277, 301 — Dutch F.ast Indies, Fig. 21
— modification by glaciers, 220-26, — Eastern Rift of Africa, 435, 438,
51-54
Pis. Pis. 45,88
— hanging, 221, Fig. 14, Pis. 52-53,
— Great Basin, 422
69
1
— Lesser Fig. 213

Antilles,
— U-shaped, 218, 220, Figs. 114, — Mexican Plateau, 421
16, PI. — mud, 351-52
— V-shaped,51
1
153, 161, Fig. 71

— new, 452-53
— Western Rift of Africa,
Varnish, desert, 270 435-30,
Varved sediments, 238-40, Fig. 127, Fig. 230, PI. 87
PI- 57 Vosges, 383, Figs. 203, 218
Vegetation, climatic regions, 124 Vulcanian type of eruption, 461, 467-
— -Coal Measures, 339, Fig. 179, 468, 470, Fig. 244
PI.
74
— destructive 115, PI. Wadis, 270-72, PI. 62
— marine, 315action, 4
Wager, L. R., on uplift of Himalayan
— peat,
332-33, 74 PI. peaks, 201
— protective action, 115-16, 253, Walls, batholiths, 89, Fig. 45
—
dykes, 82-83
261
— 122
soils, Wasatch Range, 422-23, Fig. 219
Veins, 142 Wash-outs, 341
Ventifacts, 258, Fig. 136 Washington, Henry Stephens (1867-
Vents, volcanic, 452, Fig. 239, 247, 1934), on average composition of
Pis. 88, 90 rocks, 35
Vesuvian type of eruption, 461, 466, Wastage, zone of glacial, 205, 216,
470, Fig. 244, PI. 2 226
Vesuvius, 445, 465-68, Figs. 246-48, Water gaps, 184, Fig. 93
Pis. 2, 92 Waterfalls, 86, 151, 156-60, 192-93,
Via Mala Gorge, 153, PI. 36 Figs. 67-69, 99, Pis. 38-39
Victoria Falls, 158-59, PI. 39 — extinct, 271
Victoria, Lake, 431-32, Figs. 226, — from hanging valleys, 221, Pis.
229 52-53
Vitrain, Vitrinite, 337, PI. 76 Watershed, 173
Volcanic activity, 29, 443-86 —
recession, 175, Figs. 82-83
— causes, 478-80, 484-85 Water table, 127, Fig. 55
— types, 460-62, Fig. 244 Wave base, 284
Volcanic, ashes, 444, 450, 470-71, Wave-built terrace, 289, Fig. 153
Fig. 237 Wave-cut platform, 289, Figs. 153-54,
— bombs, 450-51, Figs. 237-38 Pis. 12, 69
— breccias, 451 Waves, earthquake, 361-72, Figs.
— calderas, 445, 454-55, Fig- 240, 189-90
Pis. 88, 94 — passage through Earth, Figs. 190,
— cones, 452-54, Fig. 234 195
— craters, 444, Figs. 234, 247, Pis. Waves,
—
sea, 282, Fig. 147
oscillation, 283, 285, Fig. 148
88, 90-92
531
1 9 7
INDEX
Waves, pressure, 286 Wind, abrasion, 254, 256-58

— refraction, 149-50 Figs. — deflation, 254-56, Fig. 135
— translation, 285 — deposits, 260-69, Figs. 137-41
— tsunamis, 362 — transport, 253-55, Fig. 134
Wayland, E. J., on origin of rift — winnowing action, 254-55
valleys, 439, Fig. 231 Wind gap, 1 79, Fig. 86
Wear, River, incised meanders, 1 97-98, Wind-worn pebbles, 258, Fig. 136
Fig. 1 o Windows, tectonic, 393, 395, Fig. 203
Weather recorded in rocks, g4 Winnowing action of wind, 254-55
Weathering, 22, 24, 112 Woodward, John (1665-1728), 509
— chemical, 112, 116-20 Worms, geological work, 115
— desert, 269-70 Wttrm glaciation, 248, Fig. 133
— felspars, 40, 1
— ferromagnesian minerals, 118

1
Xenoliths, 91
— •
interglacial deposits, 247-48

— lateritic, 1
1
Yangtze Kiang, flood plain, 168,
— mechanical, 269 1
3, fig. 79
— products, 1 13, 117-20
1
Yardangs, 258
— spheroidal, 118, Pis. 24-25 Yare River, deflection of mouth, 298-
— valley widening, 144, 160, PI. 31 299, Fig. 160
Wegener, Alfred (1880-1930), on con- Yellowstone Falls, 158, PI. 39
tinental drift, 489, 492-96, 503, Yellowstone National Park, fossil wood,
Figs. 257, 260 141
Wells, 129-32, Fig. 54 —
—
geysers, 138, PI. 29
— artesian, 132, Fig. 56 terraces of Mammoth Hot Springs,
Werner, Abraham Gottlob ( 1 749-1817) PI. 28
erroneous views on granite, 46 ; Yenangyaung oilfield, 352, PI. 77
and basalt, 47 Yorkshire, Carboniferous Limestone,
Western Alps, 390-97, Figs. 203-7 Pis. 8, 21, 24
Western Rift Valley, 434, 436, Figs. — coast erosion, 291, Fig. 155
226, 229-30 — rivers, 179, Fig. 87
— volcanoes, 435, PI. 87 — swallow holes, PI. 26
Whin Sill, 84-85, 157, Figs. 39-40, Yosemite Falls, PI. 53
PI. 18 Yosemite Valley, 221, Fig. 115, PI. 52
Youth stage, normal cycle, 145, 186-87,
— metamorphism of limestone, Fig. F>? s 94. 95
— -
18 arid cycle, 283

Whinstone, 84
Widening of valleys, 144, 160-66, Zambesi, plateau basalts, Fig. 252
PI. 3 1 —
Victoria Falls, 158, PI. 38
Willis, Bailey, on origin of rift valleys, Zion National Park, Fig. go
439 Zone of Roots (Alps), 396-97, Fig. 204
Printed in Great Britain by

Thomas Nelson and Sons Ltd, Edinburgh
53 ^
Central Archaeological Library,
NEW DELHI.
Call No. r *7

Principles of Physical Geology

Uploaded by

Document Informationclick to expand document informationbook by arthur holmes

Document Informationclick to expand document information

Copyright:

Available Formats

Principles of Physical Geology

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Principles of Physical Geology

Uploaded by

Copyright:

Available Formats

GOVERNMENT OF INDIA

Call No._ 551 JlQrfc

ARTHUR HOLMES D.Sc. F.R.S.

THOMAS NELSON AND SONS LTD

Thomas Nelson and Sons (Canada) Ltd

Thomas Nelson and Sons

Societe Fran^aise d’Editions Nelson

First published September 1944

book on Physical Geology. I therefore readily responded when

interpretation,and must on no account be confused with the

cause of mountain building, the source of volcanic activity and

The Director of H.M. Geological Survey, for fifty subjects

and the Burmah Oil Company Limited (Plate 77b).

III THE CHANGING FACE OF THE EARTH

Volcanic and Igneous Acu\

Sunimarv of the Geological Processes

IV MATERIALS OF THE EARTH'S CRUST: MINERALS

V MATERIALS OF THE EARTH'S CRUST: COMMON

VI ARCHITECTURAL FEATURES OF THE EARTH’S CRUST 69

VII ROCKS AS THE PAGES OF EARTH HISTORY ... 93

PART II EXTERNAL PROCESSES AND THEIR EFFECTS

IX UNDERGROUND WATERS 126

X RIVER ACTION AND VALLEY DEV ELOPMENT 143

Grading of Rivers ....

XI DEVELOPMENT OF RIVER SYSTEMS AN D ASSOCIATED

Superimposed Drainage ....

I he Normal C> cle of Erosion

XIII WIND ACTION AND DESERT LANDSCAPES . . .253

XIV COASTAL SCENERY AND THE WORK OF THE SEA 277

XV LIFE AS A ROCK BUILDER ”

XVI LIFE AS A FUEL MAKER COAL AND OIL 330

The Sources of Natural Fuels 33 °

PART III INTERNAL PROCESSES AND THEIR EFFECTS

The Nature of Orogenic Belts 377

The world we live in presents an endless variety of fascinating

Poseidon assisted them by shaking the earth and Assuring the

from the heating of the air over sun-scorched lands. The

earth’s internal heat on and through the surrounding crust.

The Major Fields of Scientific Study

The questions we ask when faced with a volcano in eruption

specific fields of interest. Astronomy claims the unfathomed

From the earliest days of exploration Geography has been

of the surface relief of the globe (Geomorphology), of the seas

The Scope and Subdivisions of Geology

to divide the subject into a number

cesses or machinery of the earth, by means of which changes of

the succession of these changes, or the history of the earth

the growth and movements of plants and animals and the ;

movements of hot materials inside the earth, as witnessed by

autobiography with a story to unfold, if only he can read the

the history of the earth. Moreover, it is familiar knowledge

SUGGESTIONS FOR FURTHER READING

THE SHAPE AND SURFACE RELIEF OF THE

animals. Myriads of these minute organisms are present in

our conception of the crust and of what lies beneath. It is

(a) Light rocks, including granite and related types, having

__ ... . . the mass of any volume of the substance

than in granite (70 per cent.). In the heavier rocks of this