l"' )■ 
 
 '*','• ■['.' 0;.. 1 , ■ .■ .,■,•'• •,', 
 
 
 .w!:!niim^;ii-ri!;-M:'''!l!'';', ''■■':•■■■' 
 Mh5)!;!'.!i'l!i>>f^'l!|i;)'!!!i;;.::M; ■: 
 
 ! ( 
 
 * \ ' 
 
 ,^i'.M'!- 
 

 •v ■■' I 
 
 A V 
 
 o 0^ 
 
 
 xV 
 
 
 
 V^^^ 
 
 ^A V^' 
 
 ..-^^ 
 
 -^o 
 
 '^^■ 
 
 ^ ,^^ 
 
 .nX^' '-. 
 
 
 ^. .~^ 
 
 
 .-f* 
 
 
 A^^ 
 
 ^ 
 
 C-^b 
 
PHYSICAL GEOGRAPHY. 
 
ATLASES OF PHYSICAL GEOGRAPHY. 
 
 lu Demy Svo, Cloth Limp, 2s., 
 
 THE POCKET ATLAS OP PHYSICAL GEOGEAPHY. 
 
 16 Maps, mounted on Guards, 
 
 In Imp. Svo, Cloth Lettered, Ss. GJ., 
 
 THE POETABLE ATLAS OP PHYSICAL GEOGRAPHY, 
 20 Maps, 11 by 13 inches, mounted on Guards, 
 
 In Imp, Svo, Cloth Lettered, Hs., 
 
 THE STUDENT'S ATLAS OP PHYSICAL GEOGRAPHY. 
 
 Twenty Maps, mounted on Guards. 
 
 With Letterpress Description and Wood EDgraving5« 
 
 By James Bryce, LL.D., F.G.S. 
 
^utuam's 3.bhant£tr Scieixa Series. 
 
 PHYSICAL GEOGRAPHY 
 
 y 
 
 BT 
 
 JOHN YOUNG, M.D., L.il.C.S. Edin., F.G.S., F.KS.E. 
 
 •» 
 
 r.EGItS PROFESSOR OF NATURAL HISTORY IN THE UNIVERSITY OF GLASGOW, 
 FORilERLT OP THE GEOLOGICAL SURVEY OF GREAT BRITAIN. 
 
 %"> 
 
 NEW YORK: 
 
 G. P. PUTNAM'S SONS, 
 
 27 AND 29 WEST TWENTY-THIRD STREET. 
 
iJH47 
 
 ■I 
 
Willst du ins Unenclliclie Schreiten 
 Geh'nur im Endliclien nach alien Seiten, 
 
 Willst cTu dich am Ganzen erquicken? 
 So musst du das Ganze im Kleinsten erblicken. 
 
 Goethe's Spruche. 
 
 ^ teacher need make no apology for adding another to 
 
 •ist of text-books, if he supplies what he believes, rightly 
 
 ongly, to be a want, if he does his work conscientiously, 
 
 and if he is content to abide by the verdict which natural 
 
 selection will make abundantly clear to himself and his 
 
 publisher. 
 
 I have endeavoured to give greater continuity to the 
 Geological argument than is usual in books on Physical 
 Geography, and in doing so it has been necessary to intro- 
 duce topics which are still under discussion. Had space 
 permitted I should have done so to a larger extent, knowing 
 by experience that students learn more from careful analyses 
 of current controversies than from the safer but less inte- 
 resting lectures which are confined to the recapitulation of 
 ** generally accepted conclusions." 
 
 Every teacher can devise questions on each chapter for 
 himself, but the private student will derive much assistance 
 from the "Two Thousand Questions on Physical Geogi-aphy," 
 by Professor Ansted. By their aid he will be able, not 
 merely to test his own progress, but to acquire the art of 
 answering questions, which, apart from its special value in 
 this age of examinations, is an admirable training in the art 
 of studying. 
 
 I am gi-eatly indebted to Mr. James Macaulay, of St. 
 Stephen's School, Glasgow, for important assistance and valu- 
 able advice in the preparation of this book. To him I am 
 
indeLted for tlie comparative Table of Centigrade and 
 ralirenheit Degrees. 
 
 My former colleagues on the Geological Survey, especially 
 Professor Geikie, Mr. James Geikie, and Mr. Whitaker, 
 will find tliat I have borrowed largely from their conversa- 
 tions, and from their published papers. I trust they will 
 forgive me for thus laying them under contribution, if I 
 liave succeeded in faii'ly representing their views. 
 
 To Professor A. C. Pamsay, Director General of the 
 Geological Survey, I am under obliga,tions which the fre- 
 quent references to him do not exhaust, and I should have 
 asked permission to dedicate this volume to him but that the 
 association of his name might have been misconstrued into 
 an unauthorised guarantee for a mere text- book. He has, 
 by his writings and his personal influence, done more than 
 any living Geologist to establish the connection, rather the 
 identity of Geology and Physical Geography, which I have 
 tried to make clear. In his and Professor Huxley's contri- 
 butions to the Proceedings of the Geological Society, the 
 future historian of Geology will recognise the convergence 
 of the geological and zoological lines of investigation. The 
 reconciliation of the physical investigations which Hutton 
 initiated, with the zoological enquiries to which Cuvier's 
 name gave for many years undue Aveiglit, is not yet complete; 
 but Professor Pamsay has pointed out the vray from the geo- 
 logical side. I shall have earned my reward if this volume 
 is found to be entitled to a share in the work which he has 
 done so much to advance. 
 
 John Young. 
 
 IlNIVERSITy OF GlASGOV,^, 
 
 November 1873. 
 
CONTENTS. 
 
 PACK 
 
 I^'TEODUCTION", ------_9 
 
 CHAPTER I. 
 
 Section I.— The Earth, ' - -, - . - 13 
 
 ,, II. — Composition of Earth's Crust, - - - 19 
 
 ,, III.— Rocks, -----. 24 
 
 ,, IV. — Sedimentary Strata, - - - 35 
 
 CHAPTER II. 
 
 Section I.— Continents, - . ... 55 
 
 ,, II.— Islands, ------ 63 
 
 ,, III. — Varieties of Land Surfaces, - - - 70 
 
 CHAPTER III. 
 
 Section I. — Proportion of Land to Water Surface, - - 102 
 
 ,, II. — Movements of Water, - - - - 118 
 
 CHAPTER IV. 
 
 Section I. — Rivers, . • - . - - 137 
 
 ,, II. — Lakes, .-.-.. 1G6 
 
 „ III. — Water in the Interior of the Earth, - - 177 
 
 CHAPTER V. 
 
 Section I. — Forms of Water in Atmosphere, - - -, 104 
 
 ,, II.— Snow and Ice, - • - • - 212 
 
8 . CONTENTS. 
 
 CHAPTER VI. fj^oa 
 
 Section I. — The Atmosphere, - - . - . 239 
 
 ,, II. — Atmospheric Circulation, .... 257 
 ,, III. — Electricity and Magnetism, ... 278 
 
 CHAPTER VII. 
 Climate and Weather, - - - - ^ - 285 
 
 CHAPTER VIII. 
 Volcanoes, Earthquakes, etc., - . - - - 207 
 
 CHAPTER IX. 
 Distribution of Plants and Animals, - - - - 317 
 
 CHAPTER X. 
 History and Distribution of Man, . - . . 345 
 
 Comparative Taele of Fahrenheit and Centigrade 
 
 Degrees, - - - - - * • SCI 
 
 Index, ...-.--- 3G2 
 
PHYSICAL GEOGRAPHY, 
 
 INTRODUCTION, 
 
 Physical Geography is tlie liistoiy of the Earth, its past 
 and present, interpreted by the light of Astronomy, Geology, 
 and Biology. It is not a mere description of the features of 
 sea and land, for these, as we now see them, are only one 
 phase in a series of events of which we do not know the 
 beginning, nor can we foretell the end. The enumeration, 
 however accurate and exhaustive, of the movements of Ocean 
 and Atmosphere is comparatively useless, unless we can arrive 
 at some general principle to which these are subordinated. 
 Nor is it enough to know the plants and animals which 
 occupy the various regions of land and water : we must seek 
 for the explanation of theii* diversity, and ascertain if the 
 distribution of organised beings in the past throws any light 
 on their relations in the present. 
 
 Physical Geography is, to adopt the elegant figure of M. 
 Guyot, the anatomy and physiology of the earth. Geology 
 deals with the anatomical problem, what are the materials of 
 which the earth's crust is composed, and appeals to chemistry 
 and physics for aid in elucidating the details of the processes 
 by which these materials come to exhibit theii' present 
 arrangement ; it illustrates, by phenomena going on around 
 us, the agencies by which the surface configuration has been 
 modified from time to time ; but with this purely anatomical 
 investigation its labours end. It is a part of the duty of the 
 zoologist and botanist to determine, by reference to their 
 comparative structure, the animals or plants which now and 
 in the past have lived on the earth, and to group them so aa 
 
10 PHYSICAL GEOGRAPHY. 
 
 to exhibit tlieir affinities. Palaeontology is a subordinate 
 section of zoology and botany ; as a science it has no place, 
 since the task of enumerating the species which formerly 
 lived and the localities in which they occurred leads to no 
 result if separated from the study of living forms, just as these 
 are meaningless without reference to the past. Astronomy 
 informs us of the behaviour of the earth as a body in space, 
 and of the influences exerted on its mass by other bodies. 
 But the bearing of these influences on the structure and 
 features of the earth's crust, and on the distribution of 
 organised beings over its surface, it is no part of the 
 astronomer's duty to discuss. 
 
 Physical Geography takes up the results achieved in all 
 these departments and proceeds to higher generalisations. It 
 shows how the behaviour of the earth as a body in space, and 
 its relations to other bodies determine the atmospheric 
 currents, and, through them, the movements of the ocean; it 
 points out how these ocean currents modify, and are afiected 
 by the tides ; it determines the extent to which the character 
 and variation of climate are dependent on secular changes. 
 The changes of sea and land, as ascertained b}'- the geologist, 
 are used to explain the movements of organised forms, and 
 the biologist finds in atmospheric, topographical, and climatal 
 influences the key to the presence or absence, the abund- 
 ance or scarcity, of particular groups in any locality. Nor is 
 man himself excluded from this wide field of research : the 
 influence of external conditions on his migration and his 
 development is an important investigation, on which depends 
 our judgment as to his prospects in any region, while our 
 contemplation of his history would be incomplete without the 
 light which physical influences shed on his moral develop- 
 ment and actions. 
 
 Thus the earliest inhabitants of western Europe of whom 
 we have knowledge were confined to the river valleys, the 
 rigorous climate of the northern hemisphere at that early 
 time restricting theii' movements, and making them the fellow- 
 citizens of the elk, reindeer, and other forms whose proper 
 home is in northern regions. The climate, whose severity 
 rendered Europe so widely different from what it is now, was 
 the indirect result of the gi'eater distance of the earth from the 
 
urrrtoDtfCTioN". 11 
 
 sun at that time, and, as in consequence of further secular 
 change, the climate improved, arctic animals retreated to 
 theii' homes, and man was enabled to spread over a larger 
 area, and to occupy it more completely, while the land con- 
 nection of Britain with Europe allowed him to reach the 
 extreme west. Without following his liistory in detail, suffice 
 it that the subsequent isolation of Britain, and the separa- 
 tion of Ireland, cutting off the inhabitants from the good 
 and evil eflects" of contact with the men of the continent, 
 permitted the uninterrupted development of a relatively 
 high civilisation in Ireland at an early period, though the 
 very fact of this isolation led to its arrest at a certain point, 
 and to its easy overthrow when greater facilities of transport 
 brought invaders of inferior cultivation. But long before 
 the human period the British area had acquired the general 
 features it now presents, and the hills and plains offered a 
 great variety of surface and capability. The tribes, isolated 
 by these irregularities, and by the thick forests and swamps 
 which then covered large tracts, unable to deal success- 
 fully with the unproductive soil which the glacial period had 
 left as a legacy to the northern part of the island, developed 
 the character common to all the inhabitants of mountainous 
 and detached regions. Tribal warfare Avent on, and slowly 
 were the earlier occupants driven by fresh invasions on the 
 east coast to the recesses of the hills, where they remained 
 as a disturbing element, effectually retarding the progress of 
 I the peaceful arts, though these advanced somewhat in the 
 richer eastern low grounds. Soon the geographical structure 
 of the country bore fimit ; the abundant minerals of the 
 palaeozoic rocks gave Britain advantages over its less richly 
 endowed neighbours, while the necessities of their position 
 maintained that maritime spirit among the people which, at 
 first of use only for war and plunder, afterwards gave Britain 
 its pre-eminence in commerce. The climate gradually became 
 milder, and the progress of agriculture still farther improved 
 it by drainage and the diminution of the forest lands, while 
 the position of the islands in the northern seas, at the meet- 
 ing point of the faunas of several marine zoological provmces, 
 added to the supply of food, and tempted men to a sea- 
 faring life. Thus the greatness of Britain, the character of 
 
12 PHYSICAL GEOGRAPHY. 
 
 her inhabitants, even the nature of her political institutions 
 have been to a large extent dependent on events long 
 anterior to the advent of man. Similar illustrations might 
 be drawn from the history of other countries, but few fiu'nish 
 a similarly connected and simple narrative. 
 
 It is impossible within the compass of a volume such as 
 this to. do more than touch incidentally on the leading 
 generalizations which are now regarded as sufficiently estab- 
 lished to be safely put before the student. It is necessary to 
 give a brief summary of the principal astronomical facts 
 which must be borne in mind in the investigation of currents, 
 climates, and seasons, and to state as concisely as 2')ossible 
 those geological processes, the operation and results of which 
 will be considered in the immediately succeeding chapters. 
 Geology and Physical Geography are so intimately associated 
 that it seems unfortunate that both terms should be retained. 
 The dismemberment of one science was useful in those earlier 
 days when, there being no recognised past in the material 
 world with which at least man had any concern, the history 
 of the earth was summed up in mineralogy and preconcep- 
 tions, but the dismemberment is now arbitrary and mis- 
 leading. 
 
CHAPTER I. 
 SECTION I. 
 
 The Earth: its Dimensions — Equatorial Protuberance — Axis of Rota- 
 tion: Limits of its Oscillation — Ecliptic: Influence of its Obliquity 
 on the Distribution of Light and Heat — Mass and Density of 
 the Earth — Oi^inions as to Internal Fluidity ; Underground 
 Temperature ; Secular Cooling — Eccentricity of the Earth's 
 Orbit ; Limit of Variation — Precession of the Equinoxes — Dis- 
 turbing Influence of Planets — The Moon : its Distance ; its 
 !Mass. . 
 
 The Earth is one of the planetary bodies which, revolving 
 round the Sun as a centre, make up the Solar system. The 
 size of these bodies is unequal, and some of them are only 
 secondarily related to the Sun as a centre, since they revolve 
 as satellites around other bodies. 
 
 1. Dimensions of the Earth. — The mean diameter of 
 the earth is 7912 miles, the equatorial or larger diameter 
 being 7925-6 miles, and the polar 7899'1. The difference 
 between these dimensions, 26*5 miles, represents the amount 
 of polar compression which gives to the globe the figure 
 of an oblate spheroid. Nor is this the only inequality, 
 since it is asserted that the equatorial diameters do not 
 all agree, a difference even to the extent of a mile exist- 
 ing between some of them. We may compare the equa- 
 torial protuberance to a layer laid on a sphere, so that 
 we have around that region of the earth a mountain mass, 
 so to speak, with slowly sloping sides. On the assump- 
 tion of the former plasticity of the earth's mass, the oblate 
 spheroidal form is that which a sphere, revolving on its own 
 axis, would naturally assume. It has been suggested that 
 the tapering of continents southward was connected with 
 the rotatiou form of the earth's mass; but; in the fi^-st place. 
 
14 PHYSICAL GEOGRAPHY. 
 
 there is no reason to believe that the earth now alters its 
 form otherwise than by the modification of the surface 
 effected by denudation and local subterranean movements; 
 in the second place, the attenuation commences to the north 
 of the equator, and some of the pyi-aniids, in place of tending 
 away from, tend towards the equator; and thirdly, marine 
 sedimentary deposits prove that at one period more or less 
 of the region of most rapid rotation was itself the seat 
 of an ocean. The stability of the axis of the earth is a 
 necessary consequence of the equatorial protuberance, since 
 any disturbing force which would be sufficient to shift the 
 axis of rotation, must be very powerful to overcome the 
 obstacle provided by this great mass. Hence, any departure 
 from the normal position of the polar axis is within very 
 narrow limits; and, in point of fact, there is a constant 
 oscillation in the endeavour to restore equilibrium. The 
 inclination of the polar axis to the plane of the earth's orbit, 
 that is, to the ecliptic, is 23|^^, but it is known that this 
 is diminishing at the rate of 48" in the century; and, after 
 attaining its maximum, will move in the opposite dii'ection, 
 the range of variation being 1° 21'. 
 
 2. Distribution of Light and Heat. — As a consequence 
 of this obliquity the earth in its movements round the sun 
 receives heat alternately to the north and to the south 
 of the equator ; so that, in place of an equatorial band 
 of warmth and two polar zones of constant and extreme 
 cold, we have alternations of temperature and more equal 
 distribution of light ; for, if the sun were always vertical to 
 the equatorial area, a considerable tract of the arctic and 
 antarctic circles would be in deep twilight, and the duration 
 of day and night w^ould be equal all over the rest of the 
 globe. This distribution of light and heat equally over the 
 north and south hemispheres promotes equilibrium in all 
 other respects. As we shall afterwards find, the movements 
 of the atmosphere and of the ocean are, doubtless, powerfully 
 affected by the rotation of the earth, but are chiefly deter- 
 mined by the physical features of the land. 
 . 3. Mass and Density of the Earth. — The mass of the 
 earth is about 259,801 millions of cubic miles, and the 
 specific gravity has been determined by experiments to be 
 
INTERNAL FLUIDITY. 15 
 
 ft'om 5 to 5-67, tlie extreme amount having heen observed 
 by Aiiy who, from experiments at the mouth and bottom of 
 a coal pit, estimated it at 6 '5 6. Intimately associated with 
 the estimate of the density of the mass is the question of its 
 internal condition. 
 
 4. Internal Fluidity. — Different views are entertained 
 at the present time as to the existence or not of a central 
 fluid mass, and the observations upon which is based 
 the inference that there must exist such a fluid core are, 
 that at 2151 feet the temperatm^e is 75°F. (24°C.), that 
 of a stratum at 17 feet being constantly 61°F. (10'5'^C.). 
 In Cornwall 32*5° C. have been recorded at a depth of 1200 
 feet, and the borings at Creuzot show that there is an in- 
 crease of temperature by 1°F. for every 55 feet of descent, 
 but after 1800 feet have been reached the increase is 1° in 
 44 feet. The existence of hot springs, issuing from vaults 
 which extend to very considerable depths, likewise point to 
 the conclusion that a very high temperature exists some 
 distance beneath the surface ; and the outpouring of lava 
 from so many orifices has hitherto always been looked upon 
 as only explicable on the assumption that a great reservoir 
 of molten matter uniformly underlay the crust. If the 
 increase of temperature goes on increasing with depth accord- 
 t ing to the calculations at Creuzot, then we should expect to 
 j find that, at 50 miles beneath the surface, the heat would be 
 I 4600^r,, that is considerably greater than is required to fuse 
 I platinum. It has been calculated that the sedimentary 
 j strata represent a thickness of 20 miles, and, on the assump- 
 tion of a uniform increase of temperature, we should find 
 that the sedimentary strata were at their deepest part 
 subjected to a temperature capable of melting brass, and 
 probably not far short of that at which gold is melted ^ so 
 ' that a very short distance comparatively beneath the crust 
 i would bring us to a region in which every kno^vTi substance 
 J would be in the fluid condition. But it is unknown how far 
 I increase of pressure has to do with the keeping matter in a 
 solid condition even at very great temperatures, and, at the 
 > same time, when it is remembered that the specific gravity of 
 the earth is estimated at 5 or 6, while none of the con- 
 stituents of the earth's crust have a higher specific gravity 
 
16 PHYSICAL GEOGRAPHY. 
 
 than 3, it is obvious that pressure and heat have opposite 
 tendencies, and that the influence of pressure is greater than 
 that of heat ; for, while the compressed rocks would give a 
 specific gravity very much higher than that of the earth's 
 mass, the expansion consequent upon the existence of heat 
 in their substance keeps it down to the more moderate limit. 
 Sir William Thomson has investigated the question of the 
 cooling of the earth, and looks upon the increase of tempera- 
 ture from the surface downwards as proof of the constant loss 
 of heat from the globe, the heat radiating into space without 
 sensibly elevating the temperature of the upper crust through 
 which it passes. The continuance of such a loss of heat 
 involves belief in the occurrence of a period at which the 
 earth was a fluid mass, and Sir William Thomson has fixed 
 that period at not less than 200 millions, nor more than 
 400 millions, of years ago ; the probability being that 100 
 millions of years is the limit of geological history, and that, 
 prior to that time the earth's surface was unfit for the main- 
 tenance of animal or vegetable life. But he does not con- 
 sider it probable that the crust was formed, as is commonly 
 assumed, by the consolidation of an outer layer : he believes 
 that this outer layer did not acquii'e its firmness till the 
 globe had become very nearly solid ; and that, as the con- 
 solidation commenced from the centre and went towards the 
 circumference, the last portion to be formed would be the 
 external crust. He agrees with Mr. Hopkins in believing 
 that within the crust, cavities or chambers exist in which 
 molten matters are found, and that it is this cavernous layer, 
 separating the external crust from the solid core, which is the 
 source of volcanoes and their attendant phenomena, whose 
 development is excited either by the breaking in of the 
 roofs, or by the introduction of matter, chiefly water, from 
 the exterior, or by the subsidence of the crust compressing 
 the fluid matters contained in the spaces. While it is certain 
 that the earth has consolidated from the fluid state, the 
 vagueness of the limit above assigned has been made ground 
 of objection, and it has been pointed out that there is no 
 means of determining the rate at which radiation has taken 
 place in the past, since at present the atmosphere checks 
 radiation in proportion to the quantity of aqueous vapour it 
 
PRECESSION OF THE EQUINOXES. 17 
 
 contains. These and other considerations have been advanced, 
 not in disproof of the statements as to the fact that there is 
 a loss of temperature, but as suggesting that our knowledge 
 is not yet sufficient to fix the rate of loss, still less the period 
 when the globe became habitable, and when it shall cease to 
 be so. The organic world does not furnish any guide to the 
 solution of the problem, and this speculative question must, 
 therefore, be left in the meantime. 
 
 6. Eccentricity of the Earth's Orbit. — The path described 
 by the earth round the sun is an ellipse, of which the sun 
 occupies one focus. The amount of eccentricity varies, so 
 that, as calculated by Mr. Carrick Moore, the difierenco 
 between the earth's greatest and least distance from the sun 
 was as given in the following table : — 
 
 250,000 years before ISOOj A.D., 4;V millions of miles. 
 
 210,000 „ ,10i 
 
 200,000 „ lOi „ 
 
 150,000 „ 6 ,, 
 
 100,000 „ 8i „ 
 
 50,000 „ 2| 
 
 „ 3 „ 
 
 The orbit, therefore, made its nearest approach to a circulai' 
 form 50,000 j'^ears ago, and is now tending again towards 
 greater eccentricity. 
 
 6. Precession of the Equinoxes. — The rotation of the 
 'earth on its axis gives the alternations of day and night ; 
 but its revolution round the sun gives with the obliquity of 
 the axis the seasonal diiferences, as well as the alternation 
 of these seasons in the northern and southern hemispheres. 
 Twice every year in the course of its revolution the earth 
 arrives at a point where the ecliptic cuts the equator, and the 
 day and night are then equal all over the earth; twice eveiy 
 year the earth arrives at a point where the sun illuminates 
 the northern hemisphere and the southern hemisphere respec- 
 tively, and when the sun has attained its greatest northern 
 limit, the summer solstice of the northern hemisjihere takes 
 place, the sun thereafter declining to the south, till, puijsing 
 the equator, the summer solstice of the southern hemisphere 
 is reached in its turn. If the revolutions of the earth round 
 the sun were performed in equal times, the solstice would 
 23 c 
 
18 PHYSICAL GEOGRAPHY. 
 
 ahvays occur at tlie same point iii the orbit. But retardation 
 makes the earth arrive at the same point successively a little 
 later at each revolution, and thus the solstice of the northern 
 hemisphere, occurring at one period when the earth is 
 nearest tlie sun, or in perihelion, gradually comes round 
 later and Liter till it coincides with the earth's greatest dis- 
 tance from the sun, or aphelion. But another influence also 
 comes into play, namely, the revolution of the apsides, the 
 long axis of the earth's elliptic orbit slowly changing its 
 direction under tlie same disturbing influences. By the 
 joint influence of precession and the revolution of the apsides, 
 the period in which the earth completes this cycle is 21,000 
 years : within that time the north pole will have passed 
 through every intermediate position between that when its 
 winter was in aphelion back again to the same point. 
 
 7. Amount of Heat received by the Earth. — The total 
 amount of heat received by the earth is in the inverse pro- 
 poixion to the minor axis of its orbit ; a larger amount, 
 therefore, is received during extreme eccentricity of the orbit 
 than when it is more nearly circular. The obliquity of the 
 eclii^tic, or the inclination of the earth's axis to the plane of 
 its orbit, varies as has been said : at its maximum the polar 
 regions would receive ■— more of heat than they do at 
 present, and at its minimum the amount of heat would be ■ 
 correspondingly diminished. It is unnecessary to S2:)eak of 
 the theoretical consequences of these various astronomical 
 conditions, since, in discussing changes of climate the actual 
 results, controlled by the physical conditions of the earth's 
 surface, will be summed up. 
 
 8. Disturbing Influences of Planets. — The disturbing 
 influences alluded to above are due to the attraction of the 
 planets Jupiter, Saturn, and Mars, superior j^lanets, as they 
 are called, whose orbits, that is to say, are external to that 
 of the earth, and by Yenus among the inferior planets, 
 or those whose orbits are internal to that of the earth. 
 
 9. The Moon. — The moon describes an elliptic orbit, 
 having the earth in one of its foci. Its mean distance from 
 the earth is 22G,000 miles, and its orbit is inclined to the 
 ecliplic at an angle of 5'' 8' 48", so that its altitude is greater 
 and less than that of the sun at its summer and winter 
 
COMPOSITION OP earth's CRUST. ID 
 
 solstices. Its mass has beoii determlnecl as -g]-.-^ of the 
 earth ; its bulk to be •Jg- of the earth, and its density -6. 
 Its rotation is performed in the same time as its revolution 
 round the earth, the lunar month or period of these moA-e- 
 ments being 29d. 12h. 44m. 2-87sec. The attraction excited 
 by this satellite on the earth, and especially its influence on 
 the tidal phenomena, will be discussed in a future chapter. 
 
 The relations of the other planetary bodies to the earth 
 belong to the domain of astronomy, from which we borroAv 
 the fact that their attraction on the earth's mass varies witb 
 their proximity to or distance from it, according as theii 
 elliptic orbits bring them nearer or carry them fax^thex' away. 
 
 SECTION II. 
 
 Composition of Earth's Crust — Table of Formations — Interpretation 
 of such Tables — Comparative Sections in Britain — Unconformity 
 — Sedimentary Formations not all Marine. 
 
 10. Composition of Earth's Crust. — Whatever may be 
 the composition of the interior of the earth, it is certain 
 that its surface consists of various materials, aiTanged either 
 in regular masses or irregularly comingled. The chemical 
 substances entering into the composition of the earth's crust 
 are, of course, all those which are known to the chemist, but 
 the proportions in which they exist are exceedingly various; 
 thus calcium carbonate is almost universally found, whereas 
 phosphorus is comparatively I'ai'e. Tlie quantity of each 
 of the elements which enter into the composition of the 
 sti'atified deposits, and their relative impoi'tance, are fairly 
 represented by the following analysis given by Roscoe from 
 the examination of palseozoic rocks. In 100 parts, by 
 weight, there are of 
 
 Oxygen, 44 to 487. Calcium, 6 "G to 0-9. 
 
 . Silicon, 22-8 to 36-2. Magnesium, 27 to O'l. 
 
 Aluminium, 9"9to 6"1. Sodium, 2-4to2*5. 
 
 Iron, 9 9 to 2'4. Potassium, 17to3'l. 
 
 The composition and mode of combination of the diffex'ent 
 elements is likewise vexy various. The discussion of these 
 
20 
 
 PHYSICAL GEOGRAPHY. 
 
 differences "belongs to the province of daemical geology, but 
 it will be necessary to point out in a general way, the 
 manner in which the different ingxedients are presented to 
 the field observer. 
 
 11. Formation of the Earth's Crust. — Commencing at the 
 surface of the earth in temperate regions, we find a variable 
 amount of soil overlying gravel, sand, and clay, all more or 
 less incoherent. Beneath these we come to harder material, 
 disposed in layers, the inclinations of which to the horizon 
 are widely different, ranging from horizontal to pei-pendicular. 
 As every one of these layers owes its existence to the dis- 
 integration of a previously consolidated mass, it is ob^dous 
 that there must be a point beyond which it is impossible for 
 us to trace the stratified rocks, a point at which the earliest 
 formed strata have disajDpeared, and that point is reached 
 in the Laurentian series; for though these have obviously 
 been deposited as sedimentary strata, yet the sources of 
 these sediments are nowhere to be recognised now. Lauren- 
 tian rocks are therefore not the oldest, but the oldest known 
 components of the earth's crust; all the succeeding forma- 
 tions, tabulated as follows, — 
 
 e P o 
 
 s 1 
 
 Cm 
 
 
 Eecent. 
 
 Pleistocene. 
 
 Pliocene. 
 
 Miocene. 
 
 Eocene. 
 
 : f 
 
 v..S^ Cretaceous. 
 S ■Wealden. 
 
 : §1 
 
 111 Oohtic 
 
 1^ 
 
 Permian. 
 Carboniferous. 
 Old Eed. 
 Silurian. 
 Cambrian. 
 ^Laurentian. 
 
 are derived each from that which had gone before. Calcula- 
 tions have been made as to the thickness of the earth's crust, 
 the greatest thickness of each of these formations, or their 
 mean thickness, according to the views of the calculator, being 
 taken and added together; the result is very serious mis- 
 apprehension of the relations of the strata to each other. 
 For although a table, such as that above, represents every 
 formation as if it lay in contact with one above and one 
 below, and though it thus suggests the presence of all these 
 layers as the normal condition of the earth's crust, it must 
 be borne in mind that it is impossible for all of tliese strata 
 to exist at every point of the eai-tli's surface, without our 
 imagining that the seas and lands of the past had alternated, 
 
i^OU^U'TlON OF fliE EAkTH's CRUST. ^1 
 
 and Bad, each of them in turn, covered the surface of the 
 globe, a su[)position which is contrary to analogy and to ex- 
 perience. A geological section or table is true only for the 
 locality at which it is taken. Movements of elevation and 
 subsidence are constantly going on at various points, and the 
 positions of sea and land are constantly shifting; it will be 
 apparent that every change in the position of land and water 
 means addition to, or diminution of, the amount of sedi- 
 mentary layers, or even the entire arrest of their deposition. 
 The persistence of continental conditions for a long period 
 necessarily prevents the formation of sedimentary accumula- 
 tions over that area, and thus we have in the British Islands 
 a remarkable difference in the succession of the rocks at 
 difibrent parts, the following being very good examples of 
 what it is desired to explain : — 
 
 COMPARATIVE SECTIONS IN UNITED KINGDOM. 
 
 North of Scotland. 
 W. ^ E. 
 
 •2 """'^n-h. H. ^^^^ds 
 
 Oolit 
 
 '•in. ^'^fi "^tf^io. 
 
 '^e. 
 
 Grajitians to Lammermoors. 
 N.W. S.E. 
 
 '% °"% . "'"'' °"tr"' . ou-^^-^"^"' 
 
 <>■ 
 
 .<i^ 
 
 LoNDOX TO Isle of "Wight. 
 N.E. S.W. 
 
 Eocc 
 
 
 Eocene. 
 Weulden. %^^ 
 
 J"' 
 
 ^^^ Wealden, 
 
 N. 
 
 K ' %. ^ . 
 
 Wales to Cornwall. 
 
 > 
 
 1 
 
 .-V 
 
 s. 
 
 
22 PHYSICAL GEOGRAPHY. 
 
 12. Unconformity of Formations. — Tlie thickness of the 
 stratified portion of crust may or may not be twenty miles, 
 but the calculations upon which it is based are futile, and 
 apt to convey an erroneous notion of physical geography. If 
 our opportunities of observation w^ere complete, we should 
 be able to form a picture at every epoch of the changes 
 of land or water; we should be able, from a comparison 
 of sedimentary deposits and the areas they occupy, to fix 
 the ]30sition of every change. But^ in the first i:)lace, the 
 geological record is not complete, denudation having re- 
 moved a great many of the strata; and, in the second place, 
 we have no guide to the absolute time which may h?vve 
 elapsed between any two events. Thus we find in Scotland 
 that the old red sandstone is covered by the oolites, W'hereas 
 in England the carboniferous epoch, the permian, and the 
 triassic intervened between these two sets of beds; but we 
 cannot tell whether the interval rei^resented by the absence 
 of deposit is greater, equal to, or less than the. interval, re- 
 presented in England by all these formations, though the 
 probability is that it was about the same. 
 
 If every one of the formations mentioned in the table 
 represents the sea bottom, and if we find that any one ot 
 these formations is in contact Avith, not one, but several of 
 those which were formed before it, it is obvious that the sea 
 bottom upon Avhich it is laid down w^as not uniform, and was 
 not made up of the same parts. This relation, which is 
 known as unconformity, may be represented by the subjoined 
 diagram, in which the. planes of the strata are indicated by 
 the slope of the printed names. 
 
 TPvANSVERSE SECTION OF SCOTLAND. 
 
 
 
 Uolitii;. 
 
 ta 
 
 ""a. 
 
 v 
 
 13. Sedimentary Formations not all Marine. — This 
 section shows a number of iniconformities, each one of which 
 repeats exactly the same steps. We have, first of all, the 
 
BEDIMEXTARY FORilATIOXS KOT ALL MARINE. 25 
 
 deposit of tlie strata from water, then elevation until 
 they have lost tlieir approximate horizontality; we have 
 them brought -within the denuding action of the waves of 
 the sea, or they may be elevated still more so as to be acted 
 on by the atmosphere; and, again submerged after their up- 
 turned edges had thus undergone considerable reduction, they 
 become the seat of fresh accumulations. The time occupied 
 in these processes is unknown; it may be long or it may be 
 short, for we cannot tell how many times these processes 
 may have been repeated, we camiot tell how many accumula- 
 tions have been laid down thus unconformably, have been 
 swept away, and upon the fresh vrorn surface of the sub- 
 jacent rocks another pile deposited. We see in fact only tho 
 last step in the process; we cannot tell how many similar 
 steps had occurred before. But from the statement that all 
 these formations represent sedimentary deposits, and by that 
 is usually meant marine accumulations, some deduction must 
 be made. Professor Ramsay has given good reason to believe 
 that the cambrian, the old red sandstone, the pcrmian, tho 
 triassic, and part of the carboniferous formations, represent 
 the remains of contmcntal areas, of land surfaces, or of land 
 locked basins in which a meagre fauna lived, and in which 
 the deposits were largely chemical. This is among the first 
 attempts that have been made to connect the continents of 
 the present with those of the extreme past; and it seems as 
 if we were very nearly able to say that the land surfaces of 
 the triassic times are continued by the present continents, 
 and that, extensive as may have been the upward and down- 
 ward movements and other changes of their surface, these 
 have not been simultaneous at all points, and thus the direct, 
 uninterrupted transmission of plant and animal life has taken 
 place as unmistakably as it has taken place in the ocean. 
 
 In a subsequent section, an attempt will be made to show 
 how a continent may gradually become developed or evolved 
 by the addition of layer to layer, by the gradual elevation of 
 each layer, as it has been formed, to become the shore of tho 
 land. 
 
21 PHYSICAL GEOGEAPHY. 
 
 SECTION III. 
 
 • 
 
 Most Common Minerals— Table of Rocks — Difficulty of Classifica- 
 tion: Transition Groups — Mechanically, Chemically, Organically 
 Formed Hocks — Hypogene Rocks. 
 
 14. Minerals. — The elements whicli enter most largely 
 into the composition of mineral masses are : hydrogen, chlo- 
 rine, sodium, potassium, oxygen, sulphur, calcium, mag- 
 nesium, carbon, silicon, aluminium, iron, fluorine, boron, 
 phosjDhorus, lithium, barium, zirconium. A mineral is a 
 chemical substance having a definite composition and" a 
 definite form; a rock is made u]y of various minerals in 
 difterent proportions, and has no definite form. It only 
 concerns the geographer to ascertain the composition and 
 structure of rock masses, or that deiDartment which the 
 geologist recognises as petrology. 
 
 15. Rocks. — Rocks may be grouped as — 
 
 A. Mechanical. 
 
 1. Sediimentaky : — 
 
 a. Arenaceous, e.g., Conglomerate. Sandstone. 
 
 b. Argillaceous, e.g., Mud. Clay. Shale. 
 
 2. ^oltan: — 
 
 Blown Sand. Dunes. 
 
 3. Subaerial: — 
 
 Superficial Moraines. 
 
 Moraine profonde. Boulder Clay. 
 
 Talus. 
 
 B. Chemical. 
 
 1. Calcareous, e.r/., Stalactite. Travertine. 
 
 2. Siliceous, e.g., Sinter. 
 
 3. Rock Salt. 
 
 4. Gypseous. 
 
 C. Organic. 
 
 1. Calcareous, e.g., Coral Reefs. Oazc. 
 
 2. Carbonaceous, e.g., Peat. Lignite. Coal. 
 
 3. Ferruginous, e.g.. Bog Ore. 
 
 D. Hypogene. 
 1. Met AMORPHIC: — 
 
 Quartzite, Ilornstone. Porphyry. 
 
 Mica Schist, Slate. Dolomite. Sei-pentine. 
 
 Gneiss, Granite. Anthracite. Graphite. 
 
ROCKS. 
 
 •J J 
 
 o p 
 
 >— .'3 - ~ O 
 
 Jt — *" P 
 
 - O P 
 
 
 5- !5 
 
 3 S?" 
 
 o 
 
 o 
 
 5j ^ 2. 
 
 E § o 
 
 
 
 re o 3 
 
 
 
 re 
 
 
 
 B c- 
 
 
 £ H5 o 
 
 
 
 JT d OJ •-( 
 
 
 
 j; - r-C 
 
 
 «< - =-3 
 
 
 ? 5 » 3 
 
 
 ^ - c^n 
 
 
 5f ?o 
 
 
 y. — J; ^ 
 ~ re (0 
 
 £2 
 
 a =♦ §• 
 
 c 
 
 E s » 
 
 
 as-g 
 
 
 
 
 
 1! 
 
 t?: « 
 
Agglomerates. 
 
 26 PHYSICAL GEOGRAPHY. 
 
 2. Igneous:— 
 
 a. Volcanic. "^ 
 
 Felspatliic Series, e.g., Tracliytes. Aslies 
 
 Augitic Series, e.r/., Dolerite, Basalt. 1 jj^eccias 
 h. Trappean. 
 
 Felstones. 
 
 Melapliyre. 
 
 16. Classification. — These groups cannot be regarded as 
 sliarply defined : in each a typical form may be selected, but, 
 excej)t among the lavas and perhaps a few granites, we do 
 not find perfectl}'- simple re23resentatives of the family char- 
 acters. The difficulty of classification will appear in the 
 sequel; but the accompanying diagram gives a general idea 
 of the mutual relations of the groups. 
 
 17. A. Mechanically Formed Rocks. — 
 
 1. Sedimentary. — The rocks formed by deposit from water 
 constitute the grea-t bulk of the stratified formations, and 
 the two ingredients, silica and alumina, which form their 
 greatest mass, characterise, according to their predominance, 
 the sandy and the clay •■series, arenaceous and argillaceous, 
 but they are seldom found pure. 
 
 The finest arenaceous rocks, the purest sands, or the 
 finest grained sandstones, present quartz masses comminuted 
 to the smallest possible size. The series commences with 
 bi-eccia, or angular fragments which have been disengaged 
 from rock faces, but have not undergone any action by 
 which their angular asperities might be removed. Water- 
 rolling is the most powerful agent in smoothing and polish- 
 ing in all directions, a pebble being, in 2D0j)ular language, 
 a rounded piece of quartz equally smooth at all points; 
 but it is convenient to use the word for any uniformly 
 rounded fragjment. The Ioniser the friction is continued 
 the smaller does the fragment become, so that the coarse 
 shingle of the sea-shore, Avhich, when consolidated, yields 
 a conglomerate, passes into gradually finer gravels — fine 
 sand being the last term in the series. The consolida- 
 tion is effected in the case of sand by compression, by in- 
 filtration of iron and lime from the strata above, or by 
 the solution of calcareous fragments, as of shells, which 
 may have been enclosed in it. Breccias and conglomerates 
 become coherent rocks by intermixture with sand, and tlic 
 
MECHANICALLY FORMED ROCKS. 27 
 
 pressure wMcli effects consolidation may be guessed from the 
 fact that the pebbles are sometimes indented at their points 
 of mutual contact. Sandstones contain varying quantities — - 
 sometimes minute, at other times considerable — of clay, iron, 
 calcareous, felspathic, and organic matters; and thus we 
 have transitions to shales, ironstones, hypogene rocks, lime- 
 stones, and coals. The debris brought into the sea by 
 streams, or derived from the Avaste of shores, is sifted in the 
 sea, and deposited according to its relative weight; gravel, 
 sand, and mud form, generally speaking, successive zones of 
 deposit parallel to the coast. Sandstones occur massive, with- 
 out stratification planes, forming the liver rock of quarry- 
 men, or divided into seams of various thickness separated 
 from each other by layers of other materials. The value of 
 sandstone as a building material, is in proportion to the 
 closeness of its grain, the presence of stratification planes 
 destroying the free working in all directions. 
 
 Clav is the characteristic material ' of aroillaceous rocks. 
 The silt of a river is a heterosfeneous substance, consistinjj 
 of fine mud, sand, organic and chemical matters, which are 
 gradually separated. The aluminous matter is always fine 
 grained, and the glacier detritus can only be said to present 
 a greater tenuity of its materials, because the sand which it 
 contains has been ground into an unusually fine powder. 
 The only pure clays are those derived from the decay of 
 felspathic rocks, the result — kaolin, porcelain clay, or meer- 
 schaum — consisting of hydrated silicate of alumina, from 
 which other ingredients have been washed out; and those 
 clays on which a similar depuratory action has been exercised 
 by vegetation. The fire-clay found along with coal seams is 
 the representative of the white clay on which jDcat rests, and 
 the principal impurity of both is a small quantity of car- 
 bonaceous matter. Clay as a stratified deposit forms shale, 
 which is usually extremely fissile; but this fine lamina- 
 tion must not- bo confounded with the division planes of 
 slate, which form a structure superinduced upon and not 
 necessarily coinciding with the planes of original stratifica- 
 tion. As an increasing cpiantity of sand would lead into 
 sandstone, so an increase in the carbonaceous or calcareous 
 admixture would give the transition from clay to coal or 
 
23 PHYSICAL GEOGRAniY. 
 
 limestone. A small quantity of limo gives the fertile marls, 
 and of sand gives to loam its pervious character. The con- 
 solidation of this last variety yields the mndstones of the 
 older formations, which are less fissile and tougher than the 
 purer shales. Lastly, the addition of volcanic ashes to shales 
 in course of formation gives a nearer or more distant aj^proach 
 to volcanic rock in proportion to the quantity introduced. 
 
 2. jEolian rocks are the blown sands of the desert and the 
 sea shore. They are sometimes, especially the latter, regu- 
 larly stratified, and shells, blown up from the beach, are 
 often found in the lamime. This kind of formation is only 
 recognisable in the present, its incoherence and the circum- 
 stances of its accumulation preventing its preservation in the 
 form in which it was first laid doAvn. 
 
 3. The term iSubaerial is intended to apply to those 
 materials which are derived from atmospheric waste, but 
 have not been reasserted in water. The talus found at the 
 foot of every cliff, consists of debris which may be washed 
 down in j)art by rain, but the quantity of water is not suffi- 
 cient to give it a stratified character. The coarser materials 
 are found at the bottom of the slope, which has the fan- 
 shape characteristic of all sediment allowed to spread with- 
 out restraint from a single point. The glacier debris will 
 be referred to in a subsequent chapter. Meanwhile, it may 
 be mentioned, that though the Moraine profonde may have 
 more or less of an alluvial aspect, an imperfect stratification 
 being discernible in it, the rearrangement of its materials 
 has not the regularity even of river deposits, and in the case 
 of the boulder clay or till, the distinction is Avell known 
 between the pell-mell aspect of the great mass, and the 
 regular stratification of those j^arts Avhich have been laid 
 down in the sea, or upon Avhich the sea has come to act. 
 
 18. B. Chemically Formed Rocks. — The limits of this 
 group are very indefinite, but well marked types of rock may 
 be found Avhich, though deposited in water or by water, have 
 their final form very unlike that of true sedimentary rocks. 
 
 1. Calcareous deposits take place in caves, and, on a small 
 scale, in cellars and under bridges, v/here water drops down 
 carrying an excess of carbonate of lime. The Avater Avhich 
 falls on the ground eA^aporates or runs off, leaving a calcareous 
 
EOCKS OF ORGANIC ORIGIIT. 29 
 
 deposit wliicli may come to form very thick layers of stalag- 
 mite. The falling drop also parts with its lime, which 
 gradually forms a dependent rod like an icicle; and the stalac- 
 tite and stalagmite may meet, finally filling up the cavern 
 with a spongy mass of variously coloured carbonate of lime, 
 in which crystallization may afterwards take place without 
 destroying the lamination. Travertine is a porous limestone 
 deposited by precipitation from the waters of calcareous 
 springs and streams. It forms masses which, as at San 
 rilijDpo, may be 250 feet thick; and a thickness of one foot 
 has been laid down in about four months. If fragments of 
 foreign matter are enclosed in it, the lime may arrange 
 itself round these in concentric masses giving a spheroidal 
 texture. 
 
 2. The Siliceous sinters found around thermal springs, are 
 precisely similar to the calcareous travertines. The silex is 
 deposited on the cooling of the warm water charged with 
 soda, which kept it in solution at a high temperature. 
 
 3 and 4. Bock salt is found in lenticular masses, 60 or 
 even 90 feet thick in England. In France the layers are 
 each of them thinner, but at Vic 180 feet of salt are found 
 in 650 feet of strata. Kock salt (sodium chloride) and 
 gypsum (anhydrite, and hydrated calcium sulphate) are 
 usually found in association ; they occur in various forma- 
 tions, but their remarkable development in the triassic strata, 
 I and others whose tints are red, yellow, or green, according 
 to the iron salt tliey contain, furnishes an important evidence 
 in favour of Professor Kamsay's theory, that these formations 
 were deposited in inland seas. Rock salt is also found in 
 considerable quantity saturating the soil of inland plains, or 
 deposited round the shores of land-locked basins. 
 
 19. C. Rocks of Organic Origin. — The segregation of 
 various substances by plants and animals is doubtless a 
 chemical process; but being determined by the living tissues 
 ' of organised bodies, and yielding rock masses of very difierent 
 forms and relations from those of the last group, it is at least 
 convenient to keep them apart. 
 
 1. Calcareous. — The coral reefs of tropical and sub-tropical 
 seas, and the isolated corals of other oceans, offer the largest 
 ratio of chemical substances to living tissues that we know 
 
30 PHYSICAL GEOGRAPHY. 
 
 of in nature. Tlie enormous thickness of some reefs makes 
 them very important members of the stratified series, while 
 their disintegration forms sedimentary deposits whose ulti- 
 mate organic origin is obscured. Their interest is increased by 
 the fact that they contain carbonate of magnesia along with 
 carbonate of lime, and thus these two minerals are found in 
 conjunction, which, as in the permian strata, are believed to 
 represent metamorphism. The oaze which covers the floor of 
 the Atlantic and other oceans, consists of about 85 per cent, 
 of calcareous matter, derived from the tests of Globigerinse 
 (Huxley); 10 per cent, of silicious matter of inorganic 
 origin, or obtained from diatoms and other lowly organisms, 
 with silicious coverings; while the remainder consists of 
 the debris of molluscs, Crustacea, and other marine animals j 
 which are enclosed in calcareous shells. This fine soft sedi- 
 ment hardens on exposure, and both in composition and 
 structure is well nigh identical with the earthy limestone 
 forming the white chalk. Mr. D. Forbes finds that white 
 chalk contains 94 to 98 per cent, of carbonate of lime, oaze 
 not more than 60 per cent. He thence concludes that oaze 
 would yield a calcareous shale. The proportion of lime 
 probably depends on depth; the generalisation may there- 
 fore in the meantime be accepted as in the main correct. 
 Lacustrine limestones or marls are formed in many small 
 basins, and are found on a large scale in Lake Superior. The 
 varieties of limestone are considerable. The white chalk has 
 already been mentioned; it is earthy in England, more 
 compact in France. The limestones of the carboniferous 
 period are seldom quite pure, giving transitions into sand- 
 stone and shale. Bich as they are in fossils, the quantity of 
 lime is greater than the fossils preserved will explain; and it 
 is probable that the great mass consists of debris of organisms 
 similar to those which have come down, while it is not 
 unlikely that the deej) seas of ancient periods contained oaze 
 similar to that of more recent times. Bituminous and fetid 
 limestones owe their characters to the presence of organic 
 matter. Limestones may be earthy, compact, or crystalline, 
 according to their purity and the conditions to which they 
 have been subjected since their formation. The oolitic and 
 pisolitic structui'es are, like the spheroidal condition of traver- 
 
ROCKS OF ORGANIC ORIGIN. 31 
 
 tine, clue to the presence of small foreign bodies; but the 
 isolation of the spherules has been effected by the rolling of 
 the first concretions to and fro, so that they come to consist 
 of concentric layers. Siliceous infiltrations may give great 
 hardness, while decomposition yields rottenstone. Conglo- 
 meratic or brecciated limestones are found under circum- 
 stances suggestive of the waste of islands in their vicinity 
 during deposition. The Broken beds of Portland have been 
 accounted for by supposing the deposit of lime on a mass of 
 vegetation, whose decay and collapse fractured the layer, 
 while continued deposit of lime recemented the whole. 
 
 2. The carbonaceous group is the most important econo- 
 mically. The following table, given by Dr. Percy, illustrates 
 the composition of various members of the coal group, the 
 ?arbon being taken as 100. The last column shows tho 
 excess of hydrogen above that required to make water : — 
 
 Carbon. 
 
 Excess of 
 Hydrogeu 
 
 ^yoocl, : loo 8307 1218 i-8o 
 
 Peat, 
 Ligiiitc, 
 
 10 Yard coal, Stafford, 
 
 Steam coal, Tyne, 
 
 Pentref elin coal, S. Wales, . . 
 Anthracite, Peiinsjdvania, . . . 
 
 Oxygen. 
 
 Hydrosen 
 
 83 07 
 
 12-18 
 
 55-67 
 
 9-85 
 
 42-42 
 
 8-37 
 
 21-23 
 
 6-12 
 
 18-32 
 
 5-91 
 
 5-28 
 
 4-75 
 
 1-74 
 
 2-81 
 
 2-39 
 
 3 07 
 3-47 
 3-62 
 
 4 09 
 2G3 
 
 This table shows the decrease of the other ingredients in 
 proportion to carbon, anthracite being the extreme form of 
 the metamorphism of coal, in which the volatile matters have 
 been largely expelled. 
 
 In the conversion of wood into coal, part of the hydrogen 
 
 is eliminated with carbon, as marsh gas (CH^), a part unites 
 
 i with oxygen to form water, and part of the remaining 
 
 (oxygen unites with the carbon to form carbonic acid (CgO^). 
 
 Coal is formed from the decay of vegetable matter, and 
 
 the process may go on where the vegetation grew, or the 
 
 finer debris may be washed down into pools, settling down 
 
 there to form the fine grained compact cannel or gas coal. 
 
 The ordinary bituminous coals, as they are called (though 
 
 they contain no bitumen, no substance, that is to say, which 
 
 is soluble in ether or benzole), have, for the most part, an 
 
 '.mperfect cuboidal structure, the planes by which the mass 
 
32 PHYSICAL GEOGRAPHY. 
 
 is divided being vertical to those of stratification. Tliese 
 joints result from the pressure to which the once soft mass 
 DQ has been subjected, the bulk of the original vege- 
 
 ^ table matter having been, it is calculated, eight or 
 
 I s twelve times that of the coal. The seams usually 
 %'^ rest on fire-clay, the soil on which the vegetation 
 B OQ flourished, while upwards they may pass into car- 
 ^ » bonaceous shales. But the coal itself is remark- 
 g I able for its freedom from sedimentary materials, a 
 P condition only compatible with its formation be- 
 § yond the reach of sea or rivers. The occasional 
 intersection of coal fields by bands of gravel, the 
 KQ ordinary debris of a river channel, confirms the 
 B i view that the coal swamps were analogous to those 
 11 of a delta like that of the Mississippi, the plants 
 ^ a^ decaying on the islands between the streams, being 
 o I! protected from incursions of alluvium by the thick 
 g- 1 undergrowth all round. The subaqueous deposit of 
 cannel coal is rendered more probable by the 
 horizontal passage of cannel into blackband iron- 
 stone, with which are associated the remains of 
 amphibians, so that the series probably was as in 
 the adjoining table. 
 
 The oil shales mentioned in the last line of this 
 diagram, are now an important article in Scottish 
 manufactures. They are ordinary shales, saturated 
 with animal matter derived from the decay of 
 minute entomostracan crustaceans and of vegetable 
 matter, the former being the characteristic inhabit- 
 ants of brackish and stagnant fresh waters. 
 
 3. Ferruginous deposits, whose position is deter- 
 mined by organic bodies, are illustrated by the bog 
 iron ore, which forms deposits sometimes of toler- 
 able size in peat swamps, concentrated on particu- 
 lar spots by the partially decaying plants. The 
 ^•" seorresfation of the iron in carboniferous times was 
 
 w o o 
 
 5 determined likewise by the vegetation. 
 
 CO p 
 
 O 3 
 
 •- a 
 a n 
 t^ a 
 
 a p 
 
 = S 
 
 ft S?^ Si. 
 
 <C J5 -S 
 
 f. v: ~*. 
 
 ^ Cti « 
 
 " S 
 
 C 3 
 
 tr n 
 Cr-q 
 
 » 
 
 The substances above mentioned form rock masses, whose 
 
HYPOGENE. 33 
 
 primitive condition is for the most part recognisable. But 
 except tliose which are now forming at the surface, none 
 are absohitely in their primitive state. Consolidation by 
 the pressure of a superjacent stratum, has effected a certain 
 change, and as pressure increases, layer being added to layer, 
 infiltrations of fluid also take place, carrying downwards, in 
 solution, substances contained in higher strata. Thns both 
 chemicad and textural changes occur in proportion to the 
 age of the deposit, Avithont, however, the characters of the 
 rock being obscured. But other changes take place beneath 
 the surface, metamorphism altering both the chemical and 
 physical composition of the masses, without, at least in 
 general, destroying their relations, while igneous fusion 
 reduces the whole to a uniform condition both of texture 
 and composition. These changes go on at different depths 
 beneath the surface. Hypogene or subterranean is therefore 
 the most convenient common designation for both kinds. 
 20. D. Hypogene: — '"' 
 
 1. Metamorpldc. — The consolidation of an ordinary sand- 
 stone still leaves the rounded form of its ultimate grains 
 recognisable; but in quartzite, or sandstone which approaches 
 more or less towards the condition of homogeneous quartz, 
 the granules become adherent or confluent, their partial solu- 
 tion having been effected by warm alkaline waters. Horn- 
 stone and lydianstone are extreme cases of this change in 
 siliceous strata containing a considerable quantity of clay. 
 
 Mica schist, chlorite schist, and the like, are sedimentary 
 rocks, some of whose ingredients have been separated out 
 and arranged in planes parallel to those of original stratifica- 
 tion. This rearrangement is known as foliation, and it 
 reaches its highest development in gneiss in v/hich quartz, 
 felspar, and mica are disposed in alternate layers. But these 
 are not extensive, the short fusiform masses showing gene- 
 ral parallelism only when looked at from a little distance. 
 The same ingredients are united in granite without the 
 slightest approach to order. Though some granites may be 
 of igneous origin, just as some rocks lithologically identical 
 with trap are of metamorphic origin, the greater part of the 
 granites are the representatives of sedimentary strata altered 
 in place. The granitic axis often spoken of in the case of 
 23 ' c 
 
34 PHYSICAL GEOGRAPHY. 
 
 mountains is tlie central, perhaps the deepest part of the 
 mass, but its presence is no proof of its antiquity, since 
 granite may be of any age, from the Laurentian down to 
 the tertiary granite of the Alps. In syenite, hornblende is 
 associated with the ingredients ; hypersthene and diallage, 
 members of the group of augitic minerals, may be present 
 in such quantity as to confer their names on the rock in 
 which they are found. 
 
 In the foregoing rochs the chemical grouping of their 
 ingredients has undergone entire change, the alteration of 
 texture being a necessary consequence. But in the true 
 slates the metamorphism consists in a primary change of 
 structure, the chemical alterations being comparatively slight. 
 Slaty cleavage is due to pressure, and consists in the re- 
 adjustment of the particles rela,tively to each other, so that 
 they split in one direction, which may or may not coincide 
 with the stratification planes. The cleavage planes are 
 straight, parallel to each other, and traverse the strata ii're- 
 spective of their curves. 
 
 The trap rocks of conspicuously igneous origin are simu- . 
 lated by certain masses of eminently felspathic character, 
 which, however, represent fine grained as well as conglome- 
 ratic sedimentary rocks, whose place they occupy, some- 
 times altcrnatino' with the unaltered strata. These rocks 
 in Soutli Ayrshire are described by Mr. James Geikie, 
 Quart. Jour. Geol. Soc, xxii. 
 
 The calcareous rocks are represented by crystalline lime- 
 stone, such as may be produced artificially by heating chalk 
 under great pressure, the carbonic acid being prevented from 
 escaping; by dolomite, in which carbonate of magnesia has 
 cither replaced carbonate of lime by infiltration from with- 
 out, or the mineral already present in the rock has been 
 gathered along particular lines or at particular points; and 
 by serj^entine, which is the extreme modification of a mag- 
 nesian limestone, the carbonates being replaced by silicates. 
 Serpentine may also result from the metamorj)liism of some 
 kinds of trap rock. 
 
 2. Igneous Hocks. — No real difierence can be established 
 between the most ancient and modern volcanic rocks, except 
 such as are consequent on the modifying power of externa} 
 
FORMATION OF SEDIMENTAHY STRATA. 35 
 
 conditions, or are connected with the absence of any con- 
 spicuous orifice of outflow. All the members of this series 
 may be divided, according to their composition, into two 
 groups, in one of which siliceous minerals j^revail, in the 
 other the alkaline bases enter more largely. The trachytes 
 contain G6 per cent, of silica, and 17 per cent, of alumina, on 
 an average; the dolerites, or basic rocks, contain 51 per 
 cent, of the former, and 14 -per cent, of the latter, with 10 
 per cent, of lime and 14 per cent, of iron and manganese. 
 The composition of the older rocks, the felstones and mela- 
 phyres, which are grouped as traps, is essentially the same, 
 the incoherent materials of the crater of outflow having been 
 removed and having left no trace of its position, but both 
 of the older types have undergone more or less change by 
 pressure and infiltration. The solid rocks have their counter- 
 part in the fragmental series of coarse and fine ashes, which 
 are siliceous or basic according to the character of the lava 
 with which they are associated. 
 
 Porphyry is a volcanic rock, in which some of the ingredi- 
 ents have crystallized so as to be prominent in the midst of 
 the matrix. Vesicular structure is due to the presence of 
 enclosed gases in the molten mass, and when these cavities 
 have become filled Avith solid matters, deposited from infil- 
 trated water, the rock becomes amygdaloidal, the vesicular 
 condition being restored if the contents are removed by 
 subsequent percolation. 
 
 Such are the princij^al kinds of rock which enter into the 
 composition of the earth's crust, so far at least as the 
 geographer requires to consider them. 
 
 SECTION IV. 
 
 Formation of Scclimcntary Strata — Disturbances of Sedimentary 
 Strata : Outcrop — Dip — Curves — Faults — Contemporaneous and 
 Intrusive Igneous Eocks — Preservation of Fossils. 
 
 21. rormation of Sedimentry Strata. — The sequence of 
 phenomena which terminates in the production of stratified 
 rocks is as follows ; Water descending from the atmosi^here 
 
36 PHYSICAL GEOGRAPHY. 
 
 loosens and removes the particles making up rock masses, 
 and it must be remembered tliat the term rock is, in geology, 
 used for any accumulation of mineral matter, whatever be 
 its density. The particles thus removed are carried by the 
 stream downwards to their resting place on the floor of the 
 ocean. But they do not travel directly to the sea; they are 
 delayed from time to time by being thrown out of the stream 
 as alluvium, or deposited in lake bottoms; in either case, 
 however, their stay is temporary, since, after a time, they 
 too are disintegrated and carried further down. Every 
 particle in its progress helps the wasting operation, and ^\e 
 have a mechanical abrading power exercised by the river in 
 proportion to the quantity and the bulk of the mineral sedi- 
 ments it carries forward. But the water contains, besides 
 the mechanically suspended matters, chemically dissolved 
 minerals, and thus the solid matter in the stream may greatly 
 exceed the apparent contents. These dissolved substances 
 may either enter into fresh combinations, giving rise to 
 chemical deposits in favourable cii'cumstances, or they may 
 reach the sea, and there assist in the formation of the solid 
 parts of marine plants and animals, and thus, in the end, 
 contribute to the formation of rock. The rain, as it descends 
 through the atmosphere, takes up various gases on its way; 
 chief amongst these is carbonic acid gas, which becomes, 
 when precipitated on the earth, a powerful solvent of cal- 
 careous matters. Absolutely pure water can do little in the 
 way of disintegration, if its action is not facilitated by 
 gravitation or by chemical change. The disintegrating influ- 
 ence of the rain is modified by the quality of the surface on 
 which it falls, being slowest u2:)on bare hard rock, into which 
 calcareous matters do not enter; and most raj)id upon soft 
 loose soils not protected by vegetation. Two extreme cases, 
 illustrative of this influence, may be referred to : the one is I 
 the rainfall on the Khasia Hills, in Bengal, which amounts 
 to 30 inches daily; it washes away the soil from the hill sides, 
 and prevents any vegetation gaining a footing; the other case 
 is that given by Lyell, who relates how deep ravines were 
 excavated in Georgia, U.S., after the cutting down of the 
 forest which had previously protected the soil from the rain- 
 fall. As rivers occupy a larger proportional area of the high 
 
POPvMATIOK OI' SEDIMENTARY STRATA. S7 
 
 gi'ounds than do the conjoined streams in the low grounds, 
 where the waters are confined in narrow but deep channels, 
 granular disintegration effected by the atmosphere is greatest 
 in the high grounds. By atmospheric denudation is meant 
 al] the waste that is wrought by means of the moisture con- 
 tained in the atmosphere. The form of water varies; the 
 small fine rain does more than the soaking mist in the way 
 of removal; the rapid torrent, hurrying forward numerous 
 pebbles, is efficient both in disintegration and in transport; 
 but in temperate regions the expansive power of frost is 
 one of the most efficient agents in breaking up and remov- 
 ing even the hardest rocks, while, in some tropical and sub- 
 tropical regions, the heat of the sun plays a very important 
 part. The large glacier, formed by compression of the snow 
 into a stream of solid ice, at once disintegrates the surface of 
 its channel, and removes the rubbish; and the features of the 
 country over which ice has travelled, either in the valley 
 streams, kno^vn as glaciers, or in the shape of a great sheet 
 of land ice, are characteristic and easily recognised in every 
 resfion. 
 
 The action of the atmosi^here, as a denuding agent, is ver- 
 tical, that is to say, it tends to lower the vertical elevation 
 of the earth by the removal of particles from its surface. 
 The next denuding agent, in point of importance, is the 
 sea, and its action is horizontal, so that, if land were sta- 
 tionary for a sufficiently long period of time, it would, by 
 the joint action of the two, be reduced to a level at the 
 suiiace of the sea. It is believed that the denuding power 
 of the sea is greatly inferior to that of the atmosj)here, and 
 that, while the latter produces the great features of land 
 surfaces, the former is continually striving to efface them, 
 the materials of waste, whether derived from the land by 
 streams or from the shore line, being deposited in the sea 
 bottom in masses which are approximately horizontal. 
 Natural gravity effects a certain kind of arrangement, the 
 heaviest blocks being those which settle first. The succession 
 of marine deposits, theoretically stated as gi'avel next the 
 land, sand farther out, and mud farthest out of all, is in 
 general terms correct; but a serious difficulty is presented hj 
 the fact that on many coasts, and especially at the mouths of 
 
38 Physical geography. 
 
 large rivers, finer grained mud is found close up to tlio coast 
 line. An ingenious observation of Mr. David Eobertson 
 explains the difficulty: he finds that the addition of a small 
 cpiantiiy of salt water to fresh "water which contains a largo 
 quantity of excessively fine sediment, causes the precipitation 
 of the sediment to take place in a very much shorter time 
 than if it were left to itself; hence the throwing down of 
 mud at the mouths of large rivers is the consequence of an 
 alteration in the density of the water, and the rule as to the 
 gradation of deposits, as above stated, is only absolute for the 
 purely fresh Avater accumulations of lakes. 
 
 The materials found upon the floor of the ocean, passing 
 from the ordinary coast line seawards, are coarse sedi- 
 ments, finer sand, and mud — the mud, as a general rule, 
 being found in deep water; but in the very deepest parts of 
 the ocean, as in the Atlantic, the sediment consists almost 
 exclusively of the calcareous oaze derived from the disinte- 
 gration of the shells of marine animals. Following up the 
 indications thus suggested, Mr. Hull has j^ointed out that 
 the sedimentary mechanically-formed strata, consisting of 
 gravels, sands, and clay, are inverse in their quantity to 
 the calcareous. 
 
 Shore. Sea level. 
 
 Gravel, 
 
 Sand. 
 
 Mud. 
 
 Calcareous. 
 
 Thus the carboniferous strata of Scotland contain a large 
 quantity of sedimentary beds alternating with the limestone, 
 whereas the calcareous rocks of Middle England are free from 
 sedimentary admixture, and of very much greater vertical 
 thickness, the inference being that deep water prevailed 
 there, while a shallower sea had its shore in the north of 
 Scotland. The carboniferous strata of the Appalachians are 
 chiefly sedimentary, the limestones of the Mississippi basin 
 indicating that a deep ocean lay towards the west, while the 
 land was eastward where the Atlantic now is. The clialk, 
 which finds its nearest living representative in the Atlantic 
 oaze, contains many fossils which were evidently pelagic, 
 the inhabitants of a deep open ocean; while towards the 
 Bouth of France the formation has a larger admixture of 
 
Distuhbances op sedimentary strata. 39 
 
 cedimentaiy materials, and the fossils are chiefly of types 
 common m shallower waters. 
 
 22. Disturbances of Sedimentary Strata. — Sedimentary 
 strata, then, consist of masses of material, dejoosited mechani- 
 cally under the influence of gravitation, and they are divisible 
 into layers of greater or less thickness, in consequence of the 
 intervals which frequently suspend the process. Por sedi- 
 ments are not brought continuously and in equal quantities 
 from the land; they vary with the seasons, and it may haj^pen 
 that by a change in the interior of a country, the amount of 
 waste may either be very much diminished permanently, or 
 very much altered in quality; thus the conversion of a well- 
 watered district into a rainless region would at once cut ofi 
 a large part of the sediment thus derived, while, conversely, 
 the subjection of a district to a larger amount of atmospheric 
 waste, whether by a change of climate or by the cutting 
 down of timber, as in the Georgian case already referred to, 
 would gradually increase the sediment thence obtained, and 
 might introduce materials of very different character from 
 those previously brought down. Hence it is not to be 
 expected that, under ordinary circumstances, a mechanical 
 deposit in the ocean should attain any very great thickness; 
 we should expect to find that different materials were from 
 time to time brought down, giving to it vertical alternations 
 of materials. It nmst be borne in mind that the thickest 
 uninterrupted deposits of one kind of material, those, namely, 
 which characterise the old red sandstone, were, as will be 
 shown in a future chapter, probably formed in enclosed basins 
 or inland lakes, and during slow subsidence. ^... 
 
 Neither is it to be expected that the same deposit will 
 have an indefinite horizontal extension. As the coast line 
 manifests the same variety, both in texture and chemical 
 composition, as the interior, it is obvious that denudation 
 yields unlike materials at different points. This horizontal 
 variation, dependent in part upon the greater amount of 
 debris brought dov/n from one district than from another, 
 is influenced likewise by the transporting power of marine 
 currents. These have the effect of delaying the settlement 
 of the lighter materials, and of carefully sifting out the 
 detritus along the shore, so that the first rough classification 
 
40 PilVSlCAL GEOGRAtHV. 
 
 by gravitation is revised and made more exact, while, at the" 
 same time, the materials are spread over a large area. 
 
 To the variation from time to time of the materials, and j 
 to the dispersive power of currents, must be added a third 
 influence, namely, that of movements of elevation and sub- i 
 sidence. Tlie most fiivourable conditions for extensive ver- 
 tical accumulation of sediments are during subsidence, the 
 least favourable during elevation; but during subsidence the 
 distance of the shore line from the deepest water is gradually 
 increasing, and consequently the limits of particular kinds of 
 materials are alterins:. 
 
 o 
 
 In this diagram, Coast 1 represents the position of the 
 margin of a country which is undergoing submergence ; 
 Coast 2 represents the distance which the shore line has re- 
 treated in consequence of the depression. In general terms 
 we should expect the various deposits to shift their limits 
 shorev^ards, in the manner suggested by the superposition of 
 the Avords in the diagram, and thus, while the deposit of 
 mud goes on, actually without interruption, a vertical section 
 would show that, apparently, it had been interrupted, since 
 we find a layer of oaze above it. If, on the other hand, 
 elevation took place, we should find the gravel gradually 
 shifting farther and farther out as the coast line gradually 
 encroached upon what had formerly been deep water, and 
 thus we should have an aj)parent interruption, as registered 
 in the vertical succession, Avhereas, in truth, the process of 
 deposit of one material had gone on continuously. The 
 difliculty, therefore, of interpreting the events of former 
 periods in the history of the earth is increased to us by the 
 manner in which the evidence is presented. Sir Henry de 
 la Beche first pointed out that the two extreme points of any 
 one bed were not contemporaneous; that, in fact, to return to 
 the diagram above, part of the gravel, of the sand, of the 
 mud, of the oaze, were laid down at the same time, and that 
 thus our lines of contemporaneous deposit, in the strictest 
 sense of the term, would not coincide with the plains of 
 
OUTCROP. 41 
 
 stratification; tlie accompanying diagram will make this 
 intelligible. , 
 
 
 G 
 
 G 
 
 G 
 
 G G G G 
 
 
 
 
 8 
 
 S 
 
 s s s s s 
 
 
 
 
 
 
 M M M M M ^I 
 
 :m 
 
 The letters G., S., M., coiTespond to gravel, sand, mud, and 
 the figures represent the mode in which these materials are 
 laid down npon the floor of a large lake. It is obvious 
 that, as the lake gradually becomes filled np, the gravel 
 occupying a larger cubic sjDace though a less horizontal area 
 than the sand (as the sand does than the mud), the gravel will 
 gradually creep out towards the centre of the lake, and thus 
 come to repose npon the successive layers of sand which have 
 crept out along with it. While therefore when the lake has 
 been filled up, the gravel occupies one layer, the sand another, 
 and the mud a third, and while the gravel, sand, and mud 
 layers appear to be each later than the other, the latest or 
 youngest being uppermost, the real order of chronology is 
 represented by the oblique lines, G., S,, M., and the contem- 
 ])orary in point of time of M. in the sixth column is G. in 
 the third. The broad statement, therefore, that the youngest 
 strata lie npon the older, can only be taken as correct for their 
 general mass, and not as absolute for the individual parts. 
 
 23. Outcrop. — The table of formations given above (Art. 
 11), represents the order of succession in which the fossili- 
 ferous strata are found. It is formed by comparing the struc- 
 ture of different regions together, and generalising the results. 
 But it must be remembered that in no country do we find 
 the succession so complete as is here suggested, and this table 
 does not, therefore, convey an actual picture of what is found 
 in nature. It is, perhaps, impossible to avoid such a general 
 statement, but it is very necessary to keep carefully in mind 
 the fact, that the relations of these difierent formations may 
 not be truly represented in all cases. The table, as it now 
 stands, represents the opinions of a time when it was thought 
 that all these formations were marine sediments, and when 
 the existence of contemporaneous dry land was disregarded. 
 Professor Kamsay, in his Physical Geogra2)hy of Great Britain^ 
 
4:^ PHYSICAL GEOGRAPHY. 
 
 has remodelled tlie table, and lias introduced into it tlie 
 cliaracteristic conditions of each formation; thus, the Cam- 
 brian, possibly the old red sandstone, the permian, and the 
 triassic strata are, as will be afterwards more fully explained, 
 probably the records of lacustrine or mediterranean conditions; 
 and, as these conditions are only possible during a continental 
 elevation, it follows that, for a large part of the palaeozoic 
 period, in this country at least, land took the place of water, 
 and that the equivalents in point of time — the strict contem- 
 poraries of these red strata — are the marine deposits of some 
 other locality. Professor Huxley urges the employment of 
 the word homotaxis to indicate the relations of the successive 
 strata at different points of the earth's surface, and the term 
 is intended to suggest that the successive strata occupy the 
 sa,me relative position to each other wherever they may be 
 found; but that those formations to which we give the same 
 name in distant localities are not, therefore, necessarily of 
 precisely the same age; to compare, for example, the section 
 of the North of Scotland, and that of South America, as 
 given by Mr. D. Forbes : — - 
 
 Scotland. South America. 
 
 Oolitic. OoHtic. 
 
 Old E.ed Sandstone. Permian, or Triassic, 
 
 Lower Silurian. Carboniferous. 
 
 Cambrian. Devonian. 
 
 Laurentian. Sihirian^ 
 
 These two sections do not consist of the same formations, 
 nor is it probable that the particular strata to which the same 
 name is applied belong to precisely the same period of time. 
 In the first place, the devonian beds are marine, the old red 
 sandstone of Scotland being, on the other hand, a lacustrine 
 deposit ; the carboniferous strata were also marine, their 
 fossils being, some of them, identical, others only similar to 
 those found in this country; volcanic outbursts of consider- 
 able importance occurred between the carboniferous and the 
 oolitic period, and the permian or triassic strata have under- 
 gone considerable disturbance. The common facts as regards 
 these two countries are : that we have in both the history of 
 frequent oscillations, terminating in a continental condition 
 in this permian or triassic mass, and that, thereafter, depres- 
 
OUTCROP. 43 
 
 slon again took place beneath, tlie water of the oolitic sea. 
 The general resemblance of the fossils is a clue to the general 
 succession of the deposits, and the drfierences are similar in 
 kind to the diflerences which exist at the present time 
 between the types of animal life at different parts of the 
 world. If it is not possible to admit of tlie universal distri- 
 bution of the same species of fossils over the whole surface of 
 the earth, such distribution being unsupported by analogy, 
 we can only account for their occurrence at distant parts of 
 
 j the earth by their migration. Migration necessarily implies 
 difference in date, and hence the very fact of the resemblance 
 of fossils at distant localities suggests alterations in the 
 
 ' physical geography of the period. It is not drflicult for us 
 to guess the successive slow changes, whereby the geography 
 of one period assumed the characters of the next succeeding 
 time, but we can only approximately conjecture the kind or 
 direction of the changes. If we could determine them in all 
 cases with precision, we should be enabled to picture tlie 
 successive phases in the physical history of the earth, and the 
 history of its life would, thereafter, be easy. In support 
 
 ' of this view, Barrande's investigations into the fossils of 
 ■ Bohemia furnish valuable evidence; from his researches it 
 appears that, at successive stages in the silurian strata, 
 fossils occur in groups which had been extinguished in that 
 area at an earlier period. This recurrence of groups of 
 
 I 'fossils would have been impossible, without at least the 
 
 * I assumption of a large number of unwarrantable hypotheses, 
 
 I if we believe that when they disappeared from the Bohemian 
 
 I I strata, they had been absolutely extinguished. Their re- 
 currence, therefore, is clear proof that, in the interval, they 
 had only travelled into adjacent areas, returning thence 
 when the conditions again became favourable. 
 
 The most important points which it is necessary for the 
 student of physical geography to bear in mind are : (1) That 
 the surface of the earth, as it at present exists, presents an 
 inequality and irregularity of distribution of land and Avater, 
 \nd there is every reason to believe that this kind of 
 inequality has existed from the most remote past; (2) that, 
 whether by movements of elevation or depression, or by the 
 action of denuding agents, the contoui's of the land surface 
 
44 niYSICAL GEOGRAPHY, 
 
 are becoming niodifiecl, and the waste materials are travelling 
 seaward, there to be deposited in stratified layers, which at 
 some future period shall be elevated into the condition of 
 dry land, and again undergo a course of disintegration ; (3) 
 that the materials of which these strata are composed vary 
 horizontally and vertically, the variations being consequent 
 upon the difference of the physical and chemical structure of 
 a country, on the difference of movement of the earth's crust, 
 and of ocean currents ; (4) that in the past, as at the 
 present time, there were differences in the rate and amount 
 of deposit at various points, and hence the thickness of an 
 accumulation over one area may differ very importantly from 
 that at another; (5) that the time occupied by the deposit 
 of these layers of different thickness may or may not have 
 been the same ; (6) that the strata of different localities 
 are classified according to the fossils which they contain, the 
 tables showing that the types of fossils follow each other 
 in every region in the same general order, and that species, 
 even though not identical but merely similar, neverthe- 
 less present a parallel order of succession, even in distant 
 localities. 
 
 24. Dip, — Sedimentary strata are deposited on surfaces 
 which are, for the most part, nearly level; their inclination 
 is, therefore, at first slight, but they are usually found to 
 incline to the horizon at an angle which may be a right 
 angle, and in some rare cases may pass over so that the in- 
 clination becomes reversed, and the originally loAver surface 
 may form the upper surface in the new position. The 
 inclination of the beds to the horizon is the dip, and the 
 direction of the edge of the bed when it cuts the ground is 
 the strike; dip and strike are, therefore, at right angles to 
 each other; they coincide when the bed is quite horizontal, 
 and the dip disappears when it is perpendicular. The edges 
 of strata are truncated, the face they present being more or 
 less abrupt ; for the effects of denudation depend on the speed 
 of the process, and the texture of the strata. The amount of 
 materials that has been removed may often be guessed by 
 help of the dip, the planes of which, when produced, give some 
 idea of the former extension of the strata. Thus the Strath- 
 blane valley gives the following section. 
 
CURVATURE. 45 
 
 South Hill of Cami-slo. / Campsie Hill. 
 
 Stratified Trap. 
 
 Sandstone. 
 
 Ballngau Beds. 
 
 Sandstone and Conglomerate. 
 
 Stratified Trap. Hand, Gravel, Alluvium, t:/ U. Old Red Sandstone. 
 
 These beds of the Campsie Hills would, if produced, lie 
 above the South Hill, but their true position is below that 
 hill, the fault having disturbed their relative positions. The 
 
 I valley itself, however, is excavated out of the sandstones and 
 traps, which have gradually been worn back on either side, 
 their edges of outcrop becoming more and more sloping. The 
 
 I transverse section of the Tay valley is that of a broad trougli, 
 
 ' which is partly filled with upper old red sandstone beds, 
 while floor and sides are of lower old red sandstone and 
 trap rocks, which the river, in remote times, had denuded 
 
 : prior to the deposit of the ujiper old red. 
 
 N. SidlawnUls. ..• -., Ochills. S. 
 
 ^vv^^ ...•• Upper Old Red. --..^^^ '%^ 
 
 ..cx^ or, ''%/> 
 
 ><; 
 
 If tlic lines of tlic strata on citlier side wore produced as 
 shown in the dotted lines, it would become apparent that 
 a conical mass had been denuded away, the quantity then 
 removed being capable of calculation from the known 
 thickness of the beds on either side. The horizontal strata 
 in the middle rest on the denuded edges of beds which had 
 been bent into an anticlinal arch. The valley of tlie Tay is, 
 tlicrefore, a geogi-aphical valley, but a geological hill, and 
 the u])])cr is violenty unconformable on the lower old red. 
 
 25. Curvature. — Elevation and depression of limited tracts 
 of country, accomplished with more or less violence and 
 suddenness, give the simplest form of disturbance by which 
 sedimentary strata are removed from their original horizon- 
 tality; for the limited elevation gives a dome or a ridge, from 
 the centre of which the beds dip away. If the dome becomes 
 
4:3 PHYSICAL GEOGRArHY. 
 
 clenuded by sea and atmospliere, and thereafter Is submerged 
 so as to be covered Avitb marine strata again, unconformity is 
 the relation between the older and newer deposits. This 
 relation is shown in the diagram given above, and in those in 
 Art. 11, and it indicates that a period of time of unknown 
 length has intervened between the dates of the two deposits. 
 The occurrence of such unconformities between formations 
 shoAvs that there has been a great change in the physical 
 geography of the district, and that this change has been great 
 enough, and the time long enough, to allow the plants and 
 animals of the earlier period to remove, and either to return 
 modified in form during their migrations, or to be replaced 
 by a new set of organisms coming in from some other locality. 
 The curvatures are usually, however, more complicated than 
 in the case first supposed. Several adjacent elevations form 
 a succession of anticlines with intervening troughs or syn- 
 clines, and in rare cases, as the Apj)alachians and the Jura, 
 these features corresjDond to the features of the ground, in 
 place of being reversed, as in the case of the Tay valley. The 
 amount of folding to which the strata have been subjected 
 varies much; a few broad undulations, for example, form 
 the Thames valley and the Solent, the intervening ridge 
 having been denuded into a valley, the Weald like the Tay 
 occupying the site of a geological hill. In the silurian 
 districts, on the other Land, the strata are thrown into many 
 narroAV curves, and in these subordinate crumplings abound. 
 26. Faults. — The curvature of the rocks in a district is 
 usually in the inverse ratio of the fractui'es which the strata 
 liave sustained. Fractures are of two kinds, both' traceable, 
 however, to the same cause. All rocks, at least all of any 
 density, are traversed by joints, are intersected by division 
 planes, which, when well developed, cut up the rock between 
 the planes of stratification into cubes. These joints facilitate 
 the work of the quarryman and miner, the "face" of the coal 
 making the process of extraction also more economical than 
 it would otherwise be. This kind of structure is due to 
 pressure, and is essentially the same as the cleavage of slate 
 already referred to. The fracture with displacement consti- 
 tutes a fault, such as is illustrated in the section of the 
 ^trathblane valley. The rocks have been broken, and cno 
 
FAULTS. 47 
 
 side has slipped past tlie other, so that coiTesponding portions 
 come to lie at very different levels — the trap in that parti- 
 cular case forming the top of one hill, and occurring far down 
 in the opposite hill. The amount of displacement, what is 
 called the throw of the fault, varies from inches to thousands 
 of feet, and the distance to which they may be traced is 
 equally varied. But in tracing a fault line for enormous 
 distances, as in tracing a volcanic chain or an earthquake 
 movement, it does not necessarily follow that only one move- 
 ment has determined the whole : the 2:>robability that there 
 were several increases with the length of the fault, and more 
 than one parallel fracture line may become united. The 
 inverse relations of faults and contortions of the strata have 
 been ingeniously explained by Mr. J. M. Wilson, who j^oints 
 out that if a portion of the earth be elevated, and thus come 
 to occupy a greater horizontal area than it did before, it cannot 
 return to its former position so as to leave everything as it 
 was. The rock has been displaced, and the earth's surface 
 is curved; hence, when the area subsides, the rocks would 
 become folded on themselves so as to occupy the former 
 space, and contortion v/ould thus be due to subsidence of a 
 curved surfiice, while elevation might occur in an adjoining 
 district, so that the strain in the first would be relieved. 
 But if, during elevation, the convex surface of the mass 
 becomes fractured by the strain, subsidence may be accom- 
 panied by the relative displacement of the portions on either 
 side of the fracture line, and thus the whole might be again 
 accommodated within the former area without contortion, 
 and without compensating movements in adjacent districts. 
 Faults, like the metamorphosis of sedimentary strata, and 
 the fusion of igneous rocks, are doubtless in progress" at the 
 present time : great elevations, as that of the Chilian coast, 
 extensive subsidences, like those of Grecidand and of the 
 Ai'alo-Caspian area, cannot have left the strata unaffected ; 
 and geologists have shown that the smooth surfaces of rock 
 on either side of faults are scratched or scored in such a way 
 as to indicate that the displacement has been effected by 
 several movements at different periods, and not always in 
 the same dii*ection. The gi-adual development even of a 
 displacement to the extent of 2000 feet, is one of the reason3 
 
48 PHYSICAL GEOGHArilT. 
 
 why faults are not indicated at the surface by any prominent 
 features. Even had the dislocation been sudden, denudation 
 would in time have smoothed away the outstanding portion; 
 still more would denudation efface the asperity of surface 
 resulting from a gradual downward movement. But fault 
 lines do sometimes coincide with the wall of a valley for a 
 considerable distance. Thus the northern boundary of the 
 Silurian hills of South Scotland is to a large extent in a line 
 of fault. In the subjoined diagram, the names in Roman 
 letters represent the present surface ; those in italics occupy 
 the place of the strata removed since the full de^'elopment of 
 the fault. 
 
 Section at Biggae. 
 
 N.] Pentlands. Old Red Sandstone. Southern Hills. [S. 
 
 Old Red Sandstone. Old Red Sandstone, 
 Old Eed Sandstone. ^^'^ ^^^ Sandstone. / Silurian. 
 
 Traps and Sandstones. Old Eed Sandstone. 
 Silurian. Silurian : not seen. 
 
 Section S.E. or HAnniNGTOisr. 
 
 N.] Old Eed Sandstone. [S. 
 
 f Silurian. 
 Carboniferous. Old Eed Sandstone. 1^ 
 
 Tn the upper of these fijifures, the italics show the former 
 relations, the old red sandstone having once covered the 
 tops of the Silurians. Near Biggar, the Silurians are now 
 bare; but to the N.E. the old red is still found in patches 
 on the Silurians. The relative displacement near Biggar is 
 very much greater than near Haddington; but the tapering 
 fonn of a fault line is perhaps best suggested by a ground 
 plan of the Campsie district, from Stirling westwards. The 
 
 Lower Carboniferous Traps. d 
 
 Ballagan Beds. |^ 
 
 W.j Up per Old Eed Sandstone. 2 [3. 
 
 Fault Line. t^ 
 
 a> 
 
 o 
 
 •'pi^^j^^g snoisjiuoq.TC;;) joAVorj pu^ oiioijsauiiT; « 
 
PRESERVATION OF FOSSILS. 49 
 
 fault line dies out eastwards; and the carboniferous lime- 
 stone, which, west of Stirling, is found only on the soutli 
 side of the line, is there found on both sides of it, forming a 
 continuous sheet across. 
 
 27. Contemporaneous and Intrusive Igneous Rocks. — 
 The relations of the igneous rocks will be discussed in a 
 future chapter; but it may be mentioned here, that by con- 
 temporaneous trap rocks are meant those sheets of lava 
 wliich have been poured out at the surface on sedimentary 
 strata, and which are afterwards covered by other sedimen- 
 tary strata, while the rising column of molten rock, which 
 overflows at the surface, is an intrusive neck which breaks 
 through sedimentary strata. But as this column ascends 
 under pressure, portions of it may force their way between 
 subterranean strata, or into the vertical fissures which may 
 travei'se them. Thus a single volcanic outburst would, if 
 we could see the whole course of the molten matter, illus- 
 trate all these phases. The subjoined diagram will suggest 
 the relations — the space below the horizontal line being 
 occupied by stratified deposits ; — 
 
 Crater. 
 Contemporaneous Lava Flow. Level of Ground. 
 
 Interbedded Trap, "g 
 
 ^1-*' 
 
 o^^^^'' 
 
 S5 I 
 
 Volcanic focus. 
 
 Tlie contents of rocks, apart from their essential compo- 
 nents, are mineral substances which are segregated after their 
 consolidation, and organic remains, or fossils. The former 
 are the province of the mineralogist; the latter are the 
 materials with which the palaeontologist has to deal. 
 
 28. Preservation of Fossils. — The remains of plants and 
 animals are either preserved in the places where they lived, 
 or are drifted to other localities, and there preserved. The 
 completeness of the remains depends on the rapidity with 
 which they have been covered up, or on the process to which 
 they have been subjected. The mammoth was frozen into 
 the Siberian clifis, and the bodies, untouched by decav, were 
 23 o 
 
5(5 PHYSICAL GEOGRAfEtr. 
 
 perfectly preserYcd till our own time; men and animals 
 buried in peat liave been found, slirivelled but entire, and 
 the tissues, deprived of their moisture, may be perfect if the 
 body has been exposed to great heat in a dry atmosphere. 
 But for the most part decay has set in before the object is 
 covered up, or goes on afterwards ; in the latter case, we 
 should expect to find the whole of the hard parts preserved 
 in their natural relations; in the former case, the amount 
 which comes down to us depends on a variety of circum- 
 stances. Thus, shell fish dying in great depths of water 
 will slowly decay, and only the hard shells be covered up; 
 but tho shells may, in a sandy material, subsequently dis- 
 appear, infiltration of water charged with carbonic acid 
 dissolving the calcareous matter; and this action usually 
 goes on in fresh waters before the shells are covered up, so 
 that only a layer oi marl represents perhaps a very abundant 
 molluscan fauna. If the mollusc dies within reach of the 
 shore waves, it may be rolled to and fro till it is ground to 
 powder, as if it were a pebble. Animals dying on land are 
 either eaten up by other animals, or decay disintegrates even 
 their skeletons, while the ha,rdest parts may be washed into 
 rivers, rolled to and fro, and finally entombed in a very 
 mutilated state. If the carcase is at once carried ofi' by a 
 stream, it may be partly devoured as it floats, and fragments 
 ot the skeleton may thus be dropped at intervals, leavmg 
 curious puzzles for the zoologist. Icebergs every year float 
 from the extreme north, carrymg away Arctic animals which 
 could not escape, and whose remains are scattered over the 
 sea floors of temperate or even sub-tro^^ical regions. Dr. 
 Buckland experimentally demonstrated the reason for the 
 comparatively large number ot lower jaws of vertebrated 
 animals contained in fossiliferous strata. These bones have 
 very slight attachments to the trunk, and drop ofl;* therefore 
 easily. If we knew the exact equivalents in time of the 
 jaw-containing strata, we might find the reinains of their 
 skeletons, and thus combine into one parts which, in their 
 fragmentary strata, have been referred to different genera. 
 Large numbers of animals are sometimes found entombed 
 together, killed probably in shoals by floods of fresh water 
 poured into the sea^ or hj outbursts of springs charged with 
 
rrvESERVATIOX OF FOSSILS. 51 
 
 noxious substances; these two phenomena having* occurred 
 in recent times in the Bay of Eundy and in the Indian 
 Ocean. The Pampean mud contains vast quantities of 
 mammals which have sunk in the swamps, and been trampled 
 down by those which followed, eager to drink after-droughts. 
 Of animals v/hich contained no firm joarts, no remains can 
 be expected to survive; yet even of the gelatinous jelly 
 fishes or medusae, the casts have been preserved in soft sedi- 
 ments with sufficient perfection to allow the zoologist to 
 determine their affinities. But these are exceptional cases, 
 and the paucity of the remains of inferior invertebrates is 
 doubtless due to the extreme softness of their tissues. The 
 zoologist, therefore, is deprived of an important kind of 
 evidence on which to rest his speculations as to the succession 
 of life. The tissues may be preserved, retaining their 
 original chemical composition; but they may undergo meta- 
 morphism, thus the carbonate of lime in stone lilies and 
 bivalve shells may be crystallized in the characteristic 
 rhombs, the form of the shell being retained. The material 
 may entirely disappear, being replaced grain for grain by 
 other matters : thus plants and animals may show all their 
 finest structure in silica, or the structure may be lost while 
 the form is intact, if grains of sand have taken the place of 
 the organic substances. 
 
 The value of fossils is twofold; they may serve as guides 
 in the identification of particular strata, the definiteness of 
 theii* forms giving them high value for this purpose ; or 
 they may afibrd the biologist the means of filling up his 
 scheme of the classification of plants and animals, their study 
 being then an integral pai-t of botany and zoology. The 
 interpretation of fossils, then, is not a mere mechanical 
 process, but depends for its value on thorough knowledge of 
 the nearest living kindred of the extinct forms. The general 
 tables of the succession of fossilifcrous strata wevQ based for 
 the most part on the more mechanical investigation of fossils. 
 But fuller knowledge has shown that tables so constructed 
 do not toll all the truth. Thus it is cputo correct that tho 
 Silurian strata contain fossils on the whole distinct from 
 those of the old red sandstone, still more distinct from those 
 of the carboniferous. But Professor Pamsay has pointed out 
 
52 PHYSICAL GEOGHAPHY. 
 
 that even the subordinate members of the silurian series are 
 separated from each other by gaps of great importance. He 
 has tabulated the sihirians as follows, from above down- 
 wards: — 
 
 Wenlock Shale. 
 
 Break and strong unconformity. 
 
 Upper Llandovery Beds. 
 
 Break and decided unconformity. 
 
 Lower Llandovery Beds. 
 
 Large break, especially in species, and probable uncon- 
 formity. 
 
 Llandeilo and Caradoc Beds. 
 
 Break very nearly complete, both in genera and species, and 
 probable unconformity. 
 
 Tremadoc Slate. ' " ' 
 
 Break very nearly complete, both in genera and species, and 
 probable unconformity. 
 
 LiNGULA Flags. 
 
 Great differences in the fossil contents of two successive 
 groups of strata are, in some cases, associated with uncon- 
 formity, and this, as has been explained, means repeated move- 
 ment of the inferior mass of strata with some amount of 
 denudation, the whole representing the lapse of a considerable 
 period of time, and a correspondingly great change in the 
 physical geography of the region. Again, he has shown that 
 the lower greensand contains 280 sj^ecies of animals, of which 
 233 are peculiar to it, while 57, or 18 per cent., are found also 
 in the upper cretaceous. Unconformity accompanies this great 
 break in the succession of life, and the 182 species not found 
 in England migrated, or were destroyed by the geographical 
 change. Again, the occurrence of recognised terrestrial con- 
 ditions in this and other areas, as during the Cambrian, old 
 red sandstone, upper carboniferous, part of the permian, 
 trias, weald, and eocene times, presents so many interrup- 
 tions in the succession of the marine life, on which our 
 classifications were almost exclusively based. 
 
 What, then, is a formation 1 The evidence points in the 
 direction of an important change in our conception of that 
 phrase. Whereas, formerly, the idea of time was inseparable 
 from itj whereas, formerly, a formation meant a group of 
 
tRES^kVATlON OP FOSSILS. flS 
 
 deposits laid down during a particular period, and the next 
 overlying formation laid do^vn during the succeeding period, 
 the obsei'vations and speculations of Godwin Austen, E. 
 Forbes, Huxley, and Ramsay, on the older rocks, and the 
 remarkable results obtained by Carpenter, Wyville Thomson, 
 and others, in deep sea soundings at the present, indicate 
 what may be called an overlap of formations, the most 
 obvious consequence of which is that they no longer represent 
 a perfectly definite chronology. The principle has already 
 been conceded by the Geological Survey of the United 
 Kingdom in the case of the old red sandstone or devonian, 
 for these formations are thus arranged on the table published 
 by that department : — 
 
 Carboniferous. 
 
 K Upper Dill Red SaTidstone. ►_ if tt.,^»« rkM v>^a ao«^..^^r,„ iS -Upper Devonian. 
 .£ Middle Old Red Sandstone. | ^^fP^^ ^M Red Sands one.-^ .^ ^/.J^j ^e^ouiau 
 S Lower Old Red Sandstone, t^ ^"^^^^ ^^'^ ^^""^ faaudstone. s ^ ^ower Devonian. 
 
 Silurian. 
 
 Two perfectly distinct kinds of sediments and types of 
 animals, represent two perfectly distinct geographical areas, 
 the one continental, the other marine, which co-existed during 
 the interval between the silurian and the carboniferous; we 
 have here, therefore, a good case, illustrating the general 
 proposition that formations represent geographical areas, not 
 2)eriods of time. Again, the wealden land existed befoi'e and 
 after the set of deposits, which now preserve its debris, were 
 laid down, and the fact might be represented in a table, 
 thus : — 
 
 Tertiary. 
 Wealden Land. j ^l^P^^ Cretacerus. 
 
 Purbeck Land. Lower Cretaceous. 
 
 ( J urassic. 
 
 The wealden land merged into the tei-tiary, just as it was 
 itself the direct continuation of the purbeck land; oi' again, 
 to take marine deposits: 
 
 EV4.' « c! „„ Existing; Continents. 
 
 Cretaceous Scaa lortiary Continents, 
 
 l^retaceous tocaa. Wealdeu Land. 
 
54 PHYSICAL G20GRAPHY. 
 
 It may be a long time yet before the details of these re- 
 lations are elaborated. But it is right to set before the 
 student such a general vieAv as may enable him to understand 
 the drift of modern investigation, and to appreciate the con- 
 nection which exists between the geography of to-day, and 
 the geography of former periods. 
 
CHAPTER IL 
 
 SECTION I.— CONTINENTS. 
 
 Their Aren,s — Honiomorpliism — Coast Lines : their Homomorphism ; 
 Mountain Chains Parallel to Coast Lines — Evolutions of Conti- 
 nents: Great Britain; North America — Persistence of Deep 
 Oceans — Theory of an Insular Period — Influence of Variations of 
 Land and AVater Surfaces on Historj'- of ISIan. 
 
 The surface of the globe is divided between land and water, 
 tlie latter covering an area nearly three times as large as that 
 of the former. The approximate measurements for the land 
 being 51 millions of square miles, for the water 146 millions 
 of square miles, makes the total area of the globe 197 
 millions of square miles. The ratio, on this calculation, is 
 1 :2-8; the proportion given by Sir Charles Lyell, on Mr. 
 Saunders' authority, is 1 : 2-42. 
 
 29. Areas of Continents. — The land is iiTegiilarly distri- 
 buted : 1° The greater mass is found in the noi-thern hemi- 
 sphere, and in that portion of it which lies between 40° W. 
 Ion. and 150° E. Ion., the area thus indicated including the 
 European and Asiatic masses. 2° The gi-eat blocks of land 
 have their northern extremities massive, while they taper 
 towards the south. 3° The mean elevation of the continents 
 follows generally their horizontal dimensions. The following 
 table gives some of the measurements adopted by various 
 
 
 Average Ilciglit. 
 
 Area in Square Miles. 
 
 Coast Line. 
 
 Europe,. . . . 
 Asia, 
 
 670 630)1010 
 1150 lOSO ' 
 
 17,200,000 
 1 
 
 14,128,000 
 
 1 
 17,000,000 
 
 30,800 
 
 33,000 
 
 America,. .. 
 
 030 S7G 
 
 16,000,000 
 
 10,60S,COO 
 
 14,000,000 
 
 47,000 
 
 44,500 
 
 Europe,.. .. 
 
 070 030 
 
 3,550,000 
 
 2,GSS,000 
 
 3,400,000 
 
 17,200 
 
 20,000 
 
 Africa,,.. .. 
 
 IGOO 
 
 11.511,000 
 
 8,720,000 
 
 11,300,000 
 
 14,000 
 
 10,500 
 
 Australia, .. 
 
 500 
 
 1 3,413,000 
 
 2,208,000 
 
 3,500,000 
 
 7,000 
 
 7,000 
 
56 PHYSICAL GEOGRAPHY. 
 
 writers ; they are thus tabulated to show the range of 
 variations between calculations, which are, after all, mere 
 aj)proximations. 
 
 Inspection of a map or a globe will show that the land 
 and water are inverse to each other in ratio as well as in 
 form in the two hemispheres. All the continental masses, 
 including Australia, have their greatest breadth towards the 
 north, and this tendency is apparent even in both divisions 
 of the American continent. Hence the northern circumpolar 
 land may be said to be prolonged in wedges southwards, 
 while the great water cii'cle of the southern hemisphere 
 presents corresponding alternations towards the north. 
 
 30. Homomorphism. — The similarity in form of the conti- 
 nental masses is one of those resemblances which have been 
 spoken of as geographical homologies; but the importation 
 into geography of this anatomical phrase is unfortunate, since 
 we are not yet in a position to affirm that the forms have 
 been impressed in all cases by the operation of the same law, 
 however probable it may be ; aad we know that neither in 
 point of age nor geological structure, are the continents 
 identical. As, however, a compendious term is required to 
 express this relation of similarity, homomorphism is less open 
 to objection, and it is used in the following pages because it 
 implies no theor}'-, and would still continue applicable even 
 if a iniiform cause were demonstrated. 
 
 31. Coast Line. — The greater breadth of continents at ( 
 their northern extremities, and their attenuation southwards, 
 are obvious, and are carried out even in considerable detail. 
 Thus N. America and S. America repeat the same figure: 
 in both the eastern shoulder projects, culminating in Cape 
 Charles in Labrador, and Cape St. Roque in Brazil. Africa 
 projects westward, the coast line between 10° and 30° N. lat. 
 overhano-ino- the Gulf of Guinea. Thus the S. Atlantic basin 
 is bounded by two masses of symmetrical form, save that 
 their leading features are not under the same parallels of 
 latitude. The Red Sea separates Africa from Arabia, and 
 the latter country is again separated by the Persian Gulf 
 from Persia. Arabia, moreover, has a north-westerly angle, 
 the Muscat peninsula, which repeats the Somali prominence 
 of Africa, so that the entrances to the Red Sea and the 
 
COAST LINE. 57 
 
 Persian Gulf present the same angular form. Cape Comorin 
 terminates a triangular area, which the valleys of the Indus 
 and the Ganges nearly separate from Central Asia, and the 
 Malayan peninsula gives numerous examples in detail as 
 well as in its general form of the same triangular shape. 
 Australia is at present separate from Asia, but there is suj3i- 
 cient evidence that this separation is of comparatively recent 
 date, according to the geological standard of time, though in 
 the ordinary language of men it is of very remote antiquity; 
 and, bearing this in mind, it is clear that the Pacific Ocean 
 is in reality bounded by two homomoi-phic masses, the main 
 axes of which are parallel, the Central American isthmus 
 finding its counterpart in the isolated lands of Sumatra, 
 Java, and Timor. The axes of these masses — not the leading 
 mountain chains, but the lines equally dividing their area — 
 are meridional, those of S. America and Australia curving 
 westwards, N. America and Asia eastwards. Europe is excep- 
 tional to what at first appears the rule, that the northern and 
 southern masses have contrary inclinations, since that mass 
 of which Spain is the south-western extremity has, at least at 
 present, its axis convergent with that of Asia. Other exce])- 
 tions are found on a minor scale, the land pyramids of the 
 northern Mediterranean shore having theii' axes iri-egularly 
 disposed, and rarely meridional. The homomorphism of land 
 masses then, striking as its leading features are, cannot be 
 maintained as an absolute rule. The exceptions are due to 
 the irregular lines along which those influences acted to 
 which the form and character of the coast line may be traced ; 
 and in the coast lines of various countries homomorphism of 
 another kind may be detected. 
 
 The extent of coast line of the principal masses of land 
 has been stated in the table, Art. 29. Not much importance 
 can be attached to these figures from a theoretical point of 
 \iew. The configm-ation of the coast depends upon the 
 geological structure of a country, upon the length of time 
 during which it has been exposed to denudation, upon the 
 uniformity or variety in hardness of the rocks, on the influ- 
 ence of prevailing winds and marine currents, the amount of 
 rainfall, or the presence of ice — in short, on all those circum- 
 stances on which depends the wasting or denudation of the 
 
58 PHYSICAL GEOGRAPHY. 
 
 dry land. The uniform outline of Africa and Australia is 
 analogous on a large scale to the configuration of the east 
 coast of England, where the sea line, being formed of rocks 
 comparatively homogeneous in texture, shows notable inden- 
 tations only at those points where streams reach the sea from 
 the interior. Again, the serrated coast lines of Norway, 
 West Scotland, and Ireland, and of western S. America, tell 
 not merely of similar conditions as regards atmospheric waste, 
 meaning thereby the influence of rain, rivers, and ice, but 
 also of similarity in the texture of the rocks themselves, 
 Avliich are, in all these cases, highly disturbed and altered 
 strata of unequal hardness, and disposed in layers at con- 
 siderable angles, often at right angles, to the horizon. The 
 characters of the coast line are those of the general surface 
 of the country; a generalization which only expresses in 
 other y^'^ords the fact that the features of the sea bottom are 
 those of a former dry land now submerged. This relation of 
 the submarine valleys to those of the dry land will be more 
 evident v/hen the origin of valleys has been discussed; mean- 
 while the student may satisfy himself, from the atlas, that 
 the intervals between the islets off the coast of N. and S. 
 America, of ISTorway, Scotland, and other countries with 
 similarly rugged outlines, may be referred to prolongation of 
 the river courses of the dry land. 
 
 33. Mountain Chains Parallel to Coast Line. — Consider- 
 able importance has been attached by geographers to the 
 fact that, in a large number of cases, the leading mountain 
 axes are close to and parallel with the coast lines. Thus, 
 the American continent is traversed by a nearly continuous 
 line of mountains; Scandinavia has its line of heights towards 
 the western shore. The margins of the S. African central 
 basin are elevated; the Australian highest grounds are on 
 the eastern shore. But the exceptions are more important : 
 thus, the European Alps have no obvious relation to a coast 
 line; the Himalayas are far from the sea; the transverse 
 chain which, there is good reason to believe, traverses Equa- 
 torial Africa, is neither close to nor parallel with any 
 adjacent shore at the present time, though it was close to 
 the margin of the geologically recent Sahara Sea. The 
 position of the highest grounds is for the most part capable 
 
feVOLUtlON OF CONTINENTS. 59 
 
 of explanation by reference to the geological history of a 
 country; for, though it is not in all cases possible to follow 
 the evolution of the continents, still a few instances occur 
 in which the course of events is tolerably clear. 
 
 33. Evolution of Continents. — On the geological map of 
 tlie British Islands {Student's Physical Atlas, PI. XX), the 
 bands of colour which represent the formations have con- 
 siderable regularity. The oldest rocks, the laurentians, are 
 found in the north-west in the Hebrides, and these are 
 themselves formed out of the debris of pre-existing dry 
 lands, whose position, however, we cannot positively deter- 
 mine. Probably it was to the west and north-west, in the 
 position of the present Atlantic basin. This at least is cer- 
 tain, that the next succeeding formation, the cambrian, lies 
 in hollovv'^s on the surface of the laurentian, and that these 
 tof>'ether formed the floor and shore, on and acjainst which 
 the Silurians were deposited eastwards and southwards. The 
 inland seas of the old red sandstone represent another eleva- 
 tion of the sea floor along a !N.E. ancl S.TV. line, the regu- 
 larity of which is, however, broken by the valleys, into 
 which rain and rivers had fashioned the plain of marine 
 denudation, and the semi-continental area had on its south 
 shores the seas in which the devonians of S.W. England and 
 Germany were deposited. Prior to this time, by inequalities 
 of movement and by denudation, the Silurians had been 
 divided into isolated masses, those of Scotland forming three 
 bands, imperfectly parallel: the one to the N.W. of the 
 Caledonian Canal; the next the wedge Avhose apex is sea- 
 wards towards Donegal Bay, and which is se]:)arated by tho 
 Inroad middle valley of Scotland from the southern uplands. 
 This Silurian ma.ss has its axis prolonged into Ireland, and 
 the hog-backed ridge it presents sinks down to the N.E. and 
 S.W., and is there covered by the later formations. But 
 already Scotland had undergone more elevation than England, 
 for the upper silurinns, which in England had been laid 
 down around the emerging lower silurians, are not found in 
 N. Scotland, and, in the south, fringe the great bank of land 
 just mentioned. At the Pentlands, S.W. of Edinburgh, at 
 Lesmahagow and Girvan, on the north side of the lower 
 Bilurian tract, and near Kirkcudbright on the south, tho 
 
60 Physical geography. 
 
 upper Silurians rest iinconformably on the lower, tlie latter 
 having therefore been disturbed and elevated — perhaps they 
 even formed an island — before the deposit of the former. The 
 slow submersion and re-elevation of the area during carbon- 
 iferous times, did not disturb the relative heights of the two 
 countries; at least the marine carboniferous rocks of Scotland 
 have more the character of shore and shallow water deposits 
 than those of England. After the partially continental con- 
 dition which characterised the close of the palaeozoic and the 
 commencement of the mesozoic periods in this area, there is 
 a steady change in the axis of elevation of the shore line, 
 and the colour bands describe a curve the convexity of which 
 is towards the S.E., while the axis of the tertiaries is east 
 and west, as are the axes of the latest elevatory movements 
 in this part of England. In general terms, therefore, the 
 land has grown towards the S.E., and the hill ground on the 
 west has come to border the coast line, because the lower 
 ground in that direction has partly subsided, and partly 
 been removed by denudation. A similar history of develop- 
 ment is furnished by the eastern portion of the N. American 
 continent; but though the laurentian rocks, the first sedi- 
 mentary deposits formed on the shores of unknown lands, 
 form a much more complete series, the rocks subsequent to 
 the carboniferous are not so varied as in Britain. The thick 
 sedimentary strata of the Appalachians contrast with the 
 thick limestones on the west of the Mississippi basin, and 
 indicate the position of land to have been out towards the 
 Atlantic. But though this ocean was formerly the site of 
 extensive dry land, it does not follow that it ever formed 
 one continuous continent. The probability is, that the 
 changes to which reference has been made left a central 
 valley always under water, that the shores of the ocean have 
 advanced and receded, but never effaced the intervening 
 trough. Palaeontology and physical geology agree in main- 
 taining the all but certainty of a continuous land having 
 connected S. Africa with N. Europe since the trias. The 
 peculiar reptiles are of the same tyP®^> ^^^ ^-^^^ necessity 
 for them of a land surface on which to pass from the one to 
 the other locality, has compelled investigations which have 
 resulted in the above-stated conclusion. But it does not 
 
INSULAR PERIOD. 61 
 
 follow that tlie land was as continuous as it is now; tlie shift- 
 ing of the continent may have been slow, but the migration 
 has been permitted, perhaps compelled by it — just as in more 
 recent times, the shifting land of the Pacific has isolated 
 the Tasmanian, who could not even build a raft, from his 
 kindred in New Caledonia. The land on which the trees 
 and the gigantic lizards of the weald flourished stretched 
 from western England across Biscay into S. France, and the 
 chalk ocean deep to the north shoaled as it approached the 
 Pyrenees. 
 
 34. Insular Period. — Conjectures have been formed as 
 to a former insular condition of the northern hemiaj^here 
 during the carboniferous period ; but it is now satisfactorily 
 ascertained that the facts of that period are more intelligible 
 on the theory that the land was continuous, and that the 
 archipelago of small islands assumed to have existed would 
 have involved a departure from the known proportions of 
 land and water, for which there is neither proof nor analogy. 
 In his Principles of Geology, Sir Charles Lyell gives a map 
 showing those parts of Europe which have been under water 
 since the commencement of the tertiary epoch. It aj^pears 
 that very little of Europe has been dry land through all 
 that period; but the relations of existing to fossil species of 
 animals show that the dry land has for a long time occupied 
 the same general area. In S. America the same relation 
 holds, for the edentates of the present day are the represen- 
 tatives of the gigantic sloths and armadilloes of the latest 
 deposits, and the more recent fossils of New Zealand and 
 Australia belong to the same peculiar types which now 
 inhabit that area. 
 
 To admit the extreme antiquity of continents, and to 
 believe that the land and water have always exhibited the 
 same general proportions, does not necessarily involve denial 
 of the possibility of other distributions having existed in the 
 remote past; but the doctrine of uniformity requires 2:)roof 
 of any departure from the ratios which, so far as our know- 
 ledge goes, have existed from the beginning of the fossilifer- 
 ous strata; and the doctrine of evolution, which is only a 
 particular case of the law of uniformity, leads us to seek in 
 the geological changes of the* past an explanation of tJi6 
 
62 PHYSICAL GEOGRAPHY. 
 
 geograpliy of to-day, tliougli we may not always be able to 
 trace all the steps. Tlie existence of tlie Atlantic as a deep 
 water a.rea, at least since tlie clialk formation, is one of tlie 
 tilings proved by tlie N. Atlantic exploring expeditions, and 
 the characteristic forms of the Atlantic oaze have been 
 recognised by Giimbel in limestones of the palDsozoic period. 
 
 35. Influence of Variations in Land and Water Surfaces 
 on History of Man. — The importance of the facts mentioned 
 in this chapter is considerable as bearing on the history of 
 man. The greatest amount of dry land is situated in th® 
 temperate regions, which are thus occupied by the most 
 active members of the human race, to whom they yield the 
 most abundant supply of animal and vegetable food, climate 
 and soil j)ermitting the adoption of methods of continuous 
 cultivation impracticable in warmer regions. The coinci- 
 dence in the temperate regions of the chief subterranean 
 stores of iron and coal, has further contributed to render 
 these areas the most important in the history of the world, 
 since they are at once the chief seats of commerce, and are 
 in possession of the means of conveying their wares, and of 
 producing the arms by the use of which the spread of com- 
 merce and civilisation is accompanied. The continuity of the 
 Europeo-Asiatic continent permitted the uninterrupted spread 
 of the various families of mankind, and their almost universa,! 
 diffusion over the whole area has reduced to very small dimen- 
 sions the tracts occupied by the comparatively uncivilized 
 peoples of the extreme north and north-east. The isolation 
 of the southern continental masses, on the other hand, has 
 helped to maintain the tribes living in their southern extremi- 
 ties in a state of low civilisation, the condition of the South 
 Africans and the Tasmanians being alike extreme cases — the 
 one being cut off from his northern neighbours by tracts of 
 desert land, the other by a sea which he had not the means 
 of crossing. It must of course be borne in mind that 
 between the northern and southern peoples constitutional 
 differences exist, but these do not explain all the facts for 
 which physical conditions sujDply the necessary key. 
 
 The intersection of the coast line by numerous indentations, 
 as in Scotland on the small scale, in the Mediterranean on 
 the large, helps to diminish the distance of the interior froni 
 
islands; definition. 63 
 
 tlie sea, and to render interconrse easier. AVLere, as in 
 the West Highlands, neither the soil nor the people were 
 adapted for commerce, the natural features only made clan 
 v.'ai'fare easier; but the northern shores of the JMeditei-ranean 
 rendered possible the creation of the vast empires, rich in 
 commerce, and powerful in Avar, the ruins of which are still 
 grand. 
 
 Thus the geological structure of a country, its antiquity 
 as a continent or part of a continent, and the nature and 
 amount of the external influences to which it has been ex- 
 posed, have an important influence on the history of man. 
 
 SECTION II.— ISLANDS. 
 
 1° Islands — Dismemberment of Continents — Australia — Malaya — 
 Polynesian Islands — Great Britain — 2^ Islands, directly or 
 indirectly due to Volcanic Action — Coral Islands — Reefs — 
 Barriers, and Atolls — Volcanic Islands — Submarine Gravel 
 Banks. 
 
 36. Definition. — The distinction between continents and 
 islands is arbitrary, both being masses surrounded by water, 
 and inequality of size does not constitute a true difference. 
 The artificial character of the distinction is further apparent 
 when chains of islands, as the Hebrides, even Great Britain 
 and Ireland themselves, New Zealand, New Caledonia, and 
 New Guinea, at the Antipodes, are found to be successive 
 dismemberments of the adjacent larger masses of land. The 
 elevation of a limited portion of land till it projected above 
 the surface of the sea would constitute an island, which, 
 though still — as a matter of definition — in the same category 
 with a continent, belongs to a different group from the 
 isolated masses formed by the ordinary dismemberment of 
 l)arts of a coast line. It has been proposed to separate 
 islands from continents by reference to the form of the sur- 
 face, a continent being a land mass, with an inner basin, like 
 Africa or Australia. But this is practically to make an 
 exception the rule. The African continent exhibits an 
 imusually large area, for its sizo^ of horizontal strata, and 
 
64 PHYSICAL GEOGEAPIIY. 
 
 the clenudation of these has left a central basin, "by processes 
 "which have occupied long periods, but of which we do not 
 know the details. The disturbed strata of the British Islands 
 and of Canada, though equally ancient as continents, have 
 not assumed a basin form, -which, indeed, was incompatible 
 with thek' structure. 
 
 37. Classification. — Islands fall naturally into two groups, 
 those which have been part of continuous land, and those 
 which have a different origin. It is anticipating what will 
 be said in a following section, but it must be j^remised that 
 the contour of the surface of the dry land is prepared either 
 by movements proceeding from the interior of the earth's 
 crust, or by operations taking place on its surface, and these 
 two phenomena find their extreme illustration in the develop- 
 ment of islands. 
 
 The two groups are — 1. Islands which were once portions 
 of continental lands. 2. Islands which never were portions 
 of continents, but which may become associated with them 
 by extension of the mainland. 
 
 38. Islands once part of Continents. — The fii-st group 
 embraces a very large series, the majority of existing islands, 
 v/hich are capable of an ajDiDroximate chronological classifica- 
 tion by reference to the plants and animals they contain. 
 Thus, Australia represents a very ancient separation from 
 all other lands. The animals which it contains, marsupials, 
 ornithorhynchus, and echidna, once had a much more exten- 
 sive diffusion over the globe, their fossil remains being found 
 in the mesozoic strata of Europe and America. At present, 
 the opossum family [Didelphidce) are the sole representa- 
 tives of the marsupials beyond the Australian area, and they 
 are found in the Southern States, in Central America and 
 the north part of South America. The i-emaining orders, 
 the ornithorhynchus and the echidna are confined to that 
 area which includes New Cuinea, New Caledonia, New 
 Zealand, Tasmania, and Australia. If, on the map, a line 
 be drawn from the N.W. extremity of Celebes to the N.E. 
 extremity of New Zealand, and if the axis of the latter be 
 prolonged to intersect a line drawn from Adehxide on the 
 south coast of Australia, so as to touch the western shores of 
 Tasmania, a pyramidal area, with its apex at the Macquarie 
 
ISLANDS ONCE PART OF CONTINENTS. 65 
 
 Islands, will be described, which coiTesj)oiids to the position 
 and form of a continental mass, similar to the Africa of the 
 present time. The north-western boundary of this continent 
 would be formed by a line comiecting Celebes and Western 
 Australia, and passing through the strait between Lombok 
 and Bali. Of this continent, the islands included within 
 this area would be the surviving representatives, while many 
 of the islands lying to the north and east of the limits here 
 given ought, probably, also to be included The definition of 
 this ancient continent is based by Wallace on the distinct- 
 ness of its fauna from that of the immediately adjacent 
 Asiatic continent; and if we take into consideration the 
 resemblance of the land and marine forms of this region to 
 those of the mesozoic strata of Europe, good reason will be 
 found for accepting his conclusion that these South Pacific 
 Islands are the fragments of a continent which occujiied an 
 extensive area at a time when an open ocean occupied the 
 place of South-Eastern Asia. On the other hand, the deep- 
 water sea chamiel which skirts the western shores of Celebes 
 and Lombok has, on its western side, a group of islands, the 
 animals of which manifest the closest agreement with those 
 of Asia, and prove their recent separation from that mass of 
 land. A plan is subjoined of the Malay o-Polynesian Archi- 
 pelago, for the purpose of showing the leading lines of the 
 insular groups, lines which may yet prove to have been deter- 
 mined by the leading features of the earlier continent from 
 which they became detached. The lines curve eastwards in 
 Java, Sumatra; and though these islands lie to the east of 
 that ancient deep channel fixed by Wallace, it is important 
 to note that the outline from New Zealand conforms to the 
 lines of high ground in Eastern and Central Africa. 
 
 Madfigascar, though separated from Africa by a channel 
 of only 300 miles in width, has a fauna so peculiar as to 
 suggest the extreme remoteness of its separation from the 
 adjacent mainland; while Britain agrees so closely with 
 Europe in its animal and vegetable inhabitants as to sug- 
 gest, independently of other evidence, its recent isolation. 
 Further, the common characters of the types of life found on 
 its western shores with those of Scandinavia, on its south- 
 western portions with those of the Iberian peninsula, and on 
 23 B 
 
66 
 
 PHYSICAL GEOGRAPIiy. 
 
ISLANDS DUE TO VOLCANIC ACTION. 67 
 
 1 its ea.stern shores witli those of Central Europe, suggest not 
 merely a recent, hut a direct connection with the various 
 regions named; and an inspection of a submarine chart 
 informs us that the United Kingdom projects from a sub- 
 marine plateau less than 100 fathoms in depth, the limit of 
 which extends from the north coast of Spain, round and 
 beyond the west coast of Ireland, the Hebrides, and the 
 Shetland Islands, passing thence to the Gulf of Christiania, 
 and skirting the Norwegian coast to the north [Student's 
 Physical Atlas, PI. II.). Physical and zoological evidence 
 thus concur in demonstrating that Great Britain and Ireland 
 are islands recently detached from the mainl-^nd. Applying 
 the zoological test, the greater difference of the fauna of Cor- 
 sica and Sardinia from that of Prance, indicates a greater 
 antiquity for these islands than for Britain, and thus a 
 graduated scale may be constructed between Britain, in 
 which the variation of species is slight, to Ceylon and Tas- 
 mania, in which the number of peculiar forms is progressively 
 greater.^ 
 I 39. Islands Directly or Indirectly Due to Volcanic 
 Action. — The second group of islands includes two distinct 
 types, those which are obviously of volcanic origin, the rocks 
 I of which they consist being volcanic products of various 
 
 ages, and those which arc indirectly due to volcanic action. 
 
 i The Coral Islands of the Pacific Ocean may be taken as 
 
 j the most convenient illustration of the latter class. The 
 
 I general conclusions to which their study by Darwin, Dana, 
 
 , and others leads, are, that the coral reefs are very slowly 
 
 . built up, the gi'owth being estimated variously from -J^ of an 
 
 inch to Yuir of an inch in the year. Allowing for the im- 
 
 I portant help given by shell fish and other marine animals 
 
 with calcareous investments, -j^- of an inch is a large estimate, 
 
 I and gives one foot in two hundred years. But as the 
 
 I zoophyte, which secretes in its walls the limestone skeleton, 
 
 requires a certain depth of water, of which the lowest limit 
 
 1 is, on an average, twenty fathoms in the open ocean (though 
 
 a greater depth is possible in warm waters, as of the Bed 
 
 Sea) to enalDle it to live luxuriantly, and as the species 
 
 have very different limits of depth, it follows that the coral 
 
 may continue growing only so long as any change which 
 
68 PHYSICAL GEOGRAPHY, 
 
 may take place in its relative level is very slow, a sudden 
 elevation or depression being fatal to it. The presence of 
 great coral reefs on dry land, as in Florida and the Ladrone 
 Islands, indicates the elevation of the surface on which the 
 mass grew; and taking coral debris, not coral reef, as the 
 test of elevation, a relative change of level is certain if the 
 debris is found in quantity at and over twenty feet above 
 high tide mark, that being the general limit of storm- 
 heaped accumulations. That depression has occurred is 
 proved by the separation, by a deep and wide channel, of 
 a coral reef from the island to which it is attached, by the 
 existence of atolls, and by the discovery of atolls beneath 
 the surface of the sea. The atolls or lagoon islands are more 
 or less complete circles of coral reef surrounding a basin of 
 salt water, in other words, representing a barrier reef, the 
 island round which -it grew having gradually sunk and dis- 
 appeared, while the downward movement has not exceeded 
 the rate of upward growth of the coral. If subsidence con- 
 tinues long, the upward growth of the coral will diminish the 
 size of the lagoon, the place of which may come to be taken by 
 a very small coral island. Thus the paradoxical statement is 
 true, that the small size of a coral island may be proof of its 
 antiquity. Recently emerged islands are necessarily destitute 
 of vegetation, and time is required before the waste of the 
 coral provides a soil on which plants may grow. The intro- 
 duction of seeds may have taken place by their drifting from 
 some wooded island, by their being carried in the soil lodged 
 in hollows of driftwood, or attached to the feet of birds, or 
 dropped undigested by them. The amount of life is small in 
 such islands, and the variety is not great, but the distinct- 
 ness of the species from those of adjacent lands is often 
 extreme. 
 
 The barrier reefs and atolls have usually one or more 
 openings, leading into the enclosed water. These correspond 
 to the position of streams descending from the island during 
 its submersion, the coral not growing at these points where 
 water, charged with impurities, flowed over its surface. 
 Dana finds in the irregular outline of the islands another 
 jDroof of subsidence; fov, the valleys having been formed by 
 streams, and the notches TfitU whicji tLey indent the coast 
 
6uBMAfei2?l: ghavel bajx^ks. CO 
 
 being produced while the ishxiid is above the "water, the 
 existence of irregularity of outline shows that submersion has 
 not lasted long enough to allow the sea to smooth away the 
 inequality. 
 
 The lands on which the reefs grow may be either volcanic 
 or, as has been said, may be a fragment of a former continent. 
 
 Volcanic islands are either entirely composed of volcanic 
 material, or of volcanic and sedimentary materials combined. 
 And in the latter case we may again recognise a distinction, 
 the volcano, in one case, being the centre around which sedi- 
 mentary materials have accumulated, in the other an orifice 
 opened from beneath in the midst of pre-existent sedimentary 
 strata. 
 
 Such islands as Saint Paul's, Amsterdam, and, above all, 
 the temporary Graham's Island, are examples of the tjqiical 
 volcanic island, the genesis of which seems to be traceable in 
 the shoalincr of the Atlantic basin alonoj a line connectino: the 
 most westerly point of Africa with the most easterly point of 
 S. America. That actual elevation of the sea floor takes place 
 is unquestionable; but in Graham's Island, and in the sub- 
 marine Atlantic formation, we have evidence of that other 
 method of raising a cone by the superposition of lava and 
 ashes which have been poured out from a crater. 
 
 The north-eastern peninsula of Celebes furnishes an ex- 
 ample of the fusion of an originally distinct volcanic island 
 with an adjacent land mass, the elevation of both ending in 
 the ultimate junction of their bases at and above sea level. 
 
 The volcanic islands; of greater antiquity, as JNIadoira, 
 Tenerifle, Sumatra, Japan, exhibit every gradation from 
 those of which the volcanic rocks form the nucleus, to those 
 of which they are, so to speak, only later accidents. The 
 animal and vegetable population of these islands suggests 
 sundry very complicated problems, which we are not yet able 
 to solve, the distinctness of the faunas of JNIadeira, Porto 
 Santo, and the Azores, not being explicable, save by the 
 conjecture that tliey have been separated for very long 
 periods of time, during which important specific variations 
 have been brought about. 
 
 40. Submarine Gravel Banks. — Tlie last point which 
 remains to be noted regarding islands is the evidence of 
 
70 Physical geography^. 
 
 tlieir ultimate disappearance. On the cliarts of tlie Englisli 
 Cliannel, slioals occur wliicli are " awasli," tliat is to say, 
 level with the surface of the sea ; others are at various depths 
 beneath the surface, and on these accumulations of gravel 
 are found, proving their submersion. Eor gravel, the coarse 
 detritus of land, is a shore accumulation, a,nd when it occurs 
 in patches at some distance from land, it obviously could not 
 have been transported, but must have been formed where it 
 is found, and in the case of Jones' Bank, must have been 
 formed on the shore of an island, of which that shoal is the 
 submerged representative. 
 
 SECTIOISr III. 
 
 Hclief of tlie Dry Land — Its Features dependent on External Agents 
 and on Subterranean Movements — Plane of Marine Denuda- 
 tions — Valley Formation — Hilly Districts of Britain — Transverse 
 and Longitudinal Valleys — Shapes of Vallej'-s — Sections Pointed, 
 Truncated, and Pectangular, or Evenly Curved — Remains of 
 Older in Newer Valleys — Axis of Elevation does not Kemain in 
 Centre of Land Mass — Hills and Mountains — Sand Dunes : 
 Gravel Mounds ; Moraines — Linear Directions of Chains : Scan- 
 dinavia ; Ourals ; Pyrenees ; Alps ; Asia Minor ; Asia ; Africa ; 
 America — Axes of Elevation not PaD'allel — River Valleys in 
 England — Western Escarpment of Mesozoic Rocks — Forms of 
 Hills : Rounded, Serrated, Precipitous — Cliffs — Escarpments: 
 their Origin hy Denudations, not Faults, nor Marine Action — 
 Grouping of Hills into Chains — Table-Lands. 
 
 4L Relief of the Dry Land. — Hitherto the horizontal 
 aiea of the dry land has been considered. We turn now to 
 consider the varieties of surface which it presents. The 
 relief of its surface depends on two influences : the upward 
 or downward movements which originate within the crust of 
 the earth, and the denudation which sea and atmosphere 
 unite in producing, while local modifications are the result 
 of volcanic outbursts. What is above the surface of the sea 
 is, however, the same in all essential clmracters with what 
 is below j the land of the present time has been at former 
 periods the floor of ocean, and the sea-bottoms of to-day 
 will in time appear, in their turii; as dry land. Vv^e are thus 
 
tLANE OF MARINE DENUDATION. Vl 
 
 enabled to take a coinpreliensive view of tlie features of the 
 land, and to reduce the phenomena of land and submarine 
 surface to one general law. It has been found convenient to 
 treat of islands as distinct from the continents. In discuss- 
 ing the contours of the land the distinction disappears. 
 
 42. Subterranean Movements. > — All rocks that are 
 neither volcanic in their origin, nor, like coral reefs, due to 
 the direct action of animals, are sedimentary, their primitive 
 aspect being sometimes obscured by that change of chemical 
 composition and texture which is known as metamorphism. 
 The land having been accumulated under water, had its 
 materials at first arranged in definite order. That order may 
 have been disturbed by subterranean movements, Avhereby 
 the strata have become tilted, crushed, fractured, and other- 
 wise disarranged, and the mass may not be elevated above 
 the sea till these changes have taken place. But if such a 
 disturbed mass comes within the infiuence of the waves, 
 denudation smooths off its asperities, and it finally emerges 
 from the ocean with an even surface. The strata, if elevated 
 without disturbance, likewise emerge with an even sui-fixce, 
 and this we must take as the starting point in our investi- 
 gation of the features of the land. 
 
 43. Plane of Marine Denudation. — Though even, the 
 surface is not necessarily flat, probably never is flat, 'but 
 presents a gentle slope from a central position, slightly 
 elevated, if we suppose a mass to be raised ch oiovo from 
 beneath the sea. If the mass is nearly circular the slope 
 will be quaquaverscd, that is, in all directions from the 
 centre ; if it is elongated, the centre will form a ridge with 
 opposite inclinations, and sinking at either extremity into 
 the sea. The gentle seaward slope, "the plane of marine 
 denudation" (Ramsay), is the result of the horizontal action 
 of the waves on the vertically ascending land, the resultant 
 of the two forces yielding a plane whose inclination is pro- 
 portional to the speed of elevation on the one hand, and the 
 intensity of wave action on the other. As soon as land 
 appears, moisture descends on it; and as the mass enlarges 
 the moisture increases, so that at last little rills are formed. 
 The first drops of water which sought a lower level took the 
 shortest route down the slope, that is, at right angles to its 
 
72 
 
 PHYSICAL GEOGRAPHY* 
 
 axis, and thenceforth the position of a valley was detei'mlned, 
 which every subsequent increase in the size of the rill deepens 
 and widens. If two such rills are immediately adjacent, 
 the deepening of their hollow^s renders sharper the ridge of 
 ground which separates them, and thus another series of 
 slopes is produced, from which again streamlets descend by 
 the shortest course nearly at right angles to their axes; thus 
 secondary streamlets are formed, and as these go on deepen- 
 ing their channels, the ridge between them becomes lowered, 
 and we may finally have two such secondary streamlets flow- 
 ing in opposite directions, in what appears at first sight a 
 
 HlVER, receiving tributaries on both sides, flowing in the direction 
 indicated by the arrows. The primary tributaries b, b, enter 
 the main stream at right angles. The deepening of the tributary 
 valleys leaves between each pair a sloping ridge which secondary 
 streamlets descend, as in &^, 6-. If these eat back the interven- 
 ing ridge a connecting valley may result, such as that connecting 
 Z>-, b^. Several of these may unite as at c, and form a valley 
 parallel to the main one, from ■which, as it deepens, it becomes 
 separated by a ridge. 
 
 single valley, or one without a water parting. If the deepen- 
 ing of the secondary streamlets proceeds more rapidly than 
 the deepening of one of the primary valleys, the secondary 
 valley may intercept the waters of the p)rimary, and thus we 
 may have the principal drainage of a district effected by a 
 channel which is, for a large part of its course, parallel to 
 the principal river of the region. In the south of Ireland, 
 Mr. Jukes has shown that the Bandon and Lee, which appear 
 as the main drainers of the district, are in reality secondary 
 streams which have inter cejDted the primary water courses, 
 and at last rejoin a primary valley at right angles to enter the 
 ^ea. Such, in brief terms, is the theory of the origin of valleys 
 
TRANSVERSE AXD LONGITUDINAL VALLEYS. 73 
 
 by river action, or, in more general terms, by atmosj^heric 
 denudation. It appears, then, that the original contour of the 
 country is that of a flat surface or plain, and that the hills, 
 Avhether isolated or in continuous chains, are in reality jior- 
 tions of the original plain, Avhich have been left outstanding 
 between grooves excavated by atmospheric denudation. The 
 hills of a region are, in other words, very rarely due to 
 elevation, but are for the most part the result of denuda- 
 tion. In Wales, Professor Kamsay showed that the plane 
 of marine denudation is still recognizable, since a line drawn 
 from the summit of Snowdon to the sea would be above every 
 peak intermediate between the tw^o ends of the line. In 
 Scotland, both in the north and south, the same observation 
 holds good, the hilly districts having central heights which 
 are the axes of the original elevations. Thus we have Great 
 Britain divided into six hilly districts: Wales, Cumberland, 
 the southern Highlands, the region between the Grampians 
 and the Caledonian Canal, and the region lying to the north- 
 west of the Canal; and each of these has the same character, 
 consisting, namely, of a central jirincipal axis, with a series 
 of peaks and ridges gradually descending therefrom. In 
 these localities we have on a small scale the same arrange- 
 ment which is met with on a large scale in other continents, 
 namely, the Cumberland hills represent a nearly circular 
 mass of heights, while the Scottish hill masses are all arranged 
 in chains, whose direction coincides with that of the longi- 
 tudinal axes of elevation. 
 
 44 Transverse and Longitudinal Valleys. — The valleys 
 descending from a doubly sloping ridge are transverse to its 
 axis, and to any valleys which may be formed parallel to 
 that axis. Madagascar, Sardinia, and some other islands, 
 consisting of a simple ridge, exhibit only transverse valleys. 
 The map of Scotland, perhaps, best illustrates the transverse 
 and longitudinal systems. The mass of high gTOund, whoso 
 middle line runs from the Koss of Mull to midway between 
 Cape Wrath and the Pentland Frith, is grooved by streams 
 which run generally to W.N.W. and E.S.E. The Sound of 
 Mull is apparently formed by two such valleys, which have 
 lowered the intervening water parting, till practically a single 
 valley has been formed. The longitudinal valleys arc formed 
 
74 PHYSICAL GEOGRAPHY. 
 
 by the deepening of tiibutaries at a more rapid rate tlian tlie 
 main stream; tluis the tributaries are intercepted and carried 
 to N.E. and S.W., while the continuance of the lowering 
 process ultimately lays these two valleys into one, and a 
 longitudinal valley is formed, Vv^hich, if the land is depressed, 
 becomes an arm of the sea, isolating a larger or smaller tract 
 of country. If such submersion takes place slowly, the 
 ab-vading action of the waves planes down the floor of the 
 valley, so that on re-emergence we should expect to find a 
 smooth surface, from which all trace of the original water 
 2)arting has disappeared. The influence of the sea on the 
 form of such a valley is exerted in widening it and in smooth- 
 ing it down, so long as its floor remains within the action of 
 its surface waters: but there is a limit to its operation; for, 
 though there is evidence that the great valley of the Forth 
 and Clyde has been beneath the sea, a distinct water parting 
 has survived the submersion, and the streams flow in oj)po- 
 site directions. 
 
 45. Shapes of Valleys. — The water parting in the imagined 
 case of a newly elevated ridge, coincides with the summit of 
 the ridge; but, as the ridge is lowered at the points corre- 
 sponding to the head waters of streams descending in opposite 
 directions, gaps will be formed, and, as the lowering process 
 goes on, till at last the valleys become continuous, the com- 
 mon groove will become a pass. If a larger body of water, 
 or softer strata, enable the stream on one side to eat down 
 the gap more rapidly than it is being worn away on the 
 other, the water parting will become sinuous, and the sources 
 of the streams would appear on the map to interlace. 
 
 In temperate regions, the shape of such valleys as have 
 been described, is that of a triangle V, the apex of winch 
 may in the larger valleys become truncated \ /, the flat- 
 tened area representing the alluvial flats formed by the flood- 
 borne detritus. But the action of the stream at the bottom 
 of the groove goes on along with the disintegrative action of 
 the atmosphere on the higher part of the walls, a slope being 
 thus maintained on either side. In rainless regions, such as 
 that of the Zambesi, in Africa, or the Colorado district of 
 western N". America, the stream wears down its channel 
 vertically, and the shape of the gorge is more nearly rectan- 
 
EEMAIXS OF OLD VALLEYS. 75 
 
 gular I I, the channel of the Zambesi below the falls as 
 
 described by Livingstone, and the cafions of the Colorado 
 district, sometimes 3000 feet deep, being stupendous examples 
 of the power of unaided streams. Such valleys, moreover, 
 have their lines very ii-regularly disposed, manifesting little 
 subordination to the axis of greatest elevation in the dis- 
 trict. In glaciated districts on the other hand, as in the 
 British Islands, the passage of glaciers down the valleys has 
 obliterated the angles, and given the regular curve — char- 
 acteristic of ice- worn channels. The difference of form of 
 valleys thus gives a guide to their age, the convergent slopes 
 being peculiar to thorje grooves which have been excavated 
 since the ice period, whether on the sides of a hill, or in the 
 glacial detritus Avith which a wide valley has been partially 
 filled up. 
 
 46. Remains of Old Valleys. — A river valley for the 
 most pai-t has the stream flowing at its lowest level; but' in 
 temperate regions, a -river bed is sometimes above the level 
 of the lowest point, is in fact scooped out on the side of the 
 valley, and parallel to its axis. These excej^tional cases are 
 confined to ice-worn districts, and suggest that the stream, 
 or the glacier, has worn dovm sl new groove, leaving the 
 original stream channel as an index of the amount of denu- 
 dation which has taken place since it formed the sole drain 
 of the district. In the same districts we often observe that 
 the walls of the valley slope in two distinct 
 planes on either side; the upper, more gentle 
 inclination representing the original form, 
 through the floor of which the latter more 
 acute an^ded vallev has been worn. 
 
 The abrupt change in the walls of a valley, from gentle 
 slopes to almost vertical gorges, is a phenomenon chiefly 
 associated Avith glaciation, the gorges being for the most part 
 rock-worn channels, which the stream has cut for itself 
 since the glacial period, having been unable to wear 
 through the detritus with which its old channel had been 
 choked up. 
 
 The walls of fiords or submerged valleys, which indent the 
 west coast of Scotland, Norway, Ireland, and S. America, 
 arc cither continuous or interrupted; for if the submersion 
 
76 PHYSICAL GEOGl^APIn^ 
 
 only bring clown the ridges nearly to the sea level, the cols 
 become shannals, isolating larger or smaller peaks. 
 
 47. Axis of Elevation. — In the imaginary case of newly 
 elevated land, we have assumed that the axis of elevation 
 coincided with the middle line of the mass. But this is not 
 the case in land masses of any antiquity. In them there is 
 a gradual rise from the coast towards the greatest heights, 
 and these, as has been already mentioned, are usually found 
 nearer to one or other coast line. Thus, in S. America there 
 is a gradual ascent on the east side, an abrupt descent to the 
 western shore. The long slope may be either uniform, or, 
 as is more frequently the case, interrupted by terraces. It 
 is usually stated that a long slope on one side has a steeper 
 counter slope on the other side; but, from what has been 
 said regarding river valleys, it is clear that this general state- 
 ment only describes the large valleys whose ridges separate 
 them from more recently excavated grooves. The gradation, 
 therefore, in point of age among these parallel or intersecting 
 valleys, explains the variety of the inclinations, and suggests 
 that the lesson often taught to military engineers, the slope 
 on one side is nearly the slope of the other side, is a rule 
 open to exception to begin with, and still more liable to be 
 inaccurate when the action of ice in one, but not in the 
 adjacent valleys, is remembered as a not unfrequent event. 
 
 48. Mountains and Hills. — The low undulations of a 
 country which has been abandoned by the sea, and on which 
 atmospheric agents have begun to work, stand to the grand 
 features of such mountain masses as the Himalayas in the 
 same relation as the islet to the continent. The height of 
 the undulating surface above the sea is in exact proj^ortion 
 to the depth of the furrows, so that, in general terms, eleva- 
 tion and denudation go together in assisting to determine the 
 relative age of the high grounds. The advocates of atmo- 
 S])heric waste as the agent to which we chiefly owe the features 
 of the dry land, do not exclude the sea from all share in 
 determining the position of future vallej's. No surface, 
 probably, is ever left absolutely smooth; and the mud bank 
 of an estuary shows, as the tide ebbs, the kind of inequality 
 to be expected on a large scale in emerging continental land. 
 The slight hollow helps the atmosphere to commence, but 
 
SAND DUNES, ETC. 77 
 
 witliout such a lielp, the atmosplieric moisture would of 
 necessity excavate grooves for itself. 
 
 The outstanding features are hills, mountains, and table- 
 lands, plateaux or terraces, terms to which certain tolerably 
 precise meanings are attached popularly, but which it is very 
 difficult to define, so as to establish a real difference between 
 them as regards their origin. The cliffs, escarpments, and 
 slopes, by which these features merge into each other, are 
 more easily dealt with. 
 
 Between a hill and a mountain the only difference is a 
 convention of language, by which the latter term is vaguely 
 limited to heights of between 2000 and 3000 feet and 
 upwards; the emplopnent of one or other term does not 
 therefore, from a geological point of view, involve error. 
 
 Hills are detached or connected by their bases into groups 
 of various forms, the linear arrangement constituting a range; 
 the junction of several ranges forms a group with more or 
 less radial arrangement, while system refers to the ranges or 
 groups Avhich, in form and position, belong to the same 
 period, to the same region, or what comes to the same thing, 
 are due to the same processes of elevation or denudation. 
 
 Inequalities of the surface are caused by denudation, by 
 the outpouring of volcanic or other materials from the interior 
 of the earth, and by the shifting of loose materials on the 
 surface of the earth. 
 
 In Scotland, isolated cones of gravel are frequent towards 
 the coast lines; and though many of these are obviously 
 denuded fragments of sand and gravel plateaux, others have 
 certainly been left as w^now see them by the sea, which, at 
 the close of the glacial period, covered the low grounds. The 
 heaping together of the gravel by cross currents of water, 
 brings sul^marine and subaerial deposits very close together. 
 
 49. Sand Dunes, etc. — Sand dunes form ranges, the rela- 
 tions of which are often very interesting. The wind blowing 
 A'om the sea carries from the water edge a roll of loose sand, 
 and drives it onward till it is arrested by the slope of the 
 shore, or by some other obstacle; the piling thou goes on, 
 the particles travelling up one side and down the other, sq 
 tliat tlie masses travel forward by stages, the breadth of 
 which is half the breadth of the base of the sand ridge. 
 
78 PHYSICAL GEOGRAPHY. 
 
 These mounds can only be formed wliere a wind blows on 
 the shore steadily, or, Avhen it is not blowing, no other wind 
 passes over the surface with equal strength and frequency. 
 Hence the parallel ridges of the dunes inland, till cultiva- 
 tion, or a wood, or a river, arrests them. In this country 
 these ridges are trifling for the most part, the gi^eatest 
 development of them being on the north coast of Cornwall. 
 But in Africa, for example, the movement of the desert 
 sand over the Egyptian plains, where gaps occur in the high 
 grounds separating the two regions, is on a grand scale of 
 destructiveness, and in the Thurr or sand desert, which lies 
 north of the Kunn of Cutch and east of the Indus, the ridges 
 attain to 400 feet in height, their summits being thus 500 
 feet above sea level. It is right to mention, however, that 
 Sir H. Bartle Frere doubts this explanation of the origin of 
 these ridges, and suggests a series of parallel fractures and 
 subsidences as one of the many volcanic phenomena of the 
 district. 
 
 The mounds of glacier detritus, the terminal moraines of 
 extinct as well as existing glaciers, form piles of no incon- 
 siderable importance in some valleys, those of the Yal d'Aosta 
 attaining a height of 1600 feet above the plain. They ai-e 
 mentioned here simply to enumerate in conjunction all the 
 surface features. 
 
 60. Bearings of Ranges. — The directions of the linear 
 ranges and the grouped systems of hills varies in different 
 regions, as the following table, an abstract of that given by 
 Jukes, ^'' shows. The orientations are reduced to Bingerloch, 
 and they have not been corrected to any British locality, as 
 it is intended to show the relations of the systems to each 
 other, not to any particular point : — 
 
 N. 1° ir W. Corsica, Sardmia. Red Sea, Hungary, Syria. 
 Upper part of Loire and Allier, Khone. 
 
 L. Miocene, 
 
 N. 2° 30' E. Malvern Hills. I. of Gothland. N. of Russia. 
 
 Permian. 
 
 N. 15° 4G' W. Greece. Italy. Sicily. 
 
 N. 21° 4' E. Vosgcs. Ireland. Scotland. Scandinavia. 
 
 N. 31° 15' E. Longmynd. Saxony. Sweden, Finland, 
 
 * Manual of Geolofiy, Second Edition,' 
 
MOUNTAIN SYSTEMS. 70 
 
 Cambro-Silukian'. ■ 
 
 E. 4''32'N. IsleofWiglit Satra (Carpatliians). E.Alps. 
 Jura, 
 
 Eocene. 
 
 E, 2' O'N. From Elbe to St. Bride's Bay. Brittany. 
 
 S. Ireland. Thuringia. S.Russia. 
 E. 15' 6' N. S.E. England. ^^. France. Spain. N. Africa. 
 
 Atlas. Caucasus. 
 E. 37° 55' N. Oolitic Escarpment, England. Saxony. ISIonte 
 
 Pilato. 
 W. 23° 3' N. Pyrenees. Italy. Sicily. Greece. Carpathians. 
 
 Weald of Kent.' 
 
 An attentive comparison of tliis table witb tlie map will 
 show, more especially if the student refers to detailed maps 
 of each locality, that it is easy to construct parallels if we 
 disrecrard the dates and amounts of denudation in favour of 
 assumed elevating movements. 
 
 61. Mountain Systems. — The Scandinavian peninsula 
 is traversed by a ridge of continuous heights, of which 
 Sulitelma is 5956 feet in height, Skagesloestinden is above 
 8000 feet in height. 
 
 The axis of this chain is nortli-east and south-west. The 
 northern portion of the chain is higher than the southern, 
 and the snow line at the seaward side descends to 3200 feet 
 above the sea level. The axis w^ould, if prolonged, intersect 
 the protracted line of the Oural mountains, while, south- 
 wards, the British Islands may be regarded as a member of 
 the chain which, bordered the European continent to the 
 nortli-west and west, as the geological structure in both 
 countries is the same. 
 
 The Ourals extend over more than twenty degrees of 
 latitude, dying out southwards in the plateau of Sakmara, 
 ' wdiile the higher peaks are situated' in the northern portion, 
 and the average height is about 2500 feet. This, the eastern 
 limit of the European area, has its longer slope to the west- 
 ward, the steeper declivity fronting the lower plains of Asia. 
 From the sea at Cape Finisterre, to the shores of the Black 
 Sea,' a nearly continuous line of mountains traverses Europe, 
 I and, through the Caucasus, the line is prolonged through 
 t Asia, running out in Burmah. The Pyreneean section of this 
 long line has been elevated since the chalk period, and now 
 
80 PHYSICAL GEOGRAPHY. 
 
 constitutes a natural barrier between France and Spain, 
 having an average heiglit of 7000 feet, without any low 
 passes intersecting it. The highest peaks are Maladetta, 
 11,168 feet, Penaranda, 10,663 feet. Pic de Nethou, 11,426 
 feet, Mont Perdu, 11,994 feet. The slopes are the reverse 
 of what is usual along the great east and west chain, for the 
 declivity into Spain is shorter and more gradual than that 
 into France. In fact, the Pyrenees are the northern rampart 
 of an elevated plateau from which several ridges approxi- 
 mately, east and west, project; of these, the most northerly 
 are the Cantabrian Mountains, which are almost a continua- 
 tion of the Pyrenees, and the most southerly, the Sierra 
 Nevada, parts of which approach 12,000 feet. The snow line 
 of the Pyrenees averages 8000 feet; in the Sierra Nevada 
 it is at 11,000 feet. 
 
 The nucleus of the Alpine system lies between the sources 
 of the Pliine, Phone, Ticino, Inn, and Adda; it gives off to 
 the S.W. the Italian Alps, of which Corsica and Sardinia 
 are appendages; to the N.W., the Gallic Alps; to the S.E. 
 the Apennines, which are separated from the central mass 
 by the valley of the Po, almost as completely as South 
 India from the Himalayas by the Ganges; on the N.E., 
 the German Alps pass towards the Sarmatian plain, of which 
 the southern boundary is formed by the Carpathians; while 
 the Hellenic Alps are the S.E. prolongations, reaching through 
 the Balkan to the Black Sea, and forming, by its southern 
 spurs, the deeply intersected south-eastern corner of Europe. 
 The heights in this extensive system are Etna, the most 
 southerly point 10,874; Mount Athros, 9628; the highest 
 of the north-easterly branch does not exceed 9000 feet; 
 the Glockner is 12,956 feet; Mount Blanc, 15,784 feet; 
 Mount Pelvoux, 13,468 feet, and the Gallic portion has an 
 average of 5000 feet. A large part of this group is the seat 
 of perpetual snow, and of large glaciers, the snow line aver- 
 aging 8500 feet, being lower on the south than on tlie north 
 slo])cs. 
 
 Tlie Caucasus, the Taurus, and Antitaurus, form the con- 
 necting liuk between the Alpine and the Himalayan nuclei. 
 The Caucasus rises above the snow line at many points, its 
 paximum, Elburz, being 17,112 feet. The Armenian or 
 
MOUNTAIN SYSTEMS. 
 
 81 
 
 eastern portion of tlie Taurus and Antitaurus culminates in 
 Mount Ararat, 17,200 feet, and Demavend, 21,000 feet, 
 is a peak of the range which merges in the Hindu Kush, 
 
 £-2 ©a 
 
 E.g^o a. 
 
 tr '-. o "o 
 
 a cc'K 5 
 
 fi" 2 n £ ''^ 
 
 E o = 05 S 
 
 2 -■ 
 
 05 N 
 
 e^ 
 
 la 
 
 »C <t = ;: 3 
 £ "' S' 11, ™ 
 ^r /- p cr CD 
 
 ST £-M 
 
 CD f^ ■ 
 
 t= - s c 5 
 
 c c:?t'^ H 
 
 C •^- — ^- ts 
 & y S .^ c^ 
 
 while the Tamils is prolonged eastwards in a sinuous, but 
 not very high range to terminate at the same point. From 
 this nucleus, the Kuen Lun range passes eastwards as tlio 
 proper continuation of tlie Europeo-Asiatic line, v/liile tlio 
 Himalayas bend to the S.E., and maintain a uniform aspect 
 23 F 
 
82 PHYSICAL GEOGRAPHY. 
 
 facing tlie Indian Ocean. The Indus and Brahmapootra 
 rise on the north side of the range, and, flowing in oj^posite 
 directions, breach the rampart at its western and eastern 
 ends, so as, with the Indo-Gangetic valley, to insulate the 
 principal part of the range. The highest peak, Everest, is 
 29,002 feet, and the mean elevation is between 15,000 and 
 16,000 feet. To the north-east, the Himalayan line is con- 
 tinued through a region the details of which are unknown, 
 save that it consists of complicated mountain chains, by 
 which the transverse system of Central Asia is connectec 
 wiih. the great Altai system, which runs from the source o 
 the Irtysh to Behring's Straits; parallel ranges lie to th< 
 north and south of tliis main line, and separate differem 
 plateaux from each other, v/hile the circle of mountains just 
 outlined encloses the great table-land of Asia, the drainage 
 of which is, to a large extent, inwards. The ridges north of 
 the Himalayas rise to a great height above the sea level, 
 the central portion attaining to 10,000 feet, while either 
 extremity of the range sinks to an average of 5000 feet, but 
 these chains are all portions of the original plateau of Asia; 
 they are fragments of plains left outstanding by denudation. 
 Modern as is, geologically speaking, the elevation of the 
 Himalayas, they present a grand illustration of atmospheric 
 waste. Bising from the plains of India by successiA^e ridges, 
 the troughs separating these parallel ramparts are longitudinal 
 valleys, while transA^erse valleys intersect the slopes with 
 remarkable regularity, and the southerly spurs which run 
 towards Cambodia and the Malayan Beninsula are channelled 
 out by the streams which flow between them. The snow 
 line of the Himalayas is about 15,500 feet on the south, 
 19,600 feet on the north slope; glacier action, therefore, is 
 intense in a region large part of which is a constant snoAv 
 field, whose valleys, beyond- the limit of the glaciers, are 
 occupied by perennial streams, and are flooded at the same 
 time that the monsoon brings rain over the plains of Bengal. 
 When to this is added the fact that the westerly or south- 
 v/esterly upper current of the atmosphere is, so to speak, 
 tapped by the mountain range which forms a natural boundary 
 of the basin of the Indian Ocean, it is apparent that on 
 this range are concentrated all the agents of atmospheric 
 
MOUNTAIN SYSTEMS, 83 
 
 T\''a3te, and tliat thus, in a comparatively short time, and 
 among rocks of considerable relative hardness, denudation 
 has produced results comparable with those which, in 
 more northern latitudes, requii-ed long ages for their elabora- 
 tion. 
 
 Our knowledfi^e of the interior of Africa is not such as to 
 enable us to speak with certainty of its mountain ranges. 
 The grounds on which an east and west chain is believed to 
 traverse Central Africa are, as Laughton sums them up, 
 (1), the known high lands of Abyssinia; (2), the equatorial 
 mountains on the west coast, the Cameroons, and others seen 
 stretching, peak after peak, into the interior; (3), the lofty 
 mountains seen by Denham to the south of Bornu; (4), the 
 intimate and singular relationship between the mountain 
 flora of Abyssinia, of the Cameroons, and of Fernando Po. 
 The Atlas, to the north of the Sahara, is parallel with the 
 S.E, chains of Europe, and its parallel ranges gradually 
 ascend from Tripoli westwards, from 2000 to 13,000 feet, the 
 summits beinu' thus within the snow line. The Straits of 
 Gibraltar are excavated through a northern spur which 
 formerly closed the Mediterranean. The Abyssinian 
 mountains are, in reality, a table-land of an average level of 
 8000 feet, from which ridges project to more than 15,000 
 feet above the sea, while the intervening valleys are of 
 great depth, the gorge of Jitta being 3500 feet deej). The 
 depressed plateau of S. Africa rises southwards, and, in 
 the Cape district, the plateau is worn into mountain ranges 
 which are parallel to each other, and which present grada- 
 tions of height up to 6000 feet. The kloofs or ravines 
 which cut into the successive tables, or karroos, are results 
 of denudation of a kind curiously characteristic of all flat 
 elevated lands. It is wi'ong to speak of the karroos as 
 terraces; their surface is interrupted by outstanding low 
 heights, denuded hillocks, which mark the rate of denudation. 
 The kloofs are transverse water channels, which will gradually 
 encroach on the karroo ground till only sharp ridges separate 
 the ravines. A glance at the map of Livingstone's explora- 
 tions shows that this kind of denudation extends into the 
 lake repion. 
 
 o 
 
 Tlio two halves of the New "World agree in having, in 
 
84 PHYSICAL GEOGRAPHY. 
 
 common, a moiuitain range tlie features of which are gene- 
 rally similar. Tlie Rocky Mountains in N. America, tlie 
 Andes in S. America, are connected across the central 
 district by an essentially volcanic region, bordering the 
 Mexican Gulf. The heights of the principal peaks are 
 variously stated, but the following are probably correct 
 enough to give the general relations of the summits. The 
 Patagonian section is of moderate height, not exceeding 
 8000 feet, the height of the volcano Minchinmadiva. The 
 Chilian division forms a single range as far north as 30° 
 lat., rising, at 41° lat., to 16,000 feet in the volcanic Villa 
 ivica; Maipo is 17,660 feet; San Jose, 18,150; Tupungatu is 
 variously estimated at 22,000 and 15,000, probably in con- 
 sequence of confusion of names between it and Aconcagua, 
 which is 22,296 or 23,910 feet. The Bolivian section 
 includes the double chain of the A-ndes, and the commence- 
 ment of the spurs which once connected the main chain with 
 the lower parallel ridge of the Cordillera Geral, but are now 
 intersected by the tributaries of the Madeira river. In this 
 region the salinas, or salt deposits of the table-lands, are 
 found. The sinuous ridge which passes eastwards in 25° 
 lat. S. is the commencement of the eastern great chain; the 
 peaks are somewhat irregularly placed, being volcanic to a 
 large extent; among them are Colorados, 12,406 feet; Cerro 
 di Potosi, 16,000; Chorolque, 17,000; Cochabamba, 17,000. 
 The range which forms the proper eastern wall of the 
 Titicaca valley commences southwards with an average of 
 15,000 feet, and rises to Illimani, 21,252 (24,200 of some 
 writers), Sorata, 21,286, and sinks north-westwards to Yil- 
 causta, 17,500 feet, whence the eastern range gradually j 
 disappears, being dismembered by the streams which form 
 the head waters of the Amazon. The western ridge may be 
 regarded as itself a double mass, of which the highest peaks 
 are Gualateiri, 22,000; Sahami, 19,450; Parinacota, 23,000; 
 Arequipa, 20,300; Chuquibamba, 22,000. From the lati- 
 tude of Cuzco northwards, the chain continues of an average 
 height, no greatly outstanding peaks occurring in this por- 
 tion, which corresponds to the area of meeting of the cold 
 southern and warm northern oceanic currents, heavy fogs 
 being the obvious sign of their mixture. Northwards rise 
 
AXES OF ELEVATION". 85 
 
 Cliimbomzo, 21,450 feet; Cotopaxi, 18,880; Antisana, 
 19,150; Cayambe, 20,140; Tolmia, 18,120.''^ 
 
 62. Axes of Elevation. — The details of the areas, of 
 •which the principal features have been summarily indicated, 
 may be learned from special treatises. But from the sketch 
 that has been given it will be apparent that, to a very lai-ge 
 extent at least, atmospheric denudation has fashioned the 
 surface of all countries. The somewhat cumbrous generalisa- 
 tion has been put forward that the continents are older than 
 the mountains; but this is only true for the mountains which 
 have been raised into prominence by subterranean move- 
 ments, since — if the theory which has been developed holds 
 true — the mountains are for the most part fragments of older 
 table-lands. The elevation of any region to such an extent 
 as to determine the direction of its waters, may take place 
 at any period subsequent to the appearance of the region as 
 dry land; that elevation which gives to one portion the 
 relative height held by convention to constitute a mountain, 
 may either be continuous or effected by many successive 
 movements, at longer or shorter intervals, and of variable 
 intensity. 
 
 The axis of a mountain chain which may have been formed 
 by elevation corresponds theoretically, as it certainly does 
 primarily, to the summit of the anticlinal ridge into which 
 
 B 
 
 its strata are bent. A represents two anticlines with inter- 
 vening syncline; but, after long ages of denudation, the 
 relative positions of these two may be reversed, and the 
 bottom of the synclinal curve may come to form the top of 
 the hill, the greater part of the anticlines on either side 
 being worn away as in B. Ben Lawers represents such a 
 condition, the strata dipping on each side towards the centre 
 of the mountain, the top of the hill being the bottom of the 
 geological trough. But a more striking illustration is oftcred 
 by the Scuir of Eigg, as described and figured by A. Geikie. 
 
 * Very few of the heights here given are accepted by all writers, 
 the differejices being due partly to defective knowledge, partly to 
 inaccurate repetitions. 
 
S6 PHYSICAL GEOGRAPHY. 
 
 The Scnir is a mass of volcanic rock with vertical sides, 
 which stands on the summit of a hill; but it is in reality a 
 lava flow which occupied the bed of a stream, whose gravel 
 is still found in place below the lava. The banks of the 
 stream have been worn away; the bed of the stream now 
 terminates a conical hill, and the lava which once filled up 
 the bottom of the valley now forms the apex of the hill. 
 
 M. Eli^ de Beaumont believed that by comparing the 
 directions of mountain axes, and determining their parallel- 
 ism, the dates of their elevation might be approximately 
 determined. But subsequent fuller information led him, in 
 1853, to modify his views by the admission of several dates 
 of elevation in the Pyrenees, and this admission for a single 
 chain is practically the adoption of the hypothesis of slow 
 and successive movements. Sir Charles Lyell urges, in 
 favour of the slow;ness, that even in such a case as the 
 Pyrenees, where one of the movements took place between 
 the cretaceous and tertiary deposits of the region, the cre- 
 taceous beds were not necessarily the latest, nor the tertiary 
 the earliest of their respective epochs, and that therefore a 
 very long interval may have given time enough for a con- 
 siderable amount of vertical movement. 
 
 It is further to be remembered, that satisfactory approxi-' 
 mations to the date of movements depend upon accurate 
 identification of the deposits, and as these are becoming 
 gradually more precise, our chronological arrangements are 
 liable to important modification. The re-examination of the 
 Longmynd gave proof that the strata of which it is com- 
 posed are silurian, and that therefore the elevation was 
 post-silurian, and the correction of one opinion founded on 
 necessarily imperfect knowledge suggests caution in the 
 adoption of a view wliich implies that there is, in the 
 chemical phenomena underground, a regularity and period- 
 icity which cannot be affirmed for the chemical or mechanical 
 phenomena by which the surface is affected. Thus, to take 
 the succession of volcanic outbursts in Britain : the silurian 
 strata are traversed by dykes, or injected masses of later 
 date than the beds which they break up; but the contem- 
 poraneous lava sheets, evidences of volcanic activity at the 
 surface of the earth dm-ing their deposit, are chiefly found in 
 
AXES OP ELEVATIOJJ. 8t 
 
 Wales, traps and asli entering largely into tlie structure of 
 Snowdon. Tlie old red sandstone period witnessed many 
 active volcanoes north of the Tweed; in the carboniferous 
 period, volcanoes emitted lava and ashes at almost every 
 stage, but the locality was not always the same, and Prof. 
 A. Geikie has very clearly set forth the alternate quiescence 
 and activity of the various districts. During permian times 
 the outflows are abundant, and volcanic vents referred to 
 this period are found at some distance even from the nearest 
 permian strata. The mesozoic strata were deposited in 
 Britain during a period of quiescence, and the tertiaries of 
 England are undisturbed. In the middle of Scotland, which 
 had been the chief seat of carboniferous volcanoes, there is 
 no later development on any considerable scale; but in the 
 west, from Antrim to the Faroe Islands, the miocene lavas 
 v.'ere not merely poured out in enormous quantity, but have 
 since undergone an astonishing amount of denudation — Mull 
 and Eigg, for example, being isolated fragments of a trappean 
 plateau, once probably well nigh continuous. To this period 
 belong the latest volcanoes of Auvergne. In the succeeding 
 times, the site of volcanic activity has been transferred to the 
 Mediterranean area, and to a zone which includes Hecla and 
 Jan Meyen. So far, then, as it is possible to estimate with 
 safety the relative length of different geological epochs, the 
 intervals between these various developments of volcanic 
 activity are unequal, and the phenomena are, for each 
 period, limited i]i the area they occuj)y. If, therefore, there 
 is reason to believe that elevatory movements and volcanic 
 outbursts have, if not identical, at least closely similar 
 causes; and if, as also seems probable, the subterranean cavi- 
 ties in which volcanic products are generated are connected 
 with each other, the irregular recurrence of the phenomena 
 is not such as to justify any general statement regarding 
 tlieir results. Bearing in mind that this forms a difficulty in 
 the way of explaining the irregular occurrence of eruptions, as 
 much as in the way of determining the periods of subterranean 
 movements, the belief in the separation from each other of 
 these subterranean reservoirs would still further complicate 
 the question. For on that hypothesis we should be com- 
 pelled to seek on the surface the cause of this development 
 
88 PHYSICAL GEOGRAPHY. 
 
 of elevatoiy movements in parallel directions at distant 
 localities. And reference to the map of tlie world given by 
 Darwin (Coral Reefs) shows that at this present time move- 
 ments in opposite directions take place in adjacent districts 
 to the south of the equator, while in the north also antagon- 
 ism of the same kind occurs — Greenland descending while 
 Disko is being elevated : the north of Scandinavia is slowly 
 rising while the extreme south is as surely sinking. 
 
 63. River Valleys in England. — Professor Ramsay, in 
 his investigation of the history of rivers, has given good 
 reason in support of the following as the course of events in 
 S. England. The mesozoic period was brought to a close 
 by a general depression, during which the cretaceous rocks 
 covered over unconformably the oolitic and all older strata, 
 abutting against the Welsh silurians. (The three following 
 sections are supposed to be seen by a spectator looking south). 
 
 £ J Cretaceous.,. V. ■yv'.- 
 
 The movements which we know to nave taken place during 
 the early tertiary times in Central Europe, were associated 
 with a tilting of the French and English areas, so that the 
 cretaceous beds dij)ped to the N.W., and thus a groove 
 was formed, at the bottom of which a stream ran, with 
 the Silurians for its westerly bank. The Severn, thus com- 
 menced, continued thereafter to flow in the same direction, 
 but the atmosphere cut back the chalk in the escarpment 
 form now seen, and afterwards the east of England was again 
 tilted, so that its incline sloped in the opi^osite direction. 
 
 ^- i — Ci-etaccou,?. 
 
 \\.v 
 
 Old Red Sandstone. 
 
 x*^* 
 
 EastNvard vivers. g 
 
 Old Red Sandstone. 
 
 The streams now flowing east on the surface of the chalk 
 
FORMS Of MOUNTAINS. 8() 
 
 maintained their direction wliile tlie clialk escarpment was 
 gradually eaten further back eastward, and thus they seem 
 to cut through a high ground, whereas in reality the escarj^- 
 ment has receded past them. 
 
 54. Forms of Mountains. — The forms of mountains, their 
 aspects, varied as they may appear, have very close relations 
 to their structure and the influences to which they have been 
 subjected. The flat summits of the fragments of table-lands, 
 as Table Mountain, and the hills of Saxon Switzerland, and 
 of trappean plateaux, as in Auvergne, are similar, because 
 both consist of rocks laid doYs'n in horizontal strata which 
 retain their relative joosition. The smoothly rounded sum- 
 mits of the hills in S. Scotland, and the loAver heights in 
 N. Scotland, Wales, Ireland, some parts of Central Europe, 
 and America, are due to the action of ice passing over ground 
 which once formed a plane of marine denudation, and which 
 still retains the primitive relation of its parts, notwithstand- 
 ing the denudation which its surface has undergone. The 
 wide, even curves of the gaps which intervene between the 
 heights, tell, in Brazil as in Europe, of the passage of 
 ice through the strait, leaving a channel of symmetrical 
 form. 
 
 The serrated ridges which give to many mountain masses 
 their peculiar character, for which restlessness is the best ex- 
 pression, are due to the similarity of texture of the ranges, as 
 much as to the length of time during which they have been 
 exposed to denudation. The highly altered slates and schists 
 of metamorphic districts weather unequally along the strike 
 of the strata, and thus give an iiTegularly rugged outline, 
 which, when large* masses of unequal hardness alternate, 
 becomes exaggerated into the pointed peaks constituting the 
 aiguilles of Alpine scenery. In a country the hills of which 
 have in general the rounded smooth outlines of ice-worn 
 summits, many of the lower heights may come to present, as 
 in the Scottish Highlands, serrated summits; while the hills 
 of S. Scotland, though exposed to atmospheric waste for an 
 equal length of time, since the glacial epoch, very rarely lose 
 their smoothness, their softer consistence permitting them to 
 weather more evenly. The outlines of trappean plateaux 
 in the last stage of disintegration are similarly ragged, but 
 
90 fSYSICAL GEOGRAPHY. 
 
 tlie peaks are more symmetrical pyramids, their texture loeing 
 uniform on all sides. 
 
 65. Cliffs and Escarpments. — Some hills, or ranges of 
 higher ground, terminate not by slopes, but by abrupt pre- 
 cipitous faces. These are sjioken of commonly as cliffs, 
 the same term being applied to them as to the similar 
 faces of rock which border a coast line. It is, however, 
 desirable to employ a different term for many of these 
 inland features which are not due to marine denudation. 
 Cliff and escarpment are the two terms which field geo- 
 logists have adopted, the cliff being the result of marine 
 action, the escarpment of atmospheric waste. The base 
 line of the sea-worn cliff is always horizontal, that of the 
 escarpment may or may not be so. If both are composed of 
 strata which undulate, the sea which can work only along 
 a horizontal plane, disregards the undulations; the atmo- 
 sphere, which Avorks most rapidly along the lines of least 
 resistance, makes a hard stratum the bass of the steep face, 
 and if the upper surface of that hard stratum undulates, the 
 base line of the steep face undulates also. The escarpments 
 of the chalk and oolites in England illustrate the departure 
 from the horizontal of the base line, and the undulations 
 determined by a harder bed; sections, moreover, drawn in 
 the chalk district by help of the Ordinance Survey maps, 
 prove the base of an escarj)ment to be higher at one than its 
 summit at another point, an impossible relation in a sea-worn 
 cliff, unless, subsequent to its formation, violent disturbances 
 can be demonstrated. 
 
 Without denying the possibility of vertical precipices of 
 great height occurring, their number must be very small. 
 A river or the sea undercuts its shores, and thus would 
 maintain, if no other action took place, constantly vertical 
 faces; but the sides of a valley, and the cliffs along the coast, 
 all slope either evenly, or by ledges which are determined 
 by the relative hardness of the strata. Perpendicular faces 
 very rarely have a greater height than the thickness of the 
 stratum in which they occur, and this form is very persis- 
 tent; it is due to the jointing of the rock, or that ( Art. 26) 
 which leaves a vertical face after each successive landslip. 
 
 Yery few cliffs or escarpments are due to faults. The 
 
J>LATEAUS. 91 
 
 steep face only rarely represents tlie mass of strata along 
 the line of fracture, towering above the portion which has 
 sunk down to a lower level. The fissure or line of parting 
 between the two masses of rock seldom if ever gapes at the 
 surface, and if the movement of the two masses past each 
 other left at the surface an inequality, atmospheric and 
 marine denudation would in their turn strive to smooth it 
 down, so that unless the feature were of recent develop- 
 ment, it would not be recognisable. In the few cases in 
 which the escarpment owes its origin to a fault, its face is 
 parallel in direction to that of the fault, but it has receded 
 to a distance proportioned to the time during which atmo- 
 spheric denudation has been at work. Thus — 
 
 Volcanic Plateau, 
 
 Volcanic Plateau. 
 
 In this diagram, where ''volcanic plateau" on the left side 
 represents the general level of the country formed by aii 
 approximately horizontal layer of trap rocks, which have 
 sunk down from the higher level of "volcanic plateau" on 
 the right, the face of this escarpment has receded from the 
 line of the fault in the same way that the escarpment of the 
 chalk has receded from the margin of the Severn valley to 
 its present position. - 
 
 56. Plateaux. — Bearing in mind that the plane of marine 
 denudation is the starting point for all modifications of 
 the surface of dry land, and further, that the extent to 
 which the interior of continents is raised above the level 
 of the sea varies in difierent regions, it is a necessary con- 
 sequence that indications of the primitive i:)lain should be 
 met with at very different heights. Geographers have en- 
 deavoured to establish a distinction between mountains 
 and hills, and have desired to modify the meanings of the 
 popular terms, so that conventionally a mountain shall 
 mean a more connected series of higher ground, such higher 
 
95 PHYSICiAL GEOGkAPHy. 
 
 ground having the form of cones or ridges, with more or less 
 rapid slopes, while hills shall indicate more isolated groups 
 of smaller relative height, and mostly consisting of detached 
 conical forms. Obviously this distinction, though convenient, 
 is unscientific, since there is not, in the case of the features 
 of the dry land, the same natural separation which furnishes 
 the distinction between continents and islands. The writers 
 referred to lose sight of the gradual transition effected by 
 denudation processes from the primitive plateau, through 
 hills and valleys, to the plain to which denudation tends 
 to reduce all prominent features. As might be expected, 
 parallel vagueness attaches to the classification of plateaux 
 Avhich are grouped according to their heights; commencing 
 with those of the first class, which are more than 4000 feet 
 above the level of the sea, we have in the profile section of 
 Asia {Student's Physical Atlas, Plate YIII., fig. 2), the 
 Thibetan Plateau of about 15,000 feet; further to the north 
 the Pamir Steppe of about 14,000 feet, and from these great 
 elevations a series of low terraces leads down to the Ai'ctic 
 Ocean, the slope of the Siberian surface being gradual. To 
 the south of the Himalayan Mountains the descent is more 
 abrupt. The plains of the Ganges have an average elevation 
 of 250 feet, forming low grounds which circle round the 
 northern extremity of the Deccan. The Mountains of Asia 
 Minor, again, project from a plateau of about 5000 feet in 
 elevation. The great sandy desert of Gobi is somewhat 
 lower, while the Aralo-Caspian depression is in the midst of a 
 plateau of about 2000 feet. In Europe (fig. 1) the Caucasian 
 Isthmus, the Steppes which border the Black Sea, and the 
 plains of the Danube form a great tract of low ground, which 
 is continuous with the Aralo-Caspian tract. But Europe 
 presents less extensive continuous plains than we find in 
 Asia, the hill surfaces being more extensive relatively to 
 the horizontal area. Minor plateaux are recognizable in 
 Hungary, in North Germany, in France, in Spain, and in 
 the British Islands. Ireland perhaps presents the largest 
 continuous expanse of low ground. Of Africa the Sahara 
 is the best known plateau; to the south the central area is 
 believed to be a table-land, inclining on all sides towards the 
 middle, constituting thus a shallow trough. In the extreme 
 
PLAINS OF DEPOSIT. 93 
 
 south, Table Mountain furnishes an illustration of the con- 
 struction of a plateau which may be afterwards compared to 
 the great desert of Western North America. The table- 
 lands which occur between the eastern and western peaks 
 of the Andes and the E-ocky Mountains, the prairies in the 
 midst of which the Mississippi flows, and the great similar 
 tracts bordering the Amazon and the River Plate repeat 
 again the general features of other regions. But to make 
 this enumeration of plateaux complete, it is necessary to 
 mention the submarine plains, of which one of the best 
 known is that siUTOunding the British Islands, and indicat- 
 ing the western extent of Europe at no very distant date. 
 Other plains or terraces occur at various points of the ocean 
 ■floor, as determined by soundings; and if our information 
 regarding them were complete, we should be enabled to 
 tabulate a graduated series of different levels from the lowest 
 point of the ocean to the summit of the Himalayas. Many 
 of the erroneous ideas which prevail regarding these plateaux 
 arise from the fact that sections are not commonly drawn on 
 a " true scale," that is, one on which the horizontal and the 
 vertical measurements are on the same scale. If such a 
 section were drawn from the Himalayas across the Pacific 
 to the Bocky Mountains, it would appear that the slope 
 from the highest to the lowest points is a very gTadual one, 
 and that our division of the slope into plateaux is somewhat 
 arbitrary. 
 
 67. Plains. — Plains fall into three natural groups, plains 
 of deposit, plains of denudation, and plains of volcanic 
 origin. 
 
 68. Plains of Deposit are illustrated by the deltas of 
 great rivers, and the alluvial flats which form the coast line 
 in many places. The deltas of great rivers are the deposits 
 which have gradually filled up estuaries running into the 
 
 I mainland; their presence, therefore, is proof of a very con- 
 siderable antiquity of the river valley. The mode of forma- 
 tion of such plains is very simple. A river carrying down 
 from the higher grounds loose materials, and spreading these 
 over the bottom of its channel, gradually elevates its channel 
 so that the banks at last overtop the level of the surrounding 
 country, as in the case of the Po and other streams. Every 
 
9^ PHYSICAL GEOGRAPHY. 
 
 overflov/ leaves upon the adjacent coimtiy a tliin layer of 
 silt, and thus gradually a plain is built up, the st?;atification 
 of which is typical in its regularity. Even where, as in the 
 rivers of our own country, the level of the stream is not 
 above that of the surrounding country, the alluvial plains 
 are still built up by a process which is essentially the same; 
 repeated floods, raising the stream surface to the level of or 
 above its banks, spread over the adjacent ground, and leave 
 behind a deposit of water-worn detritus. One of the most 
 ancient of such plains, which we may call typical or normal, 
 is Table Mountain, the strata of which are horizontal, the 
 mountain beings in fact the survivor of an enormous table- 
 land, the rest of which has been removed by denudation. 
 
 59. Planes of Denudation are the result of marine action, I ' 
 aided in many cases from time to time by other agents of 
 waste. The central plain of Ireland, the Sahara, the Cheshire 
 plains, probably large part of Ai'alo- Caspian plains, are 
 illustrations of this class. We have evidence that these 
 now level surfaces Avere formerly covered by thick masses of 
 strata, which have been entirely removed, and their apj)roxi- 
 mately level surface is to be regarded as a sign that, while 
 subjected to the denuding agent, they remained stationary 
 for considerable periods of time. When it is borne in mind 
 that whether a plain is due to deposit or to denudation, it 
 cannot be long exposed to the influence of the atmosphere 
 without suiFerino: denudation, it is evident that we have here 
 again a very ill-defined line of distinction between these and 
 the planes of marine denudation. If plains are elevated to 
 considerable heights above the sea level, their waste would 
 carve them out into deep valleys, separating hills of various 
 forms, and thus the difference of material becomes unimport- 
 ant in comparison with the identity of the processes to 
 which the materials are submitted. - 
 
 60. Plateaux of Volcanic Origin are small in. extent, 
 and few in number as compared with those already described. 
 Every volcanic region furnishes examples of sheets of vol- 
 canic material evenly spread out, and giving to the landscape 
 a characteristic aspect. In Central Prance, in Catalonia, in 
 North America, we have volcanic plateaux which still retain 
 upon their surfaces the irregularities of the scoriaceous lava. 
 
FORMATION OF DELTAS. 95 
 
 [ii . Britain we find, in many districts, fragments of such 
 ancient plateaux; but they are either still covered by sedi- 
 mentary strata accumulated upon them since their formation, 
 or Avhere these latter deposits have been removed, they may 
 perhaps be more accurately regarded as plateaux of denudation. 
 
 The importance of these plains from an economic point of 
 view, or because of their- influence on civilization, is very 
 great. Sandy deserts are even more serious barriers to pro- 
 gress than the intervention of deep and broad seas; and, 
 even within the very limited area of Belgium, we find in 
 the well-known colony of Gheel, a small population sur- 
 rounded by the Campine, shut off from the other parts of 
 the kingdom, and remaining in a state of j)rimitive simplicity 
 which it is difficult to parallel in Western Europe. The 
 Landes of Gascogne illustrate the influence of a plain, the 
 structure of whose soil renders it unfavourable to the health 
 of the district; the sand of the Landes rests upon a hard 
 cake formed of organic debris which has matted together 
 the particles of the soil, and i-enders the surface liable to be 
 converted by floods into a stagnant marsh. The removal of 
 endemic disease from the district has been largely a result 
 of the draining of this area, whereby a great addition has 
 been made to the agricultural surface of France. The slow 
 formation of a plain of deposit in the valley of the Ganges, 
 and, indeed, along various points of the Indian shore, which at 
 different localities supported a thriving population, has con- 
 verted regions once populous and wealthy into an unhealthy 
 desert. The delta of the Ganges, stretching • for 200 miles 
 through Bengal, and forming the rich plains of Bengal, is 
 made up of the anastomosing branches given off by the 
 Ganges and the Brahmapootra, and between the streams are 
 islands of alluvial mud of very various heights, and frequently 
 shifting their places as they are worn away by floods from 
 the land or the ocean side. On the islands which form the 
 lower delta, a rich vegetation gives shelter* to wild animals 
 of all kinds, and the tiger has banished the inhabitants from 
 some parts of a region where man had maintained a footing 
 against the ravages of the flood. 
 
 61. Formation of Deltas. — The formation of deltas 
 
96 PHYSICAL GEOGRAPHY. 
 
 requires some little consideration. The first requisite is 
 that the mud-bearing stream should enter the sea so slowly 
 that the detritus is not swept at once beyond the shore into 
 deeper water. A tidal river with a rapid flow like the Thames 
 is scoured out twice daily, and accumulations are not per- 
 mitted to rest, far less to consolidate. A tidal stream, again, 
 such as the Clyde, where the estuary is long and narrow, and 
 the slope of the bed not great, allows a deposit of sediment 
 which forms alluvial flats on either side, and thus ofiers the 
 transition from the ordinary mud banks of a river to the 
 delta properly so called. The commencement of a delta is 
 at that point where the stream divides under the influence 
 of slight obstacles, and its channels form a network embrac- 
 ing alluvial patches elevated above the surface of the waters.^ 
 These patches are not all at the same level ; for just as a ' 
 single stream like the Po may build up its banks above the 
 surrounding country, the channels thi'ough a delta may build 
 themselves up so as to form conspicuous projections, not, 
 however, permanent ; and Mr Ferguson records that, within 
 a very few years, the debris of a house which he had himself 
 built has been covered to a depth of 30 feet, while on the 
 new surface a village has arisen. By this alternate laying 
 down and sweeping away, the general surface is, on the whole, 
 steadily raised, and at the same time the seaward progress of 
 the triangle continues. The name delta is borrowed from 
 the Greek letter a, whose figure that of the river accumula- 
 tions repeats; but it must be noted that this form is not 
 peculiar to river mouth deposits, since, as the cone de dejec- 
 tion, it is the characteristic, fan-like pile of stones at the foot 
 of a dry gulley, and the more flattened mass which the ravine 
 lays down in the lake. The base of the triangle reaches the 
 coast line, and then one of three events takes place : it 
 pushes out seawards ; it becomes arrested, or the river carries 
 its mouths outwards without the rest of the delta pushing 
 equally far. The seaward extension, possible only when 
 coast currents are not strons^ enough to remove the sediment 
 as it is brought down, is especially noticeable in the Adriatic, 
 where the Po and Adige are steadily building their alluvia 
 seaward ; but the rate of advance has greatly increased 
 within the last ceixtury, being accelerated by the means 
 
STEPPES. 97 
 
 taken to protect the plains of Lombardy from being flooded. 
 The rivers are now confined within high artificial banks, and 
 the additional speed tlius conferred carries the sediments 
 farther outv/ards. The growth of the E-hone delta within 
 historic times is well known. Towns once on the coast are 
 now, after nine centuries, two leagues from the sea, and the 
 successive sand bars, thrown np parallel to the coast by the 
 south winds, aids a process the results of which are made 
 lasting by the infiltrated lime which gives solidity to the 
 Ager lapidosus, the stony delta of the Elione. It may give 
 an idea of the approximate horizontality of subaqueous 
 deposits to state, that off the mouth of the Rhone the bottom 
 sloj)es southwards at an angle of less than 1°, or 1 foot in QQ. 
 
 The arrest of the Nile delta at the present coast line is 
 effected partly by the current which sets eastward along the 
 African shore, and partly by the subsidence of the land 
 which is still going on, and which permits the settlement of 
 large parts of its sedimentary burden in the interior. The 
 rapid descent of the sea bottom from 12 to 380 fathoms con- 
 trasts with the slow slope of the Rhone silt, and shows that 
 the arrest of the delta is not of recent commencement. 
 
 The peculiar process by which the Mississippi terminates by 
 mouths at the extremity of a long tongue of land, which 
 spreads out like a bird's foot at the end, is the same as that 
 by Avhich it, like the Po and other rivers, builds up the levels 
 of its banks. 
 
 62. Steppes, etc. — The plains have received particular names, 
 or rather the native names are adopted into English on account 
 of the convenience of thus recalling their characteristic features. 
 The plains of Europe j)resent great diversity ; level or gently 
 undulating, they are either grassy meadows or forest lands ; 
 in winter they may be flooded, and in summer more or less 
 swampy. The Steppes of Eastern Europe and Asia, already 
 treeless in the time of Herodotus, as they now are, support 
 for a brief season a coarse but often tolerably abundant 
 vegetation; but summer and winter convert them alternately 
 into utter deserts and trackless snow plains, thus making ib 
 only too easy to understand how they have arrested civiliza- 
 tion over an area of more than a million of square miles. 
 The antiquity of these wastes is confirmed by Yon Bar's 
 23 a 
 
98 PHYSICAL GEOGRAPHY. 
 
 observation, tliat tlie squirrels whicli throng the woods to the 
 north of the Steppes in Kussia are not found in the Crimean 
 forests j the separation of the two woodlands must, therefore, 
 have been remote. The Prairies of North America, the 
 counterpart of the Steppes in the Old World, are great tracts 
 of rolling ground, mostly meadow, but, like the Steppes, 
 destitute of trees. Bounded westward by the Rocky Moun- 
 tains and the desert land at their base, the Prairies pass 
 eastwards into the forest lands of the Appalachians and the 
 Atlantic shores. NortliAvards towards the Arctic Ocean, and 
 southwards to the Mexican Gulf, the waste occupies an area 
 of 3,000,000 square miles, and in South America an area of 
 nearly the same extent is divided between the Llanos of 
 Venezuela, the Campos Geraes of Brazil, the Pampas, and, 
 farther south, the deserts of Patagonia ; v/hile, to complete 
 the series of similar tracts, the deserts of Atacama, and the 
 Salt Lake region of North America, may be mentioned. A 
 careful comparison of the conditions under which these 
 various regions are placed will shov«r that to the dryness of 
 the climate, to the long intervals between their rainfall, and 
 in the extreme cases of the Sahara, Gobi, Atacama, and the 
 like, to the total absence of rain, are their peculiar aspects 
 due. Everj^where do we find forests intermingled with the 
 steppe lands, but the trees follow the coast lines, and pass 
 into the interior only along the river courses. The full 
 meaning of this generalisation will only be intelligible after 
 a consideration of the Atmospheric Currents in a subsequent 
 chapter. But the sharp limitation of the forests to the 
 joerennially moist areas is nowhere better illustrated than in 
 America. Passing westwards from Arkansas, in 35° to 3G° 
 north latitude, the forests here and there enclose patches of 
 prairie land, which increase till the woods are only oases; 
 the vegetation becomes restricted to bufililo grass, but in 
 106° west lono'itude the forests ai>"ain commence. In South 
 America the selvas or forest lands of the Amazon are in the 
 line of the easterly winds, which blow from the Atlantic up 
 the valley, the natural moisture of which is thus enabled to 
 support a greater vegetation. The increased dryness of 
 regions in which, artificially or by accident, the timbe)' has 
 been destroyed, as in Madeira, since the beginning of the • 
 
STEPPES. 99 
 
 fifteeiitli centniy, wlieii the forests were burnt; tlie disap- 
 pearance of springs after the woods have been cut down; the 
 refilling of the lakes in Venezuela while the Creoles, during 
 their struggle for independence, neglected the cultivation of 
 the sugar cane; the formation of oases round the wells bored 
 in Algeria by the French, these and many other instances 
 might be cited in support of the view that the amount of 
 moisture is the condition on which the stepjoe or the forest 
 depends. Forests cannot be made to grow unless the climate 
 is favourable, still less can they be forced on the Kussian 
 Steppes, where for long ages no timber has grown, because 
 the climatal conditions are entirely changed. That region, 
 like the interior of Africa, Australia, Asia, has its winds 
 dried before they reach the interior, and the rainfall is at a 
 minimum. 
 
 The Tundras or peat mosses of Siberia, and the heathy 
 plains of Germany are usually grouped with the steppes ; but 
 the former, like the polders of Holland, the fens of England, 
 are the result of a supply of moisture in excess of evapora- 
 tion ; while the latter are separated from the steppes by the 
 forests, and require a greater supply of moisture, or, at least, 
 are less able to retain it than extensive woods. 
 
CHAPTER III. 
 WATER. 
 
 Forms of Water — Cycle of Water — Imperfect Analogy of Aqueous 
 and Atmospheric Envelopes of Earth — Effects of Cycle of Water. 
 
 63. Cycle of Water. — Water is tlie most important 
 geological agent of which we have knowledge. The forms in 
 which it effects change upon the surface of the earth are: 
 1°. Atmospheric moisture, whether insensible as vapour of 
 water, or condensed into dew and rain, or, lastly, the con- 
 densed moisture solidified into ice. 2°. E-ivers above and 
 below ground; and under this heading come the glaciers, 
 which, as moving rivers of ice, are denuding agents of great 
 importance in temperate regions. 3°. The ocean. 
 
 It is difficult to say with which of these forms we ought 
 to commence our investigation, for the the cycle is a con- 
 tinuous one, by which water is carried into the atmosphere 
 and returns a2:ain to the ocean through the intermediate 
 stages of rain and rivers. 
 
 64. Imperfect Analogy of Aqueous and Atmospheric 
 Envelopes of Earth. — The endless movement of water stands 
 in a somewhat peculiar relation to what, recalling M. Guyot's 
 figure, may be called the other functions of the globe, and 
 especially to that of the atmosphere. The currents by which 
 the waters of the ocean travel to and fro, maintaining the 
 balance of distribution, are determined by the- winds, their 
 direction and velocity being likewise affected by the earth's 
 rotation ; by evaporation, as a consequence of great local heat ; 
 by the tidal movement, and by the features of the coasts and 
 sea floor. The atmospheric currents again are variations of 
 the westerly winds, which form an oblate spheroidal shell 
 round the earth; blowing at the level of the se^v in high lati- 
 
EFFECTS OF CYCLE OP WATER. 101 
 
 tudes, at some distance above tlie earth at tlie equator, and 
 this permanent movement is dependent on the earth's rota- 
 tion. The parallelism usually taken for granted between 
 the two fluids, air and water, which surround the earth, is 
 therefore of the most imperfect kind, and could only be 
 perfect if both elements formed spheres around a smooth 
 globe which presented no projections, no features capable of 
 causing deflections ; and even then it would be further neces- 
 sary that both fluids should present the same physical 
 properties. 
 
 65. Efiects of Cycle of Water. — By the movement of the 
 ocean, heat and cold are distributed over extensive regions 
 with a regularity and moderation due to the slowness with 
 which its temperature is altered. Evaporation not merely 
 supplies the needful moisture for the support of animal and 
 vegetable life, it tempers the atmospheric currents — regulates 
 them, so to speak. Acting on the land on which it falls, 
 rain and rivers distribute chemical substances over larger 
 areas, and prevent them from being accumulated in one 
 locality. They carry off*, slowly but surely, the materials 
 which, spread over the ocean floor, are in their turn raised 
 into dry land, and thus the subterranean movements are 
 compensated. But these movements again owe their energy, 
 at least frequently, to the presence of water. The circulation 
 of water underground is a physical necessity of the broken 
 condition of the strata, a necessity which results in the for- 
 mation of springs, but for whose presence many regions 
 would be uninhabitable for plants and animals, while the 
 percolation of water charged with chemical solutions allows 
 the deposit, in fissures, of minerals of various kinds. But 
 part of this water, in place of returning to the surface, gets 
 access to the underground reservoirs, and either originates 
 or hastens the volcanic phenomena which we speak of as 
 violent, and whose results we term catastrophes, because we 
 cannot rightly estimate their place and proportion among 
 natural phenomena. 
 
10^ PHYSICAL GEOGRAniY. 
 
 SECTION I. 
 
 rroportion of Land to "Water Surface — General Relations of Oceans 
 — Deep Sea Soundings : How taken — Soundings in Atlantic ; 
 Mediterranean ; Indian Ocean ; Pacific — Form of the Ocean 
 Floor — Deposits on Ocean Floor — Specific Gravity and Contents 
 of Water — Pressure of Water — Temperature of Ocean — Colour 
 of the Water — Luminosity of the Sea. 
 
 66. Proportion of Land to Water Surface: General 
 Eelations of Oceans. — ^The ocean, as lias been stated, is fonr 
 times more extensive than the dry land, and' in the water 
 surface of the globe must be reckoned the area covered by 
 rivers and .lakes on the dry land. Paradoxical as it may 
 sound, the greatest water area in a continent is toward the 
 high grounds where the feeders of the rivers are most 
 numerous, uniting successively as they approach the lo^Y 
 grounds, so as ultimately to form streams, which occupy a 
 greater vertical and diminished horizontal area. 
 
 The oceans of the globe are the Pacific, the Atlantic, and 
 the Indian, and these three are connected by the southern 
 circumpolar ocean, their northern circumpolar connection 
 being very much smaller. These diflferent oceans are further 
 subdivided into regions which, in general terms, correspond 
 with the prominent features of the continents; thus the 
 Atlantic Ocean is obviously divided into two basins, marked 
 off by the constriction caused by the eastern projection of 
 South America, and the western projection of Africa. The 
 Pacific Ocean has its northern and southern basins roughly 
 marked oif by the Polynesian Islands, while the Indian 
 Ocean shows a less complete division into two by the 
 pyramidal mass of southern India. The minor areas which 
 are referred to as seas, gulfs, and bays, belong to two cate- 
 gories : in the one case they are denudation valleys, in the 
 other areas of subsidence. The German Ocean partakes of 
 both characters, but it is pre-eminently an aix)a of subsidence. 
 The Ked Sea is a denudation valley. The narrow channels 
 v/hich separate Madagascar from Africa, and divide Celebes 
 and Lombok on the one side from Borneo and Java, outliers 
 of Asia on the other, though doubtlesss to some extent valleys, 
 
DEEP SEA SOUNDINGS: HOW TAKEN. 103 
 
 are, by their antiquity and comparative depth, as well as by 
 tlie distinctness of tlie animals on either side, entitled to a 
 prominent place, if they cannot be ranked as co-ordinate 
 with the great oceanic areas. 
 
 The Pacific Ocean coA^ers an area of about 90 millions of 
 square miles, the Atlantic occupies less than a third of that 
 surface, the Indian Ocean covers a somewhat smaller area 
 than the Atlantic. The Antarctic Ocean, in which all these 
 great basins converge, is very imperfectly known, and the 
 smaller Arctic Ocean is also even yet the subject of discus- 
 sion and inquiry. 
 
 The Atlantic Ocean has in connection with it certain bays 
 or recesses, more or less shut off from the general ocean space, 
 and, in a few cases, entirely excluded from it. The North 
 Sea is simply a bay of the old European continent, the 
 Straits of Dover being a recently formed passage; the Baltic, 
 with its branches, the Gulf of Finland, and the Gulf of 
 Bothnia, is a valley v/hich opens by a narrow channel into 
 the North Sea; and from the Straits of Gibraltar eastwards a 
 series of basins is defined, the first of which is bounded by the 
 Italian Peninsula, Sicily, and Malta; the second, by the pro- 
 longation of the line of the Grecian Peninsula, to the east of 
 which is the third ; these three divisions of the Mediterranean 
 likewise receiving lateral branches, such as the Adriatic and 
 Ai'chipelago. This last named shallow basin is connected 
 with the sea of Marmora, and that again opens into the Black 
 Sea, from v/hich the Sea of Azof is shut off by a narrow 
 channel, while the Caspian Sea and the Sea of Aral are 
 obviously basins which have become closed off from the 
 Mediterranean. 
 
 67. Deep Sea Soundings : How taken. — ^The sotmdings 
 conducted in recent years have given important information 
 as to the shape of the ocean floor. The method of conduct- 
 ing these observations deserves some attention. The simplest 
 plan in still water is to let drop from a boat, w^hose position 
 can be fixed by the oars, and by reference to objects on the 
 land, a line with a weight at its extremity, the w^eiglit having 
 an " arming" of grease or other adhesive substance by which 
 evidence is obtained that the lead has reached the bottom, 
 and what the character of the bottom is. The inventions of 
 
104 PHYSICAL GEOGRAPHY. 
 
 Brooke and others have resulted in instruments by which a 
 portion of the bottom is removed and brought up for examina- 
 tion. The difficulties to be contended with in sounding are 
 the uncertainty, first, as to whether the lead touches the 
 bottom.; second, as to whether the amount of line let out is 
 much in excess of the vertical depth; and third, as to whether, 
 supposing his lead sinks properly, the observer's position has 
 shifted during the descent of the lead. This last difficulty is 
 felt at sea only where no fixed objects are visible, the bear- 
 ings of which can act as a guide, the sinking of the lead from 
 a boat (which is, under certain circumstances, less liable to 
 shift than a ship) diminishing, not removing, the chance of 
 error. Experienced observers are, for the most part, con- 
 scious of the contact of the lead with the bottom, or at least 
 of a momentary change of speed in the line as it passes 
 through the hand, even in very deep soundings ; but currents 
 may sway the line, so that it describes curves which greatly 
 exaggerate the depth. Nor does the absence of sand or other 
 matters adherent to the arming prove that the weight has 
 not touched ground; the "Hydra," "Bulldog," and other 
 machines can scarcely come up empty if once they have 
 touched soft soil. The record, " no bottom," on a chart, is 
 only negative evidence, showing that the observation is 
 imperfect, either because of the inexperience of the observer, 
 or because currents have drifted the line, or, where the 
 cleanness of the arming has been relied on, because the lead 
 has touched hard and smooth bottom. Such records are 
 especially untrustworthy when, as usually happens, they 
 suggest enormous depths. 
 
 68. Atlantic Soundings. — The Atlantic Ocean has been 
 carefully surveyed for the purpose of' ascertaining the nature 
 of the bottom on which it was proposed to lay the tele- 
 graphic cable; and, more recently, j)art of the northern 
 basin has been carefully sounded by the Porcupine and 
 Lightning Expedition, while the United States Coast Sur- 
 vey has contributed most valuable observations regarding 
 the western area. The following analyses of the soundings 
 across the N, Atlantic will show the general contour of that 
 region. 
 
ATLANTIC SOUNDINGS. 105 
 
 Section between Labrador and the Orkney Islands, 
 BY Iceland. 
 
 Lahrador, 
 67°30' Lon. W. 
 
 Greenland. 
 50' 45"— 43' 32° 25°30' 23°30' 
 
 2032 1572 117 203 
 
 Iceland. Faroe, 
 22'o0'— 16° 13'30' 13° 10' 8° 5°aO' Orkney. 
 
 632 850 2G9 181 6S3 
 
 Section from Ireland to Newfoundland. 
 
 Newfoundland 61° 44° 39° 32' 2a°30' 20° 19° 14' 
 
 161 2385 2424 1550 2400 1575 210U 216 
 
 Section bet'ween United States and France, ' 
 
 U. S. Plateau, French Plateau, 
 
 71°— 57\ 52° 49°30' 47° 25°, 17° 10° 10"— 5° 
 
 Maximum <> '^^ <^ %^ % -S 
 
 240 1250 650 2760 1600 2510 107 Maximum 84 
 
 In the northern section it appears that the curved line it 
 follows traverses an undulating surface, the greatest de23th of 
 which is in 32° lon. W. To the south, the greatest depth 
 is between 39° and 44° W., while the traverse from France 
 to the United States gives the maximum soundings, 2760 
 fathoms, at 42° and 17°. A basin on the eastern side show^s 
 its greatest depth in 19° and 26° W.; but northwards the 
 two basins are broken uj) by the banks of land forming the 
 Faroes, Iceland, and Greenland, the whole ocean in this region 
 being generally shallower. Iceland is a prominent peak of 
 a long ridge which separates the eastern and western basins 
 as far south as 40° lat. N., and Avhich is of nearly uniform 
 height for the greater part of its length. This feature of the 
 sea floor is curiously parallel to the leading lines of heights 
 on the American Continent, and if it should prove to be con- 
 nected with the volcanic West Indian group, the relation would 
 be one of great interest as increasing the number of meri- 
 dional bands of disturbance connected by an equatorial band. 
 
i06 Physical geographic 
 
 Some instruction as to tlie Atlantic depths may he obtainecl 
 from the subjoined section between the Cape of Good Hope 
 and England; but it must not be taken as true for any other 
 than one line of soundings. The undulations here shown are 
 remarkable, the alternations of deeper and shallower succeed- 
 ing each other with singular regularity. St. Helena separates 
 two slopes of unequal but gentle inclination, that to the south 
 being at the rate of 1 in 150, that to the north reaching 
 2350 fathoms in 2° N. lat., that is, at a rate of 1 in 314. If 
 a line of 5 inches be draAvn, this inclination would raise a 
 line representing it J^j of an inch above the horizontal plane 
 at one end, while, on the same scale, the slope between 2° 
 and 37° N. lat. would scarcely cause an appreciable thicken- 
 ing of the pencil line. The last slope from 2600 fathoms to 
 the English coast slightly misrepresents the facts; for the 
 line of 100 fathom soundings, the limit of the British plateau, 
 extends so far south that the inclination from its surface to 
 the depth mentioned would be in reality about 1 in 15. The 
 principal fact to be gathered from the above, and from othei 
 incidental statements scattered through books, is, that the 
 South Atlantic and IsTorth Atlantic basins have a line of 
 maximum depth which, in the latter area, bifurcates north- 
 wards, and probably does so in the south likewise, but the 
 results of the Challenger Expedition will give more accurate 
 information on this point. 
 
 Section from Cape of Good Hope to England. 
 
 Cape of Good Hope. 2900 1300 2S0O St. Helena IGOO 
 
 1 ill 73 lines 1 in 212 1 in 150 1 in 352 
 
 S. Lat. 34° 28° 26° 20° 12° 0° 
 
 2350 1400 2200 1610 2000 England. 
 
 Iin94 linSSO 1 in 176 1 in 179 1 in 106 1 in 27 
 
 2' N. Lat. 37° 41° 44° 47° 50° 
 
 69. Mediterranean Soundings.— The Mediterranean sec- 
 tion is as follows : — ■ 
 
 Section from Gibraltar to Egypt. 
 
 Gibraltar. 9C5 1535 234 Maltese IGO 2170 1100 1460 1S90 Egypt. 
 
 linllS lin29S linl56 Plateau. Iin56 linlG6 lin300 lin78 lin37 
 
 E. Long. 2" 3-30° 9 30° 15 -30° 10° 24° 27° 2S° 30° 
 
fORM OF THE OCEAN FLOOR. lO* 
 
 To tlie tliree basins liere sliown a fourtli should be added, 
 tlie islands of Corsica and Sardinia dividing the deep trough 
 between Spain and Italy. The Adriatic may at one time 
 have equalled these two eastern depressions, but long ages of 
 deposit from streams laden with Alpine detritus have con- 
 verted that arm of the sea into a very shallow valley. 
 
 70. Indian Ocean. — No continuous sections of the Indian 
 Ocean are accessible, but the following will give a general 
 idea of the form of the basin, or valleys of its eastern and 
 western troughs. 
 
 Sectioi!? from Aden to Bombay. 
 
 Bombay Plateau. 
 
 Aden, 
 
 
 1470 
 
 1020 
 
 2170 
 
 125 
 
 
 
 1 in 151 
 
 1 in 78 
 
 1 in 514 
 
 lia7 
 
 E. Long. 
 
 45' 
 
 51-50° 
 
 52-30' 
 
 • 67° 
 
 7r 
 
 Section from Ceylon to Penang. 
 
 Ceylon, 2340 1455 . 2320 545 950 59 PcnaDg Plateau. 
 
 1 in 48 1 in 252 1 in 81 1 in 15 1 in 153 1 in 85 
 E.Long. 85° 30' 90° 30' 92° 30' ... 93° 35' 95° 30' 98' r ,101' 
 
 These two sections agi'ee in representing the troughs a9 
 having their shortest slopes on the west, while their eastern 
 shores terminate in submarine plateaux. That of Bombay 
 runs out for over 100 miles before the depth of 100 fathoms 
 is exceeded; while the Penang plateau has less than 70 
 fathoms at the same distance from land. 
 
 71. Pacific Ocean. — Of the Pacific there are no soundings 
 sufficiently connected to furnish satisfactory tables, but it 
 seems certain that nowhere has a greater depth been found 
 than 3000 fathoms; so far, therefore, as our present know- 
 ledge goes, the oceans have a remarkable uniformity in their 
 vertical measurements^ Such observations as are reliable 
 mil be found in a subsequent section, under the heading 
 " Temperature of Ocean." 
 
 72. Form of the Ocean Floor. — It has been customary 
 to speak of the sea bottom as the counterpaii; of the land as 
 regards its contours, the assumed resemblance being a part 
 of the notion already alluded to^ that the ocean and the air 
 w^ere parallel in their relations. But it is now certam that 
 this resemblance holds good only for moderate depths. When 
 
108 PHYSICAL GEOGRAPHY. 
 
 an estuary widens, it retains the valley form wliich it once 
 had as a result of its denudation when exposed as dry land. 
 When we pass to depths exceeding 100 fathoms, the sea 
 floor presents inequalities, but these are of a very gentle kind; 
 no deep chasms, no abrupt descents, but long undulations, 
 and great areas the surface of wliich, when drawn ©n a true 
 scale, entitles them to be regarded as plateaux, and to be 
 compared with justice to that recent sea bed, the Sahara. It 
 is difficult to account for this in a satisfactory way. It is 
 true that the sea bed, once the surface of dry land, has under- 
 gone marine erosion as it sank, and that the process was 
 repeated during the oscillations wliich preceded the final 
 submergence of such an ocean iloor as that of the Atlantic. 
 The plane of marine denudation thus formed would, so far 
 as known, undergo no subsequent change, since the move- 
 ment of water at great depths cannot be, and is as a matter 
 of fact known not to be such as to produce any denudation, 
 especially as it had no pebbles to carry along and use as 
 grinding tools. Even granting the extreme power that has 
 been claimed for the Gulf Stream, this mass of water can 
 only exert its power over a limited area, and its effect would 
 be to create those features whose absence is so remarkable. 
 It would be impossible to imagine a series of gulf streams, 
 or similar currents, to have planed down the floor of all the 
 great oceans; and it is unnecessary, since the denudation 
 during descent would account for much at least of the result. 
 The subject is a difficult one, but the explanation suggested 
 is the best at present available. The student must again be 
 cautioned against forming his opinions from sections in Avhicli 
 the horizontal and vertical measurements are not on the same 
 scale. 
 
 73. Deposits on Ocean Floor. — Another assumption has 
 also been disproved by recent research, namely, that the sea 
 floor at a certain distance from land is bare and rocky. This 
 has been found in a small number of cases; but the sea floor 
 is almost universally covered by a layer of fine materials, 
 for the most part organic in their origiu. The oaze of the 
 Atlantic has already been spoken of as a fine calcareous mud, 
 representing the debris of animals covered with tests, and 
 containing also the siliceous cases of other similar organisms. 
 
SPECIFIC GRAVITY AND CONTENTS OF WATER. 109 
 
 The Indian Ocean and the Pacific have yielded afc their 
 greatest depths similar materials, though in the Ked Sea 
 and Indian Ocean the sand blown from the deserts has con- 
 tributed inorganic matter, and the ashes of volcanoes, slowly- 
 settling, have furnished a small amount of felspathic ingre- 
 dients. But allowing for such admixtures, which would, of 
 course, be greater in j^roportion to the vicinity of land, recent 
 observations bear out the generalisation that limestones, not 
 consisting exclusively of coral, are formed in deep water, and 
 that their horizontal area and vertical thickness are in the 
 direct ratio of their distance from shore. 
 
 74. Specific Gravity and Contents of Water. — Perfectly 
 pure water may be regarded as only an artificial compound, 
 the normal condition of all known waters being to contain 
 varying quantities of other chemical substances ; and we are 
 justified in calling this normal, since on the admixture de- 
 pends the importance of water in the physiology of the 
 earth. The rain as it descends absorbs carbonic acid gas and 
 ammonia, in quantities varying with the season, being less in 
 winter when the decomposition of organic substances is 
 least. Thus, on reaching the earth, the water is prepared to 
 act on the surface on which it falls, and to minister to vege- 
 table life by the nitrogenous matter it contains. The water 
 of rivers varies in composition according to the rocks over 
 v/hich it flows, and the amount of orga.nic matter furnished 
 to it by surface drainage. The Thames drains a region of 
 which the chalk is a princij)al rock, and at Kingston the 
 waters of the river are found to contain 19 grains of dis- 
 solved matters per gallon. The substances which thus pass 
 down, invisibly to the eye, are carbonates of lime and mag- 
 nesia, sulphates of lime, potash, and soda, chlorides of sodium 
 and potassium, silica, and traces of alumina, iron, and phos- 
 phates. Mr. Prestwich estimates the total quantity of solid 
 matter thus carried away in solution at 548,230 tons per 
 annum. Of the 1502 tons which are thus daily carried past 
 Kingston, carbonate of lime constitutes two-thirds, or about 
 1000 tons, sulphate of lime 238 tons. This estimate is apart 
 from the solid matter, mineral or organic, which is held in 
 suspension, and which may be taken at 1 '68 grains per gallon, 
 or (the daily discharge of the river at Kingston being 1250 
 
110 PHYSICAL GEOGRAPHY. 
 
 millions of gallons) 300,000 lbs., or 134 tons in the twenty- 
 four hours, being 48,91 tons in the year. This example shows 
 the greater importance, from a physiological point of view, 
 of the invisible contents of Avater, and indicates the function 
 of water as a distributer of fresh materials. But for this 
 source of supply, chemical equilibrium would tend to be 
 reached; the new matter, however, maintains chemical action 
 and all those processes, organic and inorganic, which make 
 up the life of the earth. In 100 parts of sea water 96*473 
 are water, the remainder salts, namely, chloride of sodium, 
 magnesium, and potassium, bromide of sodium, sulphates of 
 lime and magnesia. Carbonate of lime is not mentioned, 
 its quantity being very minute, though in the immediate 
 vicinity of a calcareous coast, as in the English channel, it 
 forms '0057 parts per 100 of water. The reason of the 
 small proportion in sea water compared to that poured in 
 from land, is doubtless the rapid using up of it by animals 
 and plants, the coral reefs and the more minute, but perhaps 
 more important, crustaceans and shell fishes. Taking Bischof s 
 estimate that the Rhine at Bonn would supply with lime a 
 mass of oyster shells covering four square miles to a depth 
 of a foot, it is obvious how readily the lime must be removed 
 from the sea at all points. The average proportion of chloride 
 of sodium may be taken as between 2 and 3 parts per 100; 
 but the proportion is liable to increase or decrease, in the one 
 case by evaporation, in the other by excess of fresh water. 
 The Dead Sea illustrates the extreme of saltness, the Black 
 Sea furnishes an example of a basin receiving a large supj^ly 
 of fresh water. The following table gives the specific gravity 
 of v/ater at various localities :— 
 
 Northern Ocean, '. j J q2G64 
 
 Southern Ocean, . ^f^l ^^^^^i^^^ranean, | ^y^^«*, J igl^ 
 
 Under tlie Equator, . 1-0277 Black Sea, . 1-01418 
 
 North Pacific, . . 1-0254S Bed Sea, . Si. i -no?"^ 
 
 South Pacific, .. . 1-02G53 | South 1 02/2 
 
 Indian Ocean, , , 1-02G3 
 
 These figures must not be taken as absolutely correct; for 
 in the Pacific, as an instance, the observations are few, and 
 
SPECIFIC GRAVITY AND CONTENTS OF WATER. Ill 
 
 though the lower specific gi-avity in the north, as compared 
 with the south, agrees generally with the observations in the 
 Atlantic, the difference between the Atlantic and Pacific is 
 not necessarily constant, the observations in the latter ocean 
 being neither sufficiently numerous nor made far enough 
 from the land to justify the generalisation that the waters of 
 the Pacific have a lower specific gravity than those of the 
 Atlantic. We must again look to the scientific explorations 
 now in progress for the settlement of this difficulty. That 
 the specific gravity of the surface waters may be lower than 
 that of deeper water, was first inferred from the observa- 
 tion of Edward Forbes; who explained the death of many 
 specimens dredged up, by supposing them to have been killed 
 by the fresher surface Vv^aters. Difference of pressure had 
 probably more to do with this phenomenon; but Buchan 
 quotes Dr. Pankin's observation, that the density of the 
 water at the mouth of Loch Fyne sank from its normal 
 1"0250 to 1-0210 after a heavy fall of rain, and in less 
 than twenty-four hours regained its usual value. Apart 
 from exceptional cases, where the sadden melting of great 
 masses of ice has flooded the adjacent sea with fresh water, 
 or a few constant phenomena, such as the influx of the fresh 
 waters of the Amazon and Orinoco into the ocean, it may 
 be remarked that the lagoon waters of some atolls, though 
 they rise and fall Avith the tides, the sea filtering through 
 the coral mass as through a sieve, are fresh; nor is this 
 remarkable, since it is pointed out by Wilkes in his rejDort 
 of the United States Exploring Expedition, that the surging 
 up and boiling over of a hot current, intercepted by a coral 
 barrier, gives rise to almost constant precipitation of rain. 
 The specific gravity of sui-face water has been found to bo 
 increased during the prevalence of high winds, an unexpected 
 result, since the surface waters are warmer after a storm, 
 their motion being converted into heat. The density of 
 water below the surface is not uniform; ordinarily lighter 
 water floats over denser, but the relation may be reversed if 
 the lighter water is introduced as a swift stream; thus the 
 fresh water in some cases underflows the sea water in tidal 
 rivers, and water of specific gravity 1"02G5 has been found 
 at 73 fathoms, while above it the specific gi'avity was 1*0270., 
 
112 PHYSICAL GEOGRAPHY. 
 
 Wyville Tliomson found at 49° 12' N. lat., 12° 52' W. long., 
 that tlie density increased from 1*0272 at 50 fathoms to 
 1-0277 at 800 fathoms; while at 54° 28' K hit., IV 44' W. 
 Ion., the surface showed 1'0280; the bottom (1425 fathoms) 
 showed 1*0269; and at another point the difference was, 
 surface, 1*0280; bottom (664 fathoms), 1*0272, the evapora- 
 tion of the warmer surface increasing its density. 
 
 Fresh water heated to the boiling point and allowed to 
 cool, contracts as it cools down to 4°C. (39 '2° F.); if the 
 temperature is still further lowered, it again expands till it 
 reaches 0°C. (32° F.), when it freezes. But salt water thus 
 treated does not attain its maximum density till it has 
 reached - 3*67°C. (25*4°F.), the freezing point of undis- 
 turbed, — 2*25° C, that of disturbed sea water. While fresh 
 water therefore expands under opposite extremes of heat and 
 cold, salt water expands from its freezing point uniformly to 
 its boiling point. 
 
 75. Pressure of Water. — A column of water 35 feet in 
 height is equal to a column of mercury 30 inches in height, 
 and these two represent the pressure of a column of aii* live 
 miles high, of equal density throughout. At the depth of a 
 mile, the pressure of the water equals that of 152 atmospheres, 
 and its bulk is reduced by y^; at greater depths the pressure 
 increases, so that at twenty miles the reduction in volume 
 would be y. But this acts injuriously only on compressible 
 bodies; thus air is greatly reduced in volume, but just as 
 animals on the dry land are not injured by the weight of 
 the atmosphere, so the ocean pressure does not affect those 
 animals whose bodies are supported inside by the same 
 clement as that which presses on them outside : the pressures 
 being equal, the animal is in equilibrium. Thus animals 
 are enabled to live at great depths, shell-fish having been 
 brought up from more than 2000 fathoms, while tlie same 
 species are found living also in very shallow water. But 
 though uninjured by pressure its sudden removal is fatal, 
 since the animals from great depths reached the surface dead 
 or dying, just as man would be killed by passing through the 
 u])per regions of the atmosphere. 
 
 76. Temperature of the Ocean. — The waters of the ocean 
 never siiik below -3°*5C. (24° F.), and very rarely rise at 
 
TEMPERATURE OP THE OCEAN. 
 
 113 
 
 the surface above 34° C, tMs exceptional temperature being 
 recorded near Aden. 
 
 Water has the highest specific heat of any known substance ; 
 thus, if a pound of mercury and a pound of water are both 
 raised 1°, it requires thirty -three times as much heat to 
 do this for the water as for the mercury, and if these are 
 cooled 1°, the water sets free thirty-three times as much heat 
 as the mercury. Tables have been constructed showing 
 the specific heats of different bodies. Taking the specific 
 heat of water as 1000, that of mercury is 33, or, as it is also 
 written, water 1*000, mercury 0*0333. 
 
 The following table shows the specific heat of equal weights 
 of the most important substances, the determinations being 
 those of Kegnault, cited by Tyndall : '^ — 
 
 Water, 1-000 
 
 Air, 0-237 
 
 Oxygen 0-218 
 
 Nitrogen, 244 
 
 Hydrogen, 3-409 
 
 Carbon, 0-2414 
 
 Diamond, 0-1469 
 
 Copper, 00952 
 
 Gold, 0-0324 
 
 Iron, 01138 
 
 74 
 53 
 12 
 
 Lead, 00314 
 
 Magnesium, 0-2499 
 
 Mercury, 0333 
 
 Potassium, 1 696 
 
 Silicon, 0-1774 
 
 Silver, 00570 
 
 Sodium 0-2934 
 
 Sulphur, native,. 1776 
 
 Tin 00562 
 
 Zinc, 0-0955 
 
 100 
 
 15 
 
 The figures in the second column give the relative conductivity of 
 the metals for heat. 
 
 The relation of land and water as regards heat, is such 
 that it requires four times the heat to raise water to tlie 
 same temperature as land; the quantity of heat therefore 
 •retained by the sea is greater than that retained by the 
 land, hence the sea never i-ises so high nor sinks so low in 
 temperature as does the land. A glance at the above table 
 shows that capacity for heat does not mean conductivity, the 
 molecular transfer by which each atom passes on the motion 
 to its neighbour being dependent on difterent conditions 
 from those of the specific heat. Water, though having tlie 
 highest specific heat, is a very bad conductor of heat, the 
 raising of a mass of v/atcr to the same temperature through- 
 out being effected by convection, or the transfer of masses of 
 
 23 
 
 Tyndall. Heat as a Mode of Motion, Chap. V, 
 
 11 
 
114 t>HYSICAL GEOGRAPHY. 
 
 heated water from place to place, not the transfer of motion 
 from atom to atom. 
 
 It will be api^arent from what shall be said hereafter 
 regarding currents, that no general statements can be made 
 which shall profitably represent the general temperature of 
 an ocean. The observations must be studied, map in hand, 
 and for each particular tract that has been carefully explored, 
 if it is desired to have a thorouQ:h knowledo-e of the surface 
 variations. From what has been said, it will not be unex- 
 pected that the surface temperature of the sea does not 
 folloAv that of the air in contact vv-ith it. To illustrate thia, 
 a table is subjoined, extracted from the deep sea soundings 
 of the " Porcupine." The noonday observations have been 
 taken for the month of June 1869, and they are arranged, 
 not in chronological order, but according to degree of west 
 longitude, between north latitudes 51° and 54° 30'. The 
 upper of the two lines of figures represents the sea tempera- 
 ture, the lower that of the air. Ail the temperatures are 
 
 centigrade. 
 
 
 Valentia. J^T 
 
 :o.5. yi;-: 11-3. i>|uM5.;»: ir ..-:?: 
 
 12° 20' 
 
 jr.y I2".<y jr 12- .3' l^$ 13-14' J»:$: 13-23'}^? 
 
 13- 4S' 
 
 I3.30 15- OS' ]-. 15-24' J;],^ 
 
 In 1870, Irminger gave tables shov/ing the alternations of 
 streaks of warmer and cooler water between Greenland and 
 Fair Island. These streaks vary in j)Osition and breadth, 
 and represent the interlacement of the southern w^armcr 
 and northern colder vv^aters. The Atlantic explorers have 
 shov.^n that the temperature of the sea sinks, but not 
 miiformly, till a constant temperature of 0°C. (32° F.) is 
 attained ab great dej)ths. But, as will be more fully stated 
 afterwards, this constant temperature is not every v/h ere 
 obtained; for the warm vraters may displace the colder in 
 the shallower parts, especially where channels exist between 
 outstanding land masses, and the bottom temperature may 
 thus be that proper to a lower latitude, while in the same 
 parallel of latitude very different bottom temperatures 
 co-exist. Captam Sliortland's observations of surface and 1 
 
 i 
 
TEMPEEATUEE OP THE OCEA^^ 115 
 
 bottom temperatures in the Indian Ocean, between Bombay 
 and Aden,"^ may be thus summed up : — ■ 
 
 On the Bombay Plateau. 
 
 Temperature of Air. ■* Depth. Surface Teruperatnre. Bottom Temperature. 
 
 22-6°- 24-7° C. 11-51 fms. 22-5°- 26-1" C. . 23-1- 25 8 C. 
 
 But the depths and temperatures are not regularly graduated; 
 thus the highest surface and bottom temperatures, 26-1° and 
 25-8'' respectively, correspond to a soimding in 45 fathoms, 
 and the lowest are in the shallowest water. Between Bom- 
 bay and Aden, the mean temperature of 15 days between 
 28th January and 12th February 1868 was 24°C., the sur- 
 face temperature 23-8°, and at 2170 fathoms 0-83°C. Captain 
 Shortland considered that variations in surface temperature 
 were not appreciable at greater depths than 100 fathoms, and 
 that the constant temperature might be said to begin at 1700 
 fathoms. These results agree generally with those of the 
 Atlantic Ex23edition, from which it appears that solar varia- 
 tion does not produce marked effects beyond 50 fathoms, 
 and the constant temperature begins about 1000 fathoms. 
 The rate at which the temperature sinks is shown in the 
 following series of observations, taken two miles north of the 
 equator, in 22° 16' W. Ion., the temperature of the air being 
 26-6°C. 
 
 Depth in fms, , 
 
 300 
 
 400 
 
 1000 
 
 2040 
 
 Temperature, 27 "2° C. 
 
 6-44°C, 
 
 5° a 
 
 3-3° C. 
 
 i-erc, 
 
 A remarkable anomaly is presented by the Mediterranean, 
 in which the temperature sinks during 100 fathoms to an 
 average of 13° C. (55 "5° F.), and this maybe taken as the 
 mean temperature of the Mediterranean below 100 down 
 even to 1500 fathoms. The Bed Sea is even more completely 
 land-locked than the IMediterranean, and its shores have a 
 higher average temperature. The bottom waters have a 
 mean of 21-6° C. (71°F.); now as the Bed Sea does not 
 exceed 1700 fathoms in depth, its temperature is that of the 
 water between 50 and 100 fathoms in the Indian Ocean. 
 Solar radiation therefore, doubtless, has an important pari 
 in raising the temperature, and to achieve this great hea^t it 
 
 * Journal GeograpJiical Societij, 1871, p. 53. 
 
116 PHYSICAL GEOGRAPHY. 
 
 is obvious that tlie horizontal movements must be very 
 slow, 
 
 Mr Prestv/ieli has collected and tabulated observations 
 in the Atlantic and Pacific which are here condensed, 
 the latitude but not the longitude of the stations being 
 given. 
 
 Temperatures of Atlantic. 
 
 Latitude 4-2''N. 29°N. 7°21'N. 4°25'N. 15°3'S. 25''10'S. 29°33'S. 82°20'S. 2S°12'S. 
 
 Depth (ft.) 4(5SS 8399 3030 6037 7200 5315 6310 644-1 2000? 
 
 Surface 17-7°C. 24'4° 26-G'' 27-r 25° 19-6° lOl" 21-9" 16-9° 
 
 Bottom C-G°C. 61° 2-2° S-g" 4-r S-l" 2-1'' 25° 3-1° 
 
 Tejiperatures of Pacific 
 
 Latitude 51°31'N. 2S''52'N. 18°5'N. 4°32'N, 0° 21°14'S. 32''57'S. 43°47'S. 
 
 Depth (ft.) 5741 3600 4261 12271 6000 5500 4692 6100 
 
 Surface ll^^C. 255° 24-7° 27-2° 80° 27-2° 16-3° 7-5° 
 
 Uottom 2-5X\ 9° _ 4'8° 17° 2-5° 22° 5*4° 2-3° 
 
 From the bottom temperatures a deduction of 2° or 3° has 
 to be made for error due to pressure on unprotected instru- 
 ments, but these figures show that a mass of water of low 
 temperature crosses the equator, and in the Atlantic at least, 
 the connection of the Arctic and Antarctic Oceans is pro- 
 bably due to the passage northwards of southern cold waters ; 
 for, as will be explained in describing the N. Atlantic 
 currents, the Arctic cold waters pass to the west side, and 
 are separated from those of the Bay of Biscay by a mass of 
 water warmer than either. The Arctic current enters the 
 Gulf of Mexico beneath the warm outflowing stream, but 
 the bottom temperature 4*1 6° C. between Cuba and Yucatan 
 must have a southern source. While the annual variations 
 of temperature cease to be felt at about 100 fathoms, the 
 daily variations do not probably extend beyond 10 fathoms; 
 but the influence of currents, and of local and temporary 
 disturbances due to configuration of the land, and to storms, 
 forbids our acceptance of such generalisations from the com- 
 paratively scanty data at our command. To the category of 
 premature inductions must also be referred tlie conclusion of 
 M. Aime, that the average temperature of the sea at a dis- 
 tance from land is 18°0. (Qi'-i^F.), that of the air being 
 18'22='a (G4-8°F.). 
 
 77. Colour of the Water. — Tyndall has shown that water 
 and aqueous vapour have the same absorptive power, and he 
 
COLOUR OP THE WATER. 117 
 
 regards the blue colour of distant hills as deep in proportion 
 to the aqueous vapour in the atmosphere. It is also a well 
 known fact that the blue colour of the sea is intense in propor- 
 tion to its saltness, and that the fresh water lake of Geneva is 
 of a very deep blue. The blue of the atmosphere, and of the 
 Lake of Geneva, might be capable of explanation by reference 
 to the fine particles of solid matter with which both are 
 charged, these particles reflecting the short blue waves of 
 light most readily. But the Avaters of the open oceans do 
 not contain these particles to the same amount, and in pass- 
 ing from open sea to land, the blue gives place to green, and 
 that to greenish yellow; while close to the shore, as the solid 
 matters increase in quantity, the tint becomes that charac- 
 teristic of the rocks or soil whence the fine detritus is derived. 
 The problem is not yet near a solution, and the observations 
 must remain in seeming contradiction till the key to their 
 harmony shall have been found. Closely associated with 
 colour is the permeability of the sea to light. Though it 
 cannot be supposed that the open ocean contains as many 
 suspended particles as a glacial lake, analogy would lead us 
 to expect that it contained such particles perhaps even more 
 abundantly than the atmosphere. The density of the Avater, 
 and this fine cloud, would together form an obstacle to tho 
 passage of light, and there is reason to believe that the 
 actinic rays are arrested at a comparatively short distance 
 beneath the surface. Nordenskiold found a sensitive plate 
 unchanged after twelve hours' exposure at the bottom in 
 79° 54' N. lat. amid luxuriant seaweeds, the temperature of 
 the bottom was 2"C. Plants, too, are not met with at greater 
 depths than 200 fathoms, and as their power of liberating 
 oxygen is associated on land with abundance of light, their 
 paucity between 50 and 200 fathouis, and their absence 
 beyond the lower limit has been held to favour the view 
 that light has very sparing access, if any, to the deej^er 
 waters. But Mr J. Y. Buchanan has ascertained that sea 
 water has a very retentive grasp of carbon dioxide (carbonic 
 acid), and this, rather than defect of light, may account for 
 the early disappearance of plant life. This question likewise 
 awaits further investigation, and it is useless to speculate in 
 the absence of data, 
 
118 PHYSICAL GEOGRAPHY. 
 
 78. Luminosity of the Sea. — The so-called pliospliorescence 
 of the sea has given rise to as much imaginative scientific as 
 poetical writing. The tints are lu^anium green, violet, lilac, 
 crimson lake, rarely azure, still more rarely white, and they 
 appear as steady stars, as lambent light moving along the 
 bodies of star fishes and other animals, or as intermittent 
 flashes. Various microscopic bodies possess this property, 
 the source of which is obscure, and equally so are the con- 
 ditions on which its often fitful display depend. Probably 
 variations in temperature, and stillness of the water, per- 
 mitting or forbidding the approach of the minuter organisms 
 to the surface are the most important; the phenomenon is 
 often well developed during magnetic storms. 
 
 SECTION II. 
 
 Movements of Water — Waves — Tides — Tidal Wave not a Wave of 
 Translation — Progress of Tidal Movement — Currents : Streams, 
 Drifts, Indraughts, Equatorial Drift — Gulf Stream and N. 
 Atlantic Basin — Movements beyond 45° N. Lat. : their Cause — 
 Kennel's Current — Mediterranean Currents — Baltic — N, African 
 Currents — Sargasso Sea of N. Atlantic — S. Atlantic— Antarctic 
 Drift — Pacific Currents — Pacific Equatorial Drift — Australian 
 Currents — Currents of Indian Ocean — Red Sea. 
 
 79. Movements of Water.— If the globe were uniformly 
 surrounded by a sphere of water subject to no external dis- 
 turbance, the water would move from west to east with the 
 velocity proper to each parallel of latitude. But this theo- 
 retically simple motion is disturbed by the projection of 
 masses of land; by the unequal temperatures of the oceans; 
 by the influence of steady or occasional winds; by the 
 introduction of fresh water streams from the land; by 
 subterranean movements ; and by the influence of the 
 sun and moon, to which are due the phenomena of the 
 tides. 
 
 80. Waves. — Water acted on by the wind is heaped up to 
 leeward, the friction between the moving air and the water 
 carrying forward the pai'ticlcs of the latter till they overpass 
 
WATES. 119 
 
 the limit of equilibrium and fall over, the phenomenon being 
 seen in its simplest form -where wind passes over dry sand. 
 A continued ■wind gradually involves successively deeper 
 layers in this forward movement, till finally a distiu-bance 
 is created which survives for some time the existence of its 
 cause. Two different kinds of movement are unfortunately 
 rej)resented by the word wave, and these must be distin- 
 guished. A wave is a pulsation or vibration among the 
 particles of a fluid; the particles tend apart and again return; 
 there is no displacement in any except a perpendicular plane. 
 "Water thrown into vibration, say by an impulse from below, 
 starts ujo wards; and if the blow be violent enough, a j^ortion 
 springs into the air. If the force of the blow is not sufficient 
 to overcome the cohesion of the liquid, the disrupting ten- 
 dency of the vibration is masked, but it still finds compensa- 
 tion in sinking as far below its original level as it had been 
 raised above it. A wave of water consists of two points 
 of rarefaction and one of compression — the rarefaction points 
 being at the crests of two j^^dses, the compression point at 
 the lowest point of the intervening trough. Pulse is here 
 used in the sense which it has in acoustics, but the sound- 
 wave has its compression point at; the apex of the pulse, its 
 rarefaction point in the interval betv,^een the two pulses. 
 The length, then, of a wave is from crest to crest, the height 
 from crest to trough. But it must not be imagined that 
 there is no horizontal movement; the particles of which the 
 wave is composed revolve in circles, returning to the spot 
 whence the}'- started. The length and height of waves vary 
 from the slightest ripple to those in whose trough a large 
 ship may be cradled, or Avhose crests are so far apart that a 
 lorn? vessel lifted on the summit of one has its back broken. 
 The circles in which the particles revolve are larger in pro- 
 portion to the magnitude of the disturbance, or, in other 
 words, to the disruptive tendency imparted to the pulse of 
 water. Whatever tends to compress the water diminishes 
 the diameter of the vertical circles which the particles de- 
 scribed ; hence the depth to which wave motion is propagated 
 depends on the amount of disturbance which, if slight, only 
 elevates and rarefies a small layer of water; it' great, rarefies 
 a thicker stratum. It follows from wliat has l)een said, that 
 
120 PHYSICAL GEOGRAPHY. 
 
 a column of water, starting from a position of rest, describes 
 a seiies of oscillations before returning to rest; the limit of 
 oscillation is tbe diameter of the circle in which the indi- 
 vidual particles revolve, and infinitely minute as the oscilla- 
 tion may be, it confers the tendency to actual translation or 
 forward movement which was alluded to in the first lines of 
 this paragraph. The rarefied surface layer, carried forward 
 by a wind strong enough to overcome its cohesion, exposes a 
 fresh surface to be rarefied by elevation, and thus the wave 
 and its oscillation become propagated downwards. The 
 largest waves recorded are those whose height was estimated 
 at between 30 and 40 feet from trough to crest, but masses 
 of inferior size, when arrested by an obstacle, possess enor- 
 mous force. The waves of British seas probably never 
 exceed 8 feet, even ofi" the west coast. But this height may 
 be exceeded close to shore, and a wall of water 8 or 10 feet 
 high, pushed over by the wind, strikes on the land with for- 
 midable violence. The pressure on the square foot has been 
 calculated for the Atlantic breakers on the west coast of 
 Scotland at 611 lbs, in summer, 2086 lbs. in winter; and a 
 pressure of about 6720 lbs. on the square foot was necessary 
 for the ground swell to cast its spray over the Bell Bock 
 lighthouse, on the east of Scotland, the height of the tower 
 being 112 feet. The gentlest wind ripples, and the waves 
 caused by other influences than the wind, are even curves of 
 v/ell-nigh symmetrical form; the spectator may look towards 
 or look after the wave without seeing an appreciable dififer- 
 ence; but with the velocity of the wind the steepness of the 
 leeward face of the wave increases, the front being a dark 
 green clifi", the back of the wave a long curve of smooth blue 
 or green water, a distinction well enforced by Buskin. 
 Waves whose crests are torn off by the wind, and which 
 tumble over in foam, are called breakers ; and these on fixed 
 obstacles, as a river bar or fringing reef, maintain a constant 
 mass of foam — the surf — which it is dangerous and sometimes 
 impossible to pass. When a wave has broken on the shore 
 it rolls back, the slope of the beach increasing the fall of 
 what corresponds to the trough of a wave. In caves, the 
 wave dashed to the roof crushes out the air, and when it 
 sinks sucks down what may be loose in the roof, thus tend- 
 
TIDES. 121 
 
 ing to produce bullers or blowholes, opening vertical sliafts 
 some little way from the edge of the sea-cliffs. After a storm 
 has subsided, the wind-driven waves with steep leeward faces 
 disappear, and the agitation survives in proportion to the 
 severity and continuance of the storm. The ground-swell, 
 as this residue of a storm is called, testifies to the depth to 
 which the waves have been propagated, while its pressure 
 suggests the force of the previous motion. The transporting 
 power of waves on the shore, both in their forward move- 
 ment and in their backdraught, is very remarkable. Blocks 
 of 504 cubic feet (about 40 tons weight) have been carried 
 6 feet; blocks of more than 200 tons have been turned over 
 and broken; masses of 20 tons have been carried 80 or 90 
 feet. Of course, these masses are of easier transport in 
 water, but they are still remarkable proofs of the power of 
 water. A more gentle but more important influence of 
 waves is seen in the travelling of beaches, the stones of 
 which move forward according to the set of currents and 
 wind-waves, become piled up against an obstacle, and often 
 render vain any effort to check their progress. 
 
 81. Tides. — That the sun and moon are associated with 
 the phenomena of the tides is well known; that the method 
 in which these phenomena are produced is known does not 
 seem so certain. The discussion of the tides belongs pro- 
 perly to astronomy, but it is necessary to indicate here the 
 ' bearing of the problems it involves on the history of the 
 earth. Twice every day the waters of the sea rise and fall, 
 or flow and ebb, on the shores at high and low tides; the 
 interval between high tides is 12^- liours (nearly), so that 
 each successive tide is later than that before. Once every 
 month the difference between the level of high water and 
 low water is at a minimum, and the difference gradually 
 increases till the level of high water is higher, of low water 
 lower, than at any other time. These monthly minimum 
 and maximum tides are neap and spring tides. They corre- 
 spond to, but do not coincide with, the positions of the sun 
 and moon relatively to the earth. The attraction of the 
 moon is greater on the particles of water immediately 
 beneath it than on the centre of the earth, still greater than 
 on the surface particles on the opposite side of the earth. 
 
122 PHYSICAL GEOGRAPHY. 
 
 The same is true for tlie sun, but tlie more powerful attrac- 
 tion of. the sun is counterbalanced by its greater distance, and 
 the tidal elevations due to moon and sun are in the ratio of 
 5 : 2. As ea»ch planet forms a tidal protuberance on the side 
 of the earth towards and away from it, so a correspond- 
 ing flattening occurs midway between the protuberances. 
 Now the moon describes her apparent diurnal circuit in 
 about 25 hours, and the sun in 24 : the solar tides are there- 
 fore at 12 hour intervals, the lunar at 12h. 24m. intervals; 
 thus the solar tide will overtake the lunar. If, therefore, 
 the attraction of both planets affects the same points, as 
 v/hcn the sun and moon are on the same or opposite sides of 
 the earth, the protuberances due to each will coincide, and 
 the rise and fall of the tides will be ,at a maximum; if the 
 sun and moon are so placed that the long axes of the sphe- 
 roids formed by their tidal protuberances are at right angles 
 to each other, the solar protuberance will coincide with the 
 lunar flattening, and the opposite influences reduce the differ- 
 ence of tidal elevation to the minimum, giving neap tide. 
 The height of the spring tides will obviously be affected by 
 the distance of sun and moon from the earth. Again, as the 
 orbital planes of sun and moon are oblique to the equator, 
 and as the attraction of these bodies is on the point imme- 
 diately beneath, the northern or southern declination of the 
 sun and moon will also affect the height of the tide at any 
 point. But it must be remembered that the sea thus raised 
 and depressed is not a perfect sphere, but broken up by the 
 land into irregular spaces; that there is frictional resistance 
 to the motion of the water, whose particles rub against each 
 other and against the earth, and that the earth itself rotates. 
 The tendency of sun and moon to produce the theoretical 
 tide is not fully realised. The tidal elevation does not coin- 
 cide with the passage of sun and moon across any point, 
 neither does it represent the whole elevation due to one of 
 the planets; but the joint heights of the two elevations give 
 a maximum at a point intermediate between them. Hence 
 the length of the tidal wave varies as the two attractions 
 approach or recede from coincidence, the variation beiug 
 known as the priming and lagging of the tides. It is 
 delayed by friction, so that the spring tide, when the sun 
 
PROGRESS OF TIDAL MOVEMENT. l23 
 
 and moon are in conjunction or opposition, is later tnan new 
 or full moon by an interval of more than 12 hours and less 
 than 3 days. This frictional resistance tends to retard the 
 rotation of the earth round its axis by 22 seconds in a cen- 
 tury (Adams, Tait, and Thomson). This retardation, it is 
 supposed, would in time reduce the earth to the condition of 
 the moon, turning only one side towards it. On the de- 
 creasing rotation of the earth, Sir W. Thomson rests his 
 speculations as to the former condition and probable future 
 of the earth. 
 
 82. Tidal Wave not a Wave of Translation. — Revert- 
 ing to the definition of a wave (Art. 79) as a pulsation or 
 vibration of the particles of a fluid which tend apart and 
 return to their previous relations, the tidal movement in the 
 open ocean fulfils this condition. The particles of the water 
 are raised up and let down, just as the bits of paper rise 
 and fall as the heated sealing wax is carried over them. The 
 tidal wave- is a superficial shape of the v/ater, and high and 
 low water occur according as the upper or lower portion of 
 this elevation occurs on our coasts. But when water is 
 heaped up in any place in which it cannot freely resume its 
 customary form, it is converted into a wave of translation, 
 which either travels swiftly over shallow water, or passing 
 through a narrow channel forms a rapid current. The estimated 
 rise of the tidal wave in open ocean, as in the Pacific, is 
 about 2 feet, the observed seems to give a mean of about 9 
 inches, and the difierence is due, it has been suggested, to a 
 similar movement in the solid earth itself, an hypothesis 
 which must as yet be received with caution. But the observed 
 height of the tidal rise on the shores of continents greatly 
 exceeds this amount, reaching its maximiun in the Bay of 
 Fundy, into which a mass of water 100 to 120 feet deep is 
 poured. The Bf^re of the Severn, the Hoogly, and other 
 great rivers, is the tidal wave heaped up in a narrow passage 
 and rushinsj forward in a huoje roller which breaks into foam. 
 Where long flats shoal the water for some distance, as in the 
 Solway, the tidal wave rushes forward with velocity as a 
 great bank of water. 
 
 83. Progress of Tidal Movement. — The tides travel with 
 various speed. In the open ocean swiftly, but ia narrower 
 
124 PHYSICAL GEOGilAPHV. 
 
 seas v/itli less speed and greater mass ; and friction al resist- 
 ance makes its effects very obvious on a chart showing the 
 co-tidal lines, retardation being obvious on the east side of 
 the north Atlantic basin, and in the Atlantic, as a whole, 
 compared with the freer Indian Ocean. The increase of 
 velocity due to local circumstances is illustrated by the eagre 
 or bore of the Tsien-Tang in China, the tidal wave advancing 
 up the river for 80 miles at the rate of 25 miles an hour. 
 The strong streams through the Pentland Firth and similar 
 confined channels are examples of extreme velocity ; but it 
 is only in such cases as the cul-de-sac of the Bay of Fundy 
 that the ebb tide can have any important power as a denud- 
 ing agent. The Atlantic and Indian Oceans, in which the tides 
 advance from south to north, evidently do not illustrate the 
 antipodal character of the tides, and in them the necessity 
 is ver}^ clear for ascertaining accurately what is known as 
 the Establishment of the Port. By this is meant the interval 
 between the passage of the sun and moon across the meridian 
 of the port and the occurrence of spring tide. This interval 
 varies in every locality, and Herschel urges the necessity for 
 distinguishing with care between the slack water, when the 
 tidal wave does not run in either direction, and the time at 
 which the tide ceases to rise or fall. 
 
 84. Currents, Streams, Drifts, Indraughts. — The move- 
 ment of a body of water in a definite direction, whether the 
 direction and velocity are constant or liable to variation, 
 constitutes a current, and by means of these an oceanic 
 circulation is maintained, the heat of tropical, the cold of 
 polar regions being interchanged and distributed. Currents 
 are of very different magnitudes, and it is as difficult to 
 classify them by size as by reference to their probable origin; 
 but in this and the following paragraphs it is proposed to 
 consider only the great currents whose constancy and magni- 
 tude render them as important from a climatal as from a 
 nautical point of view. From what has been said of the 
 tidal wave, and from the investigation of the causes of the 
 gi-eat currents, it will be easy for the student to understand 
 the existence of the minor, local, and often extremely uncer- 
 tain currents, the enumeration of which belongs to the 
 province of Descriptive Geography. 
 
EQUATORIAL DRIFT. 125 
 
 The discussion as to the origin of currents has brought 
 into prominence two phrases : current and indraught ; the 
 former is used by those who ascribe the primary movement 
 to the influence of prevailing winds, indraught being the 
 opposite movement, by which equilibrium is restored. Those 
 who trace the oceanic circulation to differences of temperature 
 regard all movements as currents, and urge that a difference 
 of speed is not sufiicient to define an indraught. But no defi- 
 nition by typical characters is here attempted, the terms are 
 used to indicate a difference of origin; the primary current 
 having its impulse imparted from without, the indraught 
 being secondary to and dependent on the existence of the 
 primary movement. 
 
 Currents are usually divided into drifts or the primary 
 wind movements, and streams or modifications of the drifts 
 effected by features of the land or of the sea bottom j the 
 indraughts are supplementary to both.* 
 
 85. Equatorial Drift. — The equatorial drift, commencing 
 south of the equator, off the west coast of Africa, crosses the 
 Atlantic, and, spreading out westward, splits on Cape St. 
 Koque, one branch passing southwards along the Brazil 
 coast, the other following the land north-west towards the 
 Caribbean Sea. The St. Boque Current has an average 
 velocity of 30 to 40 miles a day, the speed of the equatorial 
 drift being on an average 20 to 35 ; but before entering the 
 Caribbean Sea. a speed of 80 miles may be attained. It 
 follows the north-westerly bend of the coast, sweeps the 
 waters of the Amazon and Orinoco along with it, and enters 
 the Culf of Mexico through the Straits of Yucatan. A part 
 of the current is arrested by the Mosquito coast, and forms a 
 backwater which sweeps eastward round the bay to rejoin 
 the main stream off New Granada. The chief part of the 
 current spreads over the Mexican Gulf, and returns towards 
 Cuba, where it is joined by a portion which clung to the 
 Cuban coast, the united stream rushing through the Straits 
 
 ' * In speaking of ocean currents, eastward, westward, northward, 
 and southward are invariably used to indicate the direction in which 
 they move ; and the direction whence the winds come is indicated 
 by the ordinary terms. But the student must remember that in the 
 language of navigation a northerly wind comes from the north, but «% 
 northerly current sets towards the north. 
 
126 PHYSICAL GEOGRAPHY. 
 
 of Florida as a great river of warm water 30 miles broad, 
 2100 feet deep, and with a surface velocity of 4 miles an 
 hour. The temperature of the equatorial drift is 23*^0., the 
 initial temperature of the Gulf Stream is 30^ C, the difference 
 representing the amount of heat received in the Mexican 
 Gulf In the equatorial region a bottom temperature of 
 0°0. has been found constant, and in the Florida Straits a 
 temperature of 4-5°C. proves that cold water underlies the 
 warm. 
 
 86. The Gulf Stream and North Atlantic Basin.— The 
 mass of waters flowing past Cuba meets the equatorial drift, 
 and is deflected northwards through the Straits of Bemini, 
 following thereafter the curve of the coast as far as Cape 
 Hatteras, and joining the eastward drift at the 40th parallel, 
 its velocity there being 30 miles a day. It is separated from 
 the land by a belt of water of varying breadth, but which 
 gradually widens northward, and corresponds generally to 
 the line of deep soundings. The eastern boundary is not 
 well defined ; as on the west the cold wall forms a very 
 sharp line of demarcation, and as, a little northward, cold 
 water lies in mass below the stream, it flows as a river in a 
 channel bounded by water of lower temperature. Laughton 
 has pointed out that its surface is like that of a river convex; 
 but the friction being greatest along the cold wall and least 
 on the eastern side, the summit of the convexity is close to 
 the western side, just as at the bend of a river it is close to 
 the bank against which it is thrown. It is this easterly slope 
 of its surface which prevents wreck from the West Indies 
 from being stranded on the American coast ; a fact otherwise 
 inexplicable since surface movements in various directions 
 have been recorded. The temperature of the stream is not 
 uniform even in the early part of its course, for, in sections 
 at right angles to its flow, the American Coast Survey has 
 found alternate bands with a difference of 1° to 2°C. ; thus 
 passing seaward from the cold wall, the maximum tem- 
 perature of 27'8°C. prevails for 60 miles, for the next 
 30 miles the temperature is 25*5°C. ; the high tempera- 
 ture, 27 ^"C. occupies the next 170 miles, and is succeeded 
 by 25-5°C., then 27-4°C. The cooler bands correspond to 
 the position of deeper channels in the sea bed, parallel 
 
HOVEMENTS BEYOND 45° N. LAT. 127 
 
 to tlie coast, and containing tliickei- masses of tli3 polai' 
 current. 
 
 87. Movements beyond 45^ N. Lat. : their Cause. — 
 Beyond 45° N. lat., the course of the stream has been made 
 subject of dispute. According to the generally received 
 opinion, it spreads over the Atlantic towards the European 
 shores; a portion is deflected to the Gulf of Guinea, and 
 rejoins the equatorial drift; the rest is traceable northwards 
 between Iceland and Norway to Spitzbergen, on whose 
 shores West Indian fruits have been found, and it was even 
 thought to reach the coast of Siberia. But while the j^resence 
 of warmer water to the far north is admitted, its impulse by 
 the Gulf Stream has been denied. That current has been 
 declared to cease at about 45° N. lat,, and the difiiision of 
 warmer water, even though its speed between Scandinavia 
 and Iceland has been calculated by Irminger at 1-| to 2^ 
 miles a day, has been attributed to the influence of the anti- 
 trades, or to the contraction of the polar seas by cold, and 
 the expansion of the equatorial by heat, a slope thus being 
 created down which the water flows. A good deal of dis- 
 cussion has arisen from the claim put forward for one or 
 otlier of these influences as the exclusive cause of motion, 
 and our imperfect knowledge of the laws. which regulate 
 the movements of great masses of water, has necessitated 
 the introduction into the controversy of a large amount of 
 hypothesis. It seems, however, that gravitation alone does 
 not supply the necessary force, any more than do the westerly 
 winds, whose steadiness, however, is likely to accelerate a 
 current which "holds its way" in a mass, having started 
 v/ith a high velocity. The Lightning and Porcupine Expe- 
 ditions showed that the IST. Atlantic has an upper stratum 
 of warmer water overlying a deeper colder portion, and 
 that this upper stratum has a tolerably uniform thickness 
 of 800 fathoms between the Bay of Biscay and the Hebrides. 
 But the warm water in certain localities reaches the bottom, 
 the colder water being entirely excluded, and thus there 
 may be in close proximity two columns of water of very 
 unlike temperatures at corresponding depths. The rela- 
 tions of temperature to depth in the warm and cold areas 
 thus discovered, are shown in the following columns, from 
 
128 PHYSICAL GEOGRAPHy. 
 
 which it appears that the temperature of 5°C. is in the one 
 case at less than 200 fathoms from the surface, in the other 
 at 567 fathoms deeper. 
 
 Lat. 61° 21' N. Lat. 69* 35' N. 
 
 Lon, 3" 44' W. Lou. 9° 11' W. 
 
 Temperature. Depth. Temperature. Depth. 
 
 10° 0. fathoms 11°C. fathoma. 
 
 5° 180 „ S" 350 „ 
 
 0° 320 „ 6" 600 „ 
 
 -r 500 „ 5'2' 767 „ 
 
 -1-2" 640 „ 
 
 The existence of this relation is more striking when the line 
 of soundings and temperatures in the Faroe Channel, ^^e., 
 between the Faroe Islands and the Shetlands, is tabulated. 
 
 Latitude, 60^4' 60° 4' 59° 56' 59° 48' 59° 40' 59° 34' 
 
 Depth (fathoms), 632 605 363 445 190 155 
 
 Surface, 11-1°C. 11'4° 11-4^ 12^ 12-r ll^" 
 
 Bottom, -0-8°C. -1-2" -0'3° -V 9%" 96" 
 
 The warmer water thus dams up the colder across the 
 southern end of the channel. But this colder portion has 
 7°C. lower temperature than water at the same depth in the 
 Atlantic, while, on the other hand, it is nearly three degrees 
 colder than the bottom water in the Bay of Biscay at 2435 
 fathoms. Further, the bottom temperatures off the Nor- 
 wegian coast are 6-5°0. in depths of 700 fjithoms. This 
 unequal thickness of the warmer water is not compatible 
 with a gravitation theory, according to which the cold water 
 would be found always beneath the warm; nor, if convection 
 (i.e., the distribution of heat by the transfer of masses of 
 warm water) took a prominent place, should we expect to find 
 the distinction so marked between adjacent columns of water, 
 a distinction comparable to that of the cold wall near the 
 starting point of the Gulf Stream. It appears, then, that 
 there are two opposite movements, that of the Gulf Stream 
 holding its course and tending to the north-east, and that of 
 Arctic current tending more slowly southward. But this is 
 a very slow movement, and the water lagging behind the 
 earth in its rotation acquires a south-westerly direction, pass- 
 ing towards the American side, though a small portion moves 
 towards Scotland, becoming mixed Avith the warmer water. 
 
kennel's current. 129 
 
 Tlie Stream enters a cnl-de-sac in the nortli Atlantic and is 
 dammed up; hence its comparative depth where, unless it 
 were driven north by some impulse from behind, it would 
 thin away; hence, too, partly because there is a reflux from 
 the point of obstruction, partly because of the eastward 
 tendency of the mass, the waters of the Scandinavian coast 
 have a higher temperature at all depths than is due to their 
 latitude. While the course of the Arctic indraught to the 
 south-west is clear, the source of the cold water in the Bay 
 of Biscay is more doubtful. The Arctic fauna is carried 
 down in the cold area into compara,tively low latitudes; but as 
 tliat area and the Bay of Biscay are sejDarated by water never 
 lower than 6°C., it is probable that the low bottom tempera- 
 ture of the latter represents a cold Antarctic- indraught, a 
 view supported by the continuously low temperatures already 
 noted as crossing the equator. While the indraught from 
 the north is usually very slew, it acquires some velocity in 
 two places, namely, in the deep channels along the east coasts 
 of Iceland, and Greenland. 
 
 88. Labrador Current. — The Arctic indraught is joined 
 by another, the Labrador stream, which forms the bulk of 
 the water that skirts the American shore, maintains its 
 lower temperature at the cold wall and enters the ]\Iexican 
 Gulf itself beneath the outilov/ing stream. On this last 
 point counter statements have been made; it is at all events 
 certain, that the cold water passes underneath the warm and 
 rises towards the surface on the eastern side of the stream, 
 so that the influence of the northern waters is felt in the 
 equatorial area. The velocity of tlie coast current is small, 
 its maximum is 1 5 miles a day, and south of Cape Hatteras it 
 is slower, its movement being sometimes reversed, probably 
 in consequence of tlie friction of the Gulf Stream tearing off 
 a part of the water, as the air column in the Odorator has 
 its top torn off by the horizontal current. 
 
 89. Rennel's Current. — The ^omewliat variable Kennel's 
 current is a drift which, coinciding with the eastward move- 
 ment of the Gulf Stream, has its force and extent modified 
 by tlie strength and continuance of the westerly winds. The 
 drift impinging on Cape Einisterre is turned northward, and 
 finally is lost in the ocean, whither it is diverted westwards 
 
 23 I 
 
130 PHYSICAL GEOGRAPHY. 
 
 from Ushant. The importance of this current lies in its 
 power to drive ships as far out of their reckoning as the 
 Scilly Isles, when seeking to make the English Channel. 
 
 90. Mediterranean Currents. — The Gibraltar current, 
 which at the Straits attains a velocity of even as much as SO 
 miles a day, resembles the Gulf Stream itself, and the stream 
 thi'ough Dover's Straits, in that it is a body of water whose 
 velocity gradually increases towards the a^oex of the funnel 
 formed by the convergent trends of the Spanish and African 
 coasts. Within the Mediterranean the velocity diminishes, 
 but a current may be detected along the south shores turning- 
 northwards along the Asiatic coast, the drift eastwards being- 
 maintained by westerly and north-westerly winds. The 
 existence of a steady under current of outflow through the 
 Straits is a matter of dispute. Dr. Carpenter affirming its 
 constancy, while others regard his observations as too few to 
 warrant that conclusion. It must further be added that the 
 Gibraltar current has itself been regarded as mainly due to 
 the conversion of the tidal movement into a wave of transla- 
 tion, thus exj)laining the frequent reversal of its direction, 
 and rendering unnecessary the complicated hypotheses as to 
 the replacement of water lost by evaporation. The existence 
 of an inward current at 20 fathoms depth, from the --^Egean 
 into the Black Sea, seems to be established by the Hydro- 
 graphical Survey, the surface current setting outwards at a 
 rate of about 25 miles daily. Carpenter assigns the under 
 current to the effort to restore equilibrium in the Black Sea, 
 by pouring in salt water of higher specific gravity to rejolace 
 Y\^hat flows out at the surface. Spi-att asserts that there is 
 no movement at 20 fathoms, and that the surface current is 
 reversed in winter when the rivers are low, the removal of 
 salt water thus taking place at one season, and being com- 
 pensated at another. 
 
 91. Baltic Currents. — A surface outflow and an inflow- 
 ing under current have been affirmed on theoretical grounds 
 to exist, and Meyer of Kiel gives evidence in favour of their 
 existence. There is a great influx of fresh water into the 
 Baltic from the adjacent land, and Carj)enter assigns to this 
 an elevation of the surface, and consequent flow-oflf, the under 
 current restoring the equilibrium between the waters on 
 
ANTARCTIC DRIFT. 131 
 
 either side of tlie strait. Both the explanation and the 
 existence of these Baltic movements have been doubted, and 
 in the meantime it can only be said that the agency, appealed 
 to by Carpenter, may operate along with the impulse of the 
 Gulf Stream as it flows back from the Arctic Sea in which its 
 progress has been arrested, but that it cannot yet be regarded 
 as sufficient to produce all the phenomena, even supposing 
 the movements to be accurately known. 
 
 92. N. African Current. — The N. African current is a 
 large "body of water which clings to the coast, and, as the 
 Guinea current, rejoins the equatorial current. 
 
 93. Sargasso Sea. — The drifts and streams now mentioned 
 leave in the centre of the ocean an area of comparatively 
 little movement, in which, as in a dead water, a vast amount 
 of Sargasso (Sargacjo) or gulf weed, accumulates and forms 
 a mass sometimes 60 fathoms deep. The locality and limits 
 of this sea are variable, its average position being between 
 20° and 30° N. lat., and between 30° and 60° Ion. W. 
 
 94. S. Atlantic Basin. — The Brazilian cj^irrent, already 
 mentioned, has an average speed of 15 miles, and, after 
 receiving the Plate river, merges in the connecting current, 
 which is a portion of the great eastward drift in the Antarctic 
 Ocean. The cold water from the south enters the S. Atlantic 
 Ocean more centrally than the corresponding supply to the 
 N. Atlantic; the summer isothermal line of (60°F.) 15-5°C., 
 reaching to IS*^ lat. S., while that of the northern basin does 
 not descend below 35° N. lat. In our ignorance of Antarctic 
 geography, it is impossible to say what may be the influences 
 which give to the easterly drift its powerful noi'thwards 
 tendency in the Atlantic. But the southers which blow at 
 Cape Horn, and the icebergs which travel northward, 
 suggest the existence of currents which help to carry the 
 eastward drift to the north. 
 
 95. Antarctic Drift. — The southern circumpolar sea may 
 be regarded, as has been mentioned (Art. 65), as a single 
 zonular ocean, of which the Antarctic drift is the principal 
 movement, and when we find that the current which sets 
 into the South Atlantic round Cape Horn is deiived exclu- 
 sively from this Antarctic drift, we may say that tlie Atlantic 
 basin derives no appreciable contribution from the Pacific 
 
132 rnysiCAL G::iOGi?APHY. 
 
 area; but tlie connection of the Atlantic with, the Indian 
 Ocean, and of the h\tter again with the Pacific, is more 
 direct and obvious, none of the continental lands projecting 
 into the Southern Ocean to the same extent as does the 
 southern extremity of South America. The arch which the 
 southern connecting current describes across the Atlantic, 
 extends as has been said to about 15^^ south latitude, and 
 returns towards the Cape of Good Hope, passing along the 
 coast round the Cape into the Indian Ocean. One portion 
 of it, however, turns to the north, and runs pa,rallel with, but 
 in the opposite direction to, the Guinea current, forming, in 
 fact, the first part of the equatorial drift, and having a tem- 
 perature nearly 7*^0. lower than that of the Guinea current. 
 The equatorial current has thus a northern and a southern 
 source of supply, which converge westwards, but leave to the 
 east a triangular space which is closed by the Guinea current. 
 In this space, which corresponds in position to the Doldrums, 
 a counter current sets eastwards with considerable velocity. 
 A space is again circumscribed by the south, connecting the 
 equatorial and the Brazil currents, which has neither the 
 extent, comparative fixity, nor sea weed of the Sargasso area, 
 of which it is the southern counterpart. 
 
 This drift, in the main eastward, though its direction 
 varies a few points at various places, has a steadier set the 
 farther north it is followed into more temperate regions. 
 The variations southwards are due possibly to irregularities 
 of the Antarctic coastline, whose form is, however, unknown; 
 and, without doubt, the large quantity of fresh water pouring 
 from the great ice cap during summer must have consider- 
 able influence. 
 
 93. Antarctic Drift: Pacific — This great eastward move- 
 ment, which has an average speed of 10 to 15 miles a day, is 
 affected by the two land pyramids, South America and Aus- 
 tralia. Near Cape Horn it divides, one portion entering the 
 South Atlantic Basin, the other, which may in strictness be 
 regarded as a backwater, running up the western side of the 
 continent. This mass has a very extensive surface, the 
 landward portion being knov/n as Humboldt's, or the 
 Peruvian Cold Current, the western portion as Mentor's. 
 The main body of water runs noi'thyrard^ and i^ turned 
 
AUSTRALIAN CURRENTS. 133 
 
 towards the west partly by the projection of the Peruvian 
 coast, partly by contact with the dead water which fills 
 the indented coast line of Central America. Thus the 
 southern source of the Pacific equatorial current is, as 
 in the Atlantic, derived from the colder waters of the 
 Antarctic regions. The northern source is derived from the 
 current which follows the coast of Korth America, and whose 
 westward deflection, aided by the north-east trades, is prim- 
 arily determined by contact with the tropical dead water. 
 Between the two roots a space is left in which an equatorial 
 counter current sets eastward, and, as in the Atlantic, is in 
 efiect a backwater formed by the two. 
 
 97. Pacific Equatorial Drift. — The equatorial current 
 illustrates the homomorphism of Eastern Asia and Eastern 
 America. To the north of Torres Straits it is deflected 
 northwards, and following the curve of the land to the out- 
 side of the Philippine Islands, runs up as the Japanese coast 
 current as far as Behring's Straits. This body of water is 
 the counter part of the Gulf Stream, being like it, warm, 
 blue, and salt ; but its velocity is less, since it is not acceler- 
 ated hy being confined within a basin like the Mexican Gulf. 
 Part of the stream passes through Behring's Straits, and, 
 while the major part continues eastwards, carrying warm 
 water along the north shores of America till it finally merges, 
 in Davis Straits, in the Arctic currents of the Atlantic, a 
 small portion turns to the west, and may maintain open 
 channels in the ice, but its influence must be small since, as 
 has been well pointed out, it afiects only feebly tlie freezing 
 of the water in Norton's Sound. The second branch of the 
 current crosses to the American shores, and, aided by the 
 north and north-west winds, follows the Californian coast to 
 complete the circle by joining the equatorial current. Of 
 the gTcat area which lies within this circle little is known ; 
 there is no positive information as to whether the Sargasso 
 Sea has an equivalent in it. 
 
 98. Australian Currents. — As tlie Australian continent is 
 a diminished representative of the South American, African, 
 and Indian land j^yramids, the currents on its shores corre- 
 spond to those of the African and American apices, especipJly 
 to tjie former, which likewise projects into the Antarctic 
 
134: I'HYSICAL GEOGRAPHY. 
 
 drift, and nas a land mass to tlie eastward. The drift, 
 rushing past the south shores of Australia with a velocity 
 increased by the steady westerly and south-westerly winds 
 of high latitudes, turns northward between Australia and 
 'New Zealand, and at times has a speed of 100 miles a day. 
 It runs in a trough between two southward currents, the 
 principal one being that which skirts the south-east coast of 
 Australia with a considerable velocity, the number and mag- 
 nitude of the return currents or backwaters along the coast 
 being in proportion to its sjDeed, friction having the greater 
 influence, according to the cricketing rule which Laughton 
 quotes very happily, " a slow ball must be played, it will not 
 play itself." The meaning of this iTile, which is well worth 
 bearing in mind, is that, whereas holding the bat still at the 
 proper angle will send a swift ball where it may be desired, 
 a slow ball striking on a steady bat will be stopped ; to send 
 it, therefore, in any direction, it must be struck. The side 
 currents in the Australian channel are portions of the equa- 
 toiial drift which have been deflected southwards by the 
 Polynesian Islands, the speed being increased by passing 
 through the narrow straits between them. 
 
 99. Currents of Indian Ocean. — The currents in the 
 Indian Ocean illustrate the important influence of dominant 
 winds on the movements of the ocean. The Monsoons of 
 the northern part of the ocean are reflected in the changing 
 direction of the currents on either side of the Indian Penin- 
 sula. The south-east trades maintain a westward movement 
 whose limits shift to north and south with the trades; and 
 which corresponds to the equatorial currents in other seas. 
 This current turns to the south at Madagascar, and reaches 
 the African coast as a rapid stream, a small part of which 
 enters the Atlantic, while the great bulk turns eastwards 
 with the Antarctic drift, and again runs northward along 
 the Australian coast, thus completing a circle and circum- 
 scribing a deadwater. In winter the waters of the Mozambique 
 Channel are partly drawn out in front, partly driven from 
 behind by the northerly wind, and join the southward stream 
 with a high velocity. But when the wind blows u]) the 
 channel to the north, and when at the same time the westerly 
 drift readies to the equator, and is turned southwards ^s it 
 
RED SEA. 135 
 
 approaclies Africa, tlie waters of the channel are drawn in 
 two directions, and the currents in it become uncertain. 
 Laughton suggests that the constant southward movement 
 should be called the Natal current, and that Mozambique 
 current should be restricted to the variable movements in the 
 channel which he has shown have been mistaken for a part 
 of the Natal stream. 
 
 100. Red Sea. — The waters of the Red Sea move southwards 
 during summer, and join the eastward current in the Arabian 
 Sea, the outflow being (from June to September) at the rate 
 of 40 miles a day. This outflow occurs at the period when, 
 if excess of heat alone determines movement, an inflow should 
 take place. It is during winter that an inward current sets 
 tlirough the Straits of Babel Mandeb, thougli probably an 
 indraught passes the Straits during most of the year to 
 replace the excessive loss by evaporation. 
 
CHAPTER IV. 
 
 THE FUNCTIONS OF WATEE. 
 
 The water wliicli is removed by evaporation from the ocean 
 is restored to it directly by precipitation from tlie clouds, 
 and indirectly by tlie streams wliich drain the land. But 
 the water surface thus distributed over the land is itself ari 
 area of evaporation, and even the ice and snow of polar 
 regions contribute to the atmospheric moisture. Thus the 
 rainfall on land is derived from the land as well as from the 
 sea; that on the sea is in turn a direct or an indirect restora- 
 tion of water that has been withdrawn from it. But whereas 
 the rain reaches the earth comparatively pure, or at least 
 brino'insc down matters which have been raised for a short 
 distance above the earth, the river enters the sea heavily 
 laden. It carries matters soluble and insoluble; chemically 
 dissolved and mechanically suspended; and these are distri- 
 buted by the movements of the ocean. The insoluble por- 
 tions are laid down under the influence of gravitation; the 
 soluble are so diffused that the waters of the ocean have an 
 average composition everywhere except close to land. It is 
 most important to the geographer to knoAv the power of 
 rivers as agents of waste — to learn what they actually do. 
 From exact data man can calculate his power to avert the 
 disasters wrought by an element as remorseless as fire; or, if 
 that is beyond his power, to mitigate the effects of the 
 calamity. He can guess approximately the time required to 
 effect certain changes on the earth, and thence foretell the 
 completion of other changes, whctlier these be the wearing 
 away of a continent to the level of the sea, or tlie final 
 closure of a river presently navigable. The annals of the 
 human race are not written in sand, but part of the record is 
 built in the alluvium of the Nile; and history repays the 
 
THE FUNCTIONS OF WATER. 137 
 
 oKligation by fixing with certainty tlie dates at which tlie 
 action of rivers closed the prosperity of a region by destroy- 
 ing its ports, and thus furnishing a standard by which to 
 measure the rate of geological change. 
 
 The detailed anatomy of rivers, above and below ground, 
 is essential to our appreciation of the history of nations. 
 Whether rivers are navigable or not; whether they open into 
 seas, of which the currents or the prevailing winds offered 
 obstacles not easily overcome by uncivilized r«»an; whether 
 the underground streams reach the surface as potable springs, 
 or remain below ground till man has acquired the know- 
 ledge, the skill, and the means to render them available ; on 
 that depends the early or late development of the inhabitants 
 of a country. These determine whether a country shall con- 
 tribute to human progress, or remain unprofitable till that 
 remote time when men, crowded out of the more productive 
 regions, shall be forced to enter into a struggle with unfavour- 
 able conditions. 
 
 SECTION I.— RIVERS. 
 
 Complexity of Rivers — Drainage Area or Watershed, Waterparting 
 — Quaquaversal Watershed, how formed — Rivers cutting 
 across Mountain Chains — Antiquity of British River V^alle^ys — • 
 Shifting of River Courses, and Return to the Original Channel — • 
 Permanent Changes of Course diie to Movements of Elevation — 
 Direction of Rivers altered by Disturbance of the Earth's Crust 
 — Sources of Rivers — Motions and Form of Stream — Floods: 
 Flood Waves — Speed of Rivers — Transporting Power of Water 
 — Sediments in Rivers — Drainage Systems: their Relations to 
 Oceans — Divisions of a River Course — Waterfalls, Rapids — • 
 Drainage Areas — Korth American Rivers : Southern Slope, Mis- 
 sissippi, Arkansas, Missouri, Red River, Ohio; Eastern Slope, 
 Potomac, Susquehanna, Hudson; Northern Slope, Great Fish 
 River, INIackenzie River — Mexican Gulf — South American Rivers 
 — Eastern Tributaries of N. Atlantic; Northern Slopes of Europe ; 
 Asiatic Continent — Rivers of West Coasts of Europe — Rivers 
 of British Islands — Rivers entering Baltic and Mediterranean—- 
 East and South Shores of Mediterranean: the Nile and its Tri- 
 butaries — West Coast of Africa : Senegal, Niger, Congo, Orange 
 River— Indian Ocean: Rivers of Eastern Africa; of Arabia and 
 Persia; Rivers of Himalayan Origin— Southern India: homomor- 
 
138 Physical geography. 
 
 pliic witli Africa — Table of Eelative Geological Importance of 
 Principal Elvers — Pacific Basin : Western Shores; Hoang-Ho, 
 Yang-tze-Kiang, Amoor, Song Ha, Camboja: Eastern Shores; 
 Colorado, South America : Australia: Murray River, Macquarric, 
 Darling : New Zealand — Eivers of Inland Drainage : Africa, 
 Central Asia, Aralo-Caspian Area. 
 
 101. A River is a Complex Water Channel. — The simplest 
 type of a river may be illustrated by a stream which flows 
 down the muddy banks of a tidal estuary as the tide ebbs. 
 The plane of marine denudation, and the straight course of 
 the primary channel down the gentle slope are well illustrated, 
 while as the water sinks the secondary channels become 
 conspicuous, until at last we have on a small scale a complex 
 drainage system formed by the combination of jJi'im^iy ^nd 
 secondary channels in the way already described (Art. 43). 
 
 A number of originally distinct river systems may combine 
 into one ; as, for example, in the North American continent, 
 where the Mississippi receives the drainage of higher grounds 
 to the east, north, and west, the drainage system correspond- 
 ing to that of at least three primarily distinct areas. The 
 most important rivers are those which flow parallel to the 
 longitudinal axis of a country j and these, as has been already 
 explained, occupy the valleys formed by the junction of the 
 secondary streams which cut back the ridges that at first 
 separated them, and thus came to form continuous grooves. 
 To follow the course of the Mississippi, the first part of the 
 Missouri consists of longitudinal and transverse streams, while 
 between the forty-third and thirty-ninth degrees of latitude 
 the longitudinal sti-eam is the principal. From about the 
 List mentioned point a transverse channel pours the waters of 
 tlie JMissouri into a common stream with the Mississippi. 
 Lower down, the Ohio, chiefly a longitudinal stream, enters 
 the same valley, while the Arkansas and Ked Rivers enter 
 directly as transverse channels into the lower part of the 
 Mississippi. The Mississippi, below the mouth of the Red 
 River, flows through a delta laid down in a deep estuary, 
 which runs into the mainland, and which, if prolonged, 
 would join the lake region, and help to isolate an eastern 
 from a western mass of land. If this valley were submerged, 
 the Mississippi system would be broken up into a large 
 
RIVERS CUTTING ACROSS MOUNTAIN CHAINS. 139 
 
 number of distinct water systems. But if ^tlie processes by 
 "whicli the Mississippi is at present building out its delta into 
 tlie Mexican Gulf be continued sufficiently long, we should 
 have the complexity of the river system increased, and the 
 same alternation of transverse and longitudinal streams would 
 recur. The Amazon offers an exactly similar picture. 
 
 102. Drainage Area or Watershed, Waterparting. — By 
 the term drainage area is meant all that surface of country 
 the slopes of w^hich direct the waters flowing from them 
 towards a common point; and the separation of drainage 
 areas is the waterparting, a term more strictly correct than 
 watershed, which is commonly employed. It would avoid 
 confusion were versant employed for the surface which inter- 
 venes between the waterparting and the watercourse. 
 
 103. Quaquaversal Watershed, how Formed. — It is not 
 necessary that the summit of the waterparting should be at 
 any great elevation; in fact we frequently find the boundary 
 between two contiguous drainage areas formed by a feature 
 of the ground so slight as to be a scarcely appreciable dis- 
 turbance of the level of the plain ; and from what has been 
 already said regarding the mode in which two adjacent 
 valleys open into each other by wearing down the higher 
 ground separating the head waters of two streams, it is 
 obvious that by the removal of the whole of such an inter- 
 vening ridge two valleys may be laid into each other, and 
 the last trace of a waterparting effaced. Examples of this 
 are found in several parts of the world ; and a notable instance 
 is met with in South America, where the Casiquiare connects 
 the Orinoco and Rio Negro. These quaquaversal Avatersheds, 
 the existence of which has been denied, but erroneously so, 
 are of frequent occurrence on a small scale in most countries. 
 One such is found in the centre of Scotland, where a small 
 plain in the Pentland chain forms a connection between two 
 streams, one of which enters the Tweed, the other the Clyde ; 
 and through this common feeding-ground salmon have passed 
 from the one side of the country to the other, their ascent by 
 the Clyde being a physical impossibility on account of the 
 Falls, while the transport of the ova by birds has an exceed- 
 ingly slender probability. 
 
 104. Rivers cutting across Mountain Chains.— One step 
 
l40 PHYSICAL GEOGRAPHY. 
 
 more in the lowering of a watershed may cany the whole of 
 a river into another channel, and it is to this extreme result 
 of atmospheric denudation that we owe the intersection of 
 mountain chains by streams that flow in one direction. Thus 
 in Derbyshire, and other parts of central England, streams 
 which take their origin in the centre of a hill range, after 
 travelling across the low grounds for some distance, pass 
 right tlirough another mass of high ground in their progress 
 toYv^ards the sea. There are only two possible explanations 
 of such an occurrence, the one being that the original plain 
 of marine denudation was traversed by a watercourse in the 
 direction of the existing stream : that later denudation carved 
 out a channel at right angles to it, and that this channel was 
 worn down so rapidly as to fall below the level of tlie primary 
 valley, whose waters were diverted into this new course. The 
 later valley helped to isolate a mass of the primitive plain, 
 and leave it as a hill of circumdenudation, the drainage of 
 which, falling into the primitive rivercourse, went in ojDposite 
 directions, and thus created a waterparting in the middle of 
 what had been a continuous channel, so that the relations of 
 the primary and secondary valleys became reversed. By the 
 slow lowering of this waterparting the continuity of the 
 valley might be restored, and the waters of the secondary 
 valley again carried down in the direction they had originally 
 followed. Another explanation is that the valleys which 
 isolate the hill masses are the primary channels, and that 
 those which intersect the hills are the secondary streams, 
 whose more rapid denudation has enabled them to divert the 
 waters of the primary river. 
 
 
 C3 
 
 
 
 •xi 
 
 
 
 a 
 
 
 
 o 
 
 % 
 
 
 o 
 
 g> 
 
 A 
 
 CO 
 
 C3 
 
 1 * T*! m fl VTT" _ 
 
 
 "-^ 
 
 X X XlllciX V "- 
 
 
 8 
 
 
 e3 
 
 
 'C 
 
 
 
 PI 
 
 
 
 O 
 
 
 
 o 
 
 
 
 <D 
 
 
 
 U2 
 
 
 Primary 
 
 Primary. 
 
 A river channel llomng in the dii'ection Primary — Primary 
 
ANTIQUITY OF BRITISH RIVER VALLEYS. 
 
 141 
 
 Sinuous waterparting : longitudinal val- 
 leys formed by wearing down of ridges be- 
 tween the tributaries of a a ; stream liow- 
 ins in direction h. 
 
 — Primary, and traversing a valley between tlie liigli grounds 
 A, in wliicli it originates, and B, through which it flows, 
 may have its direction thus fixed: Is^, Because both A and 
 B were parts of the 
 original marine denu- 
 dation from w^hich 
 the primary stream 
 flowed, and continued 
 to flow, while tribu- 
 taries to right and 
 left, marked secondary, 
 deepened their chan- 
 nels. 2nd, But A and 
 B may have furnished 
 tributaries to a primi- 
 tive stream in the posi- 
 tion marked secondary, 
 and by a lowering process such as formed the channel h (adjoin- 
 ing figure), the two channels on either side of the water- 
 parting B may have been laid into each other, and so carried 
 ofi" the whole waters in a dii'ection at ri^-ht ansjles to their 
 original direction. In these two cases Primary and Secondary 
 are reversed. A third case is when the secondary channels 
 have become one, as in the figures Arts. 43 and 104, and tho 
 Primary stream which traversed the original plain has its 
 upper waters deflected into this new channel : the lower course 
 Primary — Primary, now conveying less water, gradually, 
 forms two streams in opposite directions from a water[)arting 
 in the middle of their once continuous bed at B. The first 
 of these cases is that of the Wye in Derbyshire ; the second 
 is that of the Lyne w^ater, which rises in tlie Pentlands, 
 crosses the low grounds, and enters the Tweed through a gap 
 in the Silurians. The third is possibly the relation of some 
 of the N. American rivers, the change being, however, due in 
 part to glacial obstruction or earthquake movements. 
 
 105. Ant^'quity of British River Valleys. — The history of 
 the river valleys of the United Kingdom is comi)licated by 
 tlie numerous subsidences and re-elevations which have 
 occurred in that area; but the labours of Kamsay, Jukes, 
 Geikie, and others have proved that the river valleys of thisj 
 
142 PHYSICAL GEOGRAPHY. 
 
 country are of very great antiquity. The inspection of tlie 
 geological map of Scotland sliows that patches of old red 
 sandstone are found in the hollows of the Lauder, in the south 
 of Scotland, and far up the estuary of the Tay; that car- 
 boniferous strata are met with at Thornhill, in Dumfriesshire; 
 and that the permians lie in the heart of the hills in the 
 valleys of the Annan and the Nith. As these are all 
 sedimentary deposits, it is clear that at the time when they 
 were accumulated a valley existed and had generally the same 
 form as we now see. In fact, the plain of marine denudation 
 as it is seen in the silurian uplands of north and south Scot- 
 land, and the reasons which justify the belief that that plain 
 vras once covered by the old red sandstone formation, demon- 
 strate the extreme antiquity of Scotland as a mass of dry land, 
 and cc-isequently the extreme antiquity of the first valley 
 grooves which were excavated on its surface. The valleys, 
 therefore, Avhich must have existed before the deposit of the 
 old red sandstone strata, now exist, and many of them are 
 still the leading drainage channels of the country. The 
 history thus unravelled is essentially the same as that of the 
 valley formations in Norway and Sweden. But in Britain 
 the more frequent submergences, or at least the greater depth 
 of more recent depressions, somewhat obscure the details of 
 the original valley systems. Perhaps the most satisfactory 
 proof of the tendency of streams to return to their original 
 channels is furnished by the events subsequent to the glacial 
 period, at the close of which the country was covered by a layer 
 of glacial detritus upwards of 2,000 feet in thickness. But this 
 deposit conformed generally to the principal inequalities of 
 the ground, so that the leading depressions, such as the groove 
 of the Forth and Clyde, the Caledonian Canal, and the Great 
 Glen, murst have been indicated at all times. In the smaller 
 streams, however, as in the Clyde, the Tweed, and the Tay, 
 we find that while part of the course is through solid rock, 
 at other points it is cut in masses of glacial detritus. Tlie 
 river has, in fact, returned to its original line; it is still en- 
 gaged scouring out a chamiel which was choked up by the 
 glacial detritus, but having failed in some places to resume 
 its original direction from the compactness of the till, it has 
 found an easier passage through the stratified I'ocks (Art. 115). 
 
MOTIONS AND FORM OF STREAM. 143 
 
 106. Direction of Rivers Altered by Disturbances of the 
 Earth's Crust. — But while in North Britain this tendency 
 to reversion is thus manifest, in the south of England 
 we find proofs of an entire change in the dii-ections of the 
 rivers, besides that case mentioned in Art. 53. Thus 
 the Weald of Kent represents the delta of a great river, 
 which, flowing from the west, entered the German Ocean 
 probably in conjunction with some large continental stream, 
 perhaps the Rhine, the joint estuary being far to the north 
 of the present coast lines. These remains of former laud 
 surfaces date from the period of the chalk, the dry land at 
 that time forming a continental mass which stretched west- 
 wards towards the Atlantic, and was continuous with elevated 
 ground in the south of Europe. Thus the evidence of change 
 in the trend, as well as in the position of the former conti- 
 nental areas, is very complete. 
 
 107 Sources of Rivers. — The gathering grounds of rivers 
 are necessarily those portions of the country which are above 
 tiie level of the stream; or, to put it more plainly, water could 
 not scoop out a channel for itself unless it descended a slope, 
 and the excavating power of a stream is therefore proportional 
 to its velocity, that is, to the inclination of its bed. The 
 source of a stream is fan-shaped, all the rills converge to 
 the lowest point, and their junction will be nearer to or 
 farther from tlie waterparting according as the versant slopes 
 rapidly or gently. It is impossible to fix any one rill as the 
 source of a river; and this is especially the case where the 
 drainage of two or three distinct tracts unites in one great 
 stream, as the Mississippi and Amazon. Atmospheric mois- 
 ture is the ultimate source of all streams, whether it contri- 
 butes to them dh'ectly as rainfall, less directly by the melting 
 of ice and snow, or still more indirectly by means of the 
 springs wliicli return to the surfiice the water that has soaked 
 into the ground at higher levels, and has smik to various 
 depths. 
 
 108. Motions and Form of Stream. — In no case is the volume 
 of water in a stream equal throughout its length and at all 
 times. Seasonal difibrences are of regular recurrence, but the 
 variations may be very frequent in glacier regions, where the 
 cold during the night diminishes the streams Avhich the heat 
 
144 PHYSICAL GEOGHAPHY. 
 
 of the day again enlarges; or they may be at long intervals, as 
 in sub-tropical countries, where a drought may last for years 
 in place of a season. The size and volume of a stream are 
 determined by the rainfall. It may enlarge and diminish 
 gradually, or the enlargement and diminution may take place 
 suddenly, constituting what is known as a flood. Tlie surface 
 of a stream is not flat : the water moves more rapidly in the 
 centre than at those points where it is in contact with its channel. 
 The foam lines are seen to be curved, the convexity of the 
 curves pointing down stream ; and the curves are more or less 
 rapid in proportion to the speed of the river or the narrowness 
 . of its channel. While, therefore, friction against the banks 
 retards the marginal water, the bottom water is likewise held 
 back, so that a descending mass of water presents a curve 
 from side to side, and from the bottom of the stream to the 
 centre of the curve. A river, therefore, descends its slope ia 
 the form of a succession of shells of water. When its volume 
 increases, the difierence is first felt in the centre of the stream, 
 which rises more than friction allows the sides to rise, just as 
 the mercury in the barometer, when tending upv/ards, has its 
 surface convex, and friction gives to the river and to the 
 mercury a concave surface when the tendency is downwards. 
 Objects thrown into the centre of a rising stream travel 
 towards the margin; as the stream sinks they return towards 
 the centre. At the bend of a river floating bodies are carried 
 with the stream against the bank, but their speed increases 
 as they apjoroach it, and they are not thrown ofl" at an angle 
 (unless they are heavy), but rush down the curved surface of 
 the water, and rise to the surface lov»^er down and nearer the 
 centre of the stream. A backwater or reverse current alon<Tf 
 the sides of a river is found below an obstruction, such as a 
 projecting bit of the bank : the water in the lee of the obstacle 
 would be still, but the stream skirting it drags away its 
 margin, and thus, by friction, creates a movement, the water 
 moving up along the bank to replace that which is carried 
 away by the downward flow. If the projection is slight, there 
 is only one backwater. If the bay on its up side is wide, a 
 backwater is directly formed by a portion of the stream 
 sweeping into and round the bay. Thus the St. Roque current 
 strikes on the Mosc][uito goa.3t iii^ the Caribbean Sea, and forms 
 
SPEED OF RIVERS. 145 
 
 a backwater, wliile the peninsula of Yucatan gives rise to 
 another in the Bay of Vera Cruz. 
 
 109. Floods — Flood-Waves. — When a quantity of Avater 
 is suddenly poured into a stream it rises, but unless the supply 
 is kept up it immediately sinks. The suddenly formed eleva- 
 tion travels forward, but as it travels, the convexity gradually 
 flattens, till at last the excess of water is evenly distributed. 
 The form of the flood- wave is that of the wind-wave on the 
 sea — a long slope up stream, and a shorter slope down stream. 
 If it enters a lake, its progress is checked, partly because it 
 subsides by spreading, partly because its way is stopped by 
 impinging on the still water. The wave which begins at the 
 outflow of the lake is less violent, because it does not start 
 with any impetus. The same regulative influence is exerted 
 even more strongly when the stream passes through a swamp, 
 or pours part of its excess of water on the adjacent ground, 
 or escapes into sidings. These are sometimes natural; but 
 their artificial preparation is recommended where the 
 ground which it is pro^oosed to flood is unproductive, or of 
 less value than localities farther down, which require protec- 
 tion from the worst fury of the waters. The most mis- 
 chievous floods are those which occur where rivers flow from 
 temperate to arctic regions, like the Mackenzie Hiver of North 
 America, and those of Siberia, or where they descend from 
 the snow line to warm plains. In the former case, the flood 
 wave descends at the season when the lower waters are 
 frozen, and must therefore spread over the adjacent region; 
 in the latter, the rapid summer melting pours great bodies 
 of water suddenly on the low grounds. 
 
 110. Speed of Rivers. — The velocity of a stream is in 
 proportion to the inclination of its bed; but, under excep- 
 tional circumstances, as when it is checked by having to pass 
 through a narrow gorge, it is in proportion to the slope of its 
 surface, for there is thus created a constant and fixed flood- 
 wave. The average slope of a river bed is about 6 inches in 
 the mile, or 1 in 880, but in particular streams this estimate 
 may be too much or too little. Even in the same stream 
 the measurements vary. Thus the Missouri, from its source 
 to its junction with the Mississippi, measures 2,908 miles, 
 and it descends 6,800 feet; its average slope therefore is 28 
 
 23 K 
 
14G PHYSICAL GEOGUAPHY. 
 
 inclies per mile. In the first 264 miles it has a slope of IG 
 inches per mile; in the next 750 miles its slope is 10*56 
 inches per mile; for 648 miles its slope is 13-20 inches per 
 mile; for 404, 12*12; for 358 miles, 10*32; for 484 miles, 
 9*24. The first 1,330 miles of the Mississippi slope at 11*74 
 inches; between Ca,iro and the mouth, 1,088 miles, the incK- 
 nation is 6*94 inches per mile; but of this, the last portion, 
 855 miles in length, has a slope of only 4*82 inches per mile; 
 and the lower half of this portion gives the following : month 
 to 100 miles, 1*8 inch per mile; 100 to 200, 2 inches; 200 to 
 300, 2*30 inches; 300 to 400, 2*57 inches. The slope of the 
 Bhine is 1 in 414, but the lower 225 miles show only 1 in 
 2,324, the upper 375 descend at the rate of 1 in 609. The 
 velocity is, of course, least in the lower portions of a stream, 
 where the course is for the most part sinuous and the slope 
 of the gentlest, as in the Mississippi, where it is 1 in 1,100; 
 the Seine, 1 in 5,200. The middle course of the Khine 
 flows, it is estimated, at 3 miles an hour, the lower portion 
 of the Mississippi at 2-|- miles an hour when the river is at 
 mean height; but these calculations are somewhat uncertain 
 from want of agreement as to the method of observation, and 
 very unequal care in conducting the experiments. 
 
 111. Abrading and Transporting Power of Rivers. — The 
 quantity of material removed by a stream is usually in propor- 
 tion to its volume and speed, but not alwtiys. Thus, Livingstone 
 forded rivers in which there was more sand than water, and 
 the freshets of some mountain torrents carry down debris out 
 of all proportion even to the flood size of the stream ; but in the 
 latter case the rapid slope of the bed enables a small body of 
 water to set a very large mass of solid material in motion, 
 and the transporting power of the torrent is apt to be ex- 
 aggerated. Much depends on the form even more than on the 
 vreight of fragments — a rounded mass travelling more easily 
 than a flat one, a smooth fragment more easily than an 
 angular one. In comparing the transporting power of dif- 
 ferent streams this element is frequently overlooked. Sir 
 Charles Lyell gives many instances of the enormous power 
 of water in rapid motion, the flood of the Dee carrying blocks 
 of more than a hundredweight up slopes of 1 in 8 or 10. The 
 flood of Bagnes, occasioned by the giving way of an ice 
 
fiEDIMENTS IN EIVERS. 147 
 
 barrier, wliich liad converted part of the valley into a lake, 
 started with a speed of 22 miles an hour, which diminished 
 to 4 miles; a journey of 45 miles being accomplished by this 
 flood wave in 6 hours. The power of a river is usually 
 stated as follows: — Velocity 900 ft. per hour tears up fine 
 clay; 1,800 ft. ca-rries fine sand; 3,600 ft., fine gravel; 2 
 miles an hour, pebbles as large as a hen's egg. It is therefore in 
 proportion to the velocity and to the materials it hurries along 
 with it. Mr. Login believes that when a river has the proper 
 load of sediment for its velocity it looses abrading power. 
 
 112. Sediments in Rivers. — But these strildng cases are 
 imimportant geographically in comparison with what is going 
 on daily in every river, solid matter travelling seaward in 
 quantities which are not appreciated till they are reduced to 
 cubic dimensions and compared with familiar objects. Pro- 
 fessor Geikie and Mr. CroU have devoted much labour to 
 the investigation of the solid contents of streams, desiring to 
 calculate thence the thickness of the layer annually removed 
 from the drainage bed. Thus, Lyell calculates that the tota,l 
 quantity of suspended sediments in the Ganges, 6,368,077,440 
 cubic feet annually, could be removed by 2,000 Indiamen, each 
 of 1,400 tons, starting every day of the year. Humphreys 
 and Abbot calculated that the sediments carried annually by 
 the Mississippi into the Gulf of Mexico, 750,000,000 cubic 
 feet, would cover a square mile to the depth of 268 feet. 
 Now, the Ganges drains an area of 432,480 square miles, and 
 the sediments above mentioned represent a lowering of the 
 surface of that area by aaVTjths of a foot yearly, or 1 foot 
 in 2,358 years, the chemically dissolved matters not being 
 included in this estimate. The importance of this latter kind 
 of waste may be inferred from the fact that, whereas the 
 Clyde is estimated to remove a foot in 6,658 years from a 
 basin of 1,580 square miles, the Thames, which drains an 
 area three times larger, and discharges more than twice the 
 volume of water, v\^oukl require 10,144 years to do the same 
 work, if its actions were tested only by the solid matter it 
 carries in suspension. But the Thames drains a chalk area ; 
 the Clyde has comparatively little soluble strata to deal with ; 
 hence the work done by the latter is exaggerated in compari- 
 son with that done by the former, if wo omit the chemical 
 
148 PHYSICAL GEOGRAPHY. 
 
 action of the former. The details of this inquiry are to be 
 found in Croll's papers in the Philosopliical Magazine, and in 
 Geikie's Essay, which forms Chapter XXV. of Jukes' s Manual 
 of Geology. The general conclusions in that chapter are that, 
 calculating from the sediments carried into the sea by rivers, 
 the dry land loses a vertical foot in 6,000 years. But this 
 loss is unequal on the high and low grounds, the valleys losing 
 5 feet in that time, while the more level grounds lose -^ths 
 of a foot; or the table-lands lose -^^ih. of an inch in 75 years, 
 while the valleys part with the same amount in 8|- years. 
 Some uncertainty exists as to the averages relied on, some 
 districts losing more than others in consequence of the less 
 coherence of their materials. The sediments of rivers such as 
 the Mississippi are derived from areas under very unlike 
 climatal conditions, and an average of the whole does not 
 necessarily give the average of each. Changes in climatal 
 conditions may accelerate or suspend the rate of waste ; and 
 the majority of the rivers on which calculations are based 
 drain regions in which the hand of man has greatly modified 
 the processes of nature. But the method of research has 
 already brought out many unlooked-for results, and it will, 
 when observations are multiplied, bring us nearer to a reliable 
 standard for the calculation of geological time. 
 
 113. Drainage Systems: their Relations to Oceans. — It is 
 most convenient to reverse the natural method of the forma- 
 tion of a stream, and, commencing at the point at which a 
 river enters the sea, to classify all such streams into systems 
 which will thus give some idea of the amount contributed by 
 various masses of dry land to the adjacent oceans. The 
 North Atlantic basin receives the drainage of Europe and of 
 America, but to this must be added the vast quantities of 
 water which pour from the Siberian rivers into the Arctic 
 Ocean, and thence into the North Atlantic, foi'ming part of 
 the cold area in that basin. The contribution from the 
 extreme part of the North American continent is very much 
 larger than at first sight appears, since nearly the whole of 
 the Arctic waters enter the North Atlantic. The Gulf of 
 Mexico must also be regarded as a tributary of the North 
 Atlantic, and the waters of the Amazon and Orinoco are also 
 carried into the Caribbean Sea, and become a part of the 
 
DIVISIONS OP A RIVER COURSE. 149 
 
 Gulf Stream. Viewed in this manner, the whole West 
 Coast of Africa likewise contributes to the North Atlantic 
 waters; the River La Plata on the other hand passing south- 
 ward to join the southern connecting current. The principal 
 navigable rivers of the world are those associated with the 
 Atlantic basin, but large as their area may be, they never- 
 theless do not represent the whole of the drainage, since a 
 vastly greater quantity of water is contained in the small 
 streams which descend to the coast line without uniting into 
 large channels. 
 
 114. Divisions of a River Course.— Stream courses have 
 been divided into various sections, artificially characterized as 
 first, second, and third divisions, or upper, middle, and lower. 
 In the case of very large rivers the arrangement is sometimes 
 convenient, because in reality it answers to certain prominent 
 features of tlie dry land. The upper course of a large river 
 is for the most part highly inclined, and consists of many 
 small branches successively uniting. This is the condition of 
 all streams in recently emerged grounds, or of -those streams 
 which flow into arms of the sea that represent submerged 
 valleys. Thus, the Scandinavian rivers flow by tolerably 
 direct courses into the sea, the fiords of that coast line de- 
 monstrating the greater extension of the land at a former 
 period. Probably, were the Atlantic basin elevated, we 
 should find a middle and lower course for these streams — 
 courses which they formerly possessed, though we cannot 
 exactly tell at what period. What is known as the middle 
 course of a river is that part of its route in which the inclina- 
 tion of its bed is somewhat less, while the lower course of a 
 river corresponds practically to the alluvial plains. Thus, 
 the Ganges has the first part of its course broken up into 
 many converging streams; the middle part of its course is 
 comparatively short, and lies in the elevated terrace land 
 fringing the Himalayas, while its lowest portion flows through 
 the vast alluvial plains, which have been slowly increasing as 
 the land was gradually elevated. The history of that river 
 is in fact the history of a stream whose estuary ran far into 
 the continent. As the alluvial matters were laid down along 
 its banks, slow submergence maintained the level of the new 
 deposits very little above that of the stream, Evidence of 
 
150 PHYSICAL GEOGRAPH"^. 
 
 tliis mode of formation of an alluvial plain is founcl in tlie 
 soundings of the Indian Ocean, wMcIi show that opposite the 
 mouth of the Ganges there is a very deep depression, in 
 v/hich lies a " swatch of no ground," a deep submarine valley, 
 about 15 miles broad and more than 300 fathoms in depth. 
 We have in this hollow an index to the former depth of the 
 Ganges valley, the seaward growth of the alluvium of the 
 river being noAV checked by the current which travels along 
 the north part of the Bay of Bengal. 
 
 115. Waterfalls, Rapids. — In general terms it may be 
 said that the possibility of dividing a stream according to 
 variations in the slope of its bed furnishes an index of its 
 antiquity, though we shall find that the Mississippi furnishes 
 a notable exception to this rule. The continuity of a slope 
 is interrupted either by the entry of a river into a lake or 
 into level ground, or by its falling suddenly in cataracts 
 from higher to lower levels. The origin of many cataracts is 
 somewhat difficult of explanation, but in some of the falls of 
 the St. Louis Hiver of North America it has been ascertained 
 that the present position of the river is not that which it 
 occupied prior to the glacial period, the preglacial bed, choked 
 up with detritus, being still recognizable in the vicinity. 
 The river, which formerly flowed at a considerably lower 
 level, resumed its course in the same general direction, but 
 becoming deflected towards a ridge of rock, the summit of 
 a former escarpment, it was there retarded, dammed back 
 till it overtoj)ped the obstacle and poured across it, gradually 
 deepening the groove. The escarpment was doubtless the slo2:)e 
 of the hill looking towards the i)lain into which the St. Louis 
 flowed at its former lower level The general tendency in such 
 cases would be to lower the height of the fall, till at last a steep 
 sloping river bed should be provided. But in exceptional cases, 
 such as that of the Niagara, the vertical face of the cliff" is 
 maintained by a peculiarity in the method of disintegration of 
 the rock. A bed of hard limestone about 90 feet in thickness 
 forms the summit of the cliff, and rests upon a mass of soft 
 shales which are disintegrated by tlie spray from the base of 
 the fall, crumble readily away, and thus undermine the top 
 bed, letting it fall in blocks which, as is the case with most 
 very compact strata that have been subjected to modification, 
 
WATERFALLS, RAPIDS. 151 
 
 are of a rectangular form, lience the perpendicular face of tlio 
 cliff is maintained, and has been maintained during all the 
 time that has elapsed since the falls tumbled over above 
 Queenstov/n. But the "waters of Lake Ontario probably 
 flowed, through a channel now filled up, into the Hudson 
 Ixiver, and the selection of the route by Queenstown vms due 
 to the piling of rubbish on the south bank of the overflo"wing 
 lake, so that its waters were driven over a steep slope to the 
 north-east. On a small scale cataracts may arise by a process 
 exactly similar to that by which watersheds are lov^ered till 
 valleys become united. Thus, in hill regions where a catch- 
 ment basin occupies the summits, whence slopes descend with 
 different inclinations to the valley, the waters of the catch- 
 ment basin may be carried over the edge of the steeper face 
 by the wearing down of grooves on its margin, and thus a 
 waterfall may arise in a secondary fashion and be maintained 
 by the progressive eating back of its steep face ; or the same 
 result may be attained by the undermining of a cliff so that 
 at last a steep face comes to intercept or open into the bed . 
 of some mountain torrent which originally flowed in another 
 direction. The lowering of waterfalls, we repeat, is a normal 
 event ; the permanence of their precipitous face is only expli- 
 cable by special stratigraphical conditions, such as the nearly 
 horizontal hard bed with its crumbling luiderlying shale at 
 Niagara. A continuous slope may be converted into a water- 
 fall if, by a difference in the hardness of the strata, one part 
 v/ears away more rapidly than another, the abrupt face once 
 formed tending to pei^Detuate itself so long as a difference of 
 texture exists. A waterfall may be finally abolished by the 
 diminution of its area of supply, in the same way as we shall 
 find a glacier may be reduced. 
 
 The falls of Niagara, with a height of 154 feet on the 
 Canadian side, of 163 feet on the American side, and the 
 Victoria Falls on the Zambesi, with a vertical height of only 
 100 feet, are among the grandest displays of natural power 
 in the world, a sheet of water 2,400 feet in ^vidth being 
 seen at Niagara, while the Zambesi falls over in a mass 
 more than 3,000 feet wide. There are grander cataracts so 
 far as mere heio-ht is concerned. The StoAibbach descends 
 1,000 feet, and the Yohamite Falls in California have a single 
 
152 PHYSICAL GEOGRAPHY. 
 
 leap of 2,100 feet, the whole descent being 3,100 feet. The 
 Kaieteaiir Fall in Essequibo is 742 feet high, and even in 
 June was 370 feet wide, and discharged 500,000 cubic feet 
 per hour — that is, one-sixth part of the volume of the Clyde. 
 An attentive observer will find that the water falls in con- 
 centric shells like those of the flowing stream ; the marginal 
 waters do not fall at the same moment as the central, and 
 this relation is preserved from top to bottom even of a high 
 fall, if the water drops sheer. But vertical clifls are not so 
 common as is supposed ; they are usually stepped, and when 
 the gradations of the descent give a low angle to the whole, a 
 rapid is produced — that is to say, an incline of water with 
 a velocity which renders it impossible to ascend, and even 
 makes descent precarious, or forms a permanent obstacle to 
 navigation, compelling the traveller to carry his boat and 
 baggage overland by a " portage " to the calmer water above. 
 Another kind of rapid occurs where a permanent shingle bank 
 is developed, or where a constant supply of boulders in the 
 bed of the stream is maintained ; the slope of water being in 
 both cases swift, but in one smooth, in the other tumultuous. 
 Between the cascade and the ordinary slope of the river bed, 
 then, there are many intermediate steps which are intelligible 
 only by reference to the j^ast history of the river, or to the 
 character of the rocks of which its bed is composed. 
 
 116. Drainage Areas. — The North Atlantic basin receives 
 on the west the drainage of — a, Central America and part 
 of Brazil, the Gulf of Mexico being the interposed reservoir ; 
 h, the coast to the east of the AjDpalachians ; c, the great 
 lakes, through the St. Lawrence ; d, the land surrounding 
 Hudson's Bay ; e, Temperate and Arctic America, through 
 Davis' Straits. On the east the rivers of the Kussian Empire 
 reach the Atlantic by the North Cape; the western 
 European coasts drain directly into the Atlantic, while the 
 drainage of the interior is either direct — as in the case of the 
 Rhine, Seine, Tagus — or indirect, as where Sweden, Russia, 
 and Prussia jiour their waters into the Baltic in the north, 
 or where in the south the Mediterranean is the drainage 
 reservoir of an irregular area, the outlying points of which 
 are the Oural IMountains, the Spanish plateau, Equatorial 
 Africa, and the Caucasus. Lastly, the western shores of 
 
KORTH AMEniCAj EASTERN SLOPE. 153 
 
 j^frica contribute to the North Atlantic directly, or by the 
 equatorial drift through the Gulf of Mexico. 
 
 The South Atlantic receives the waters of South America 
 to the east of the Andes and to the south of Cape St. Roque. 
 
 The basin of the Indian Ocean includes part of the African 
 table-land, the borders of the Red Sea, of the Persian Gulf, 
 and the Arabian Sea, and the Bay of Bengal ; the outer limits 
 of the Asiatic area being the Taurus, the Himalayas, and the 
 axis of the Malayan peninsula; while with this basin must 
 be reckoned the western shores of Australia. 
 
 The Pacific basin is more simple in its outline, the high 
 grounds near the coast lines limiting the area of supply, ex- 
 cept w^here the Chinese rivers start from the Thibetan plateau. 
 
 117. North American Rivers — Southern Slope. — The 
 Mississippi drainage area covers 1,147,000 square miles. 
 Only a small part of the waters of this, the southern slope of 
 the North American continent, enters the Mexican Gulf 
 directly, the lower part of the Mississippi being the common 
 channel of the Mississippi, the Missouri, Arkansas, E,ed 
 Kiver, Ohio, and each of these in its turn having many large 
 tributaries. The commercial importance of this great system 
 of rivers may be gathered from the fact that the Missouri is 
 navigable for 2,570 miles above its junction with the Missis- 
 sippi, which is itself navigable for 1,500 miles, so that water 
 communication connects the interior of the continent for 
 4,000 miles with the ocean. 
 
 118. Eastern Slope. — The Potomac and Susquehanna arc 
 the chief among the many streams which drain the Alle- 
 ghanies to the east. The Hudson is now a stream of minor 
 consequence, but was probably in the line of the former 
 drainage of that area whose waters now pass away by the St. 
 Lawrence. This estuary, rather than river, now is the out- 
 flow for the great lakes from Lake Superior to Ontario. 
 Newberry has shown that the rocky surface of the watershed 
 or versant of the Mississippi is far below the level of that of 
 Michigan, and that at Bloomington, in Illinois, an old channel 
 has been sunk through to the depth of 230 feet: it is probable, 
 therefore, that the lakes Superior, INIichigan, Erie, and 
 Huron drained into the INIississippi, and that the overflow at 
 Niagara was due to the closure of this channel, while tho 
 
154 PHYSICAL GEOGRAPHY. 
 
 Lake Ontario, tlius largely increased, was, in its turn, driven 
 into the St. Lawrence by the closure of the Hudson valley. 
 
 119. Northern Slope. — The Great Fish River and the Mac- 
 kenzie are the chief streams in this direction, the latter 
 having a length of more than 2,000 miles, and a drainage 
 area of 441,000 square miles. This stream, whose floods 
 have already been mentioned, is the northern drain of a 
 region once continuous with the Gulf of Mexico. The valley 
 of the Mississippi, or more strictly that of the Missouri, was 
 under the waters of the later cretaceous sea, and the Missouri 
 only began to exist in tertiary times, when the cretaceous 
 rocks became uplifted as we now find them to the west of 
 that river. The comparatively slight separation of the south 
 and north slo]3es is due therefore to inequality of elevation, 
 while the length of the Mackenzie Kiver has been increased 
 by the great amount of detritus carried down by it from a 
 region covered with glacial detritus, just as the Siberian plains 
 have extended northwards. 
 
 120. Mexican Gulf. — It is unnecessary to speak of the 
 streams which enter this basin : none of them are very im- 
 portant, save as they illustrate the transverse type of river 
 channels, their course being ?vt right angles to the axis of the 
 respective districts. 
 
 121. South America. — ^The Magdalena, Orinoco, Amazon, 
 drain this continent to the west of the Andes, and as far 
 south as lat. 18°. The two latter rivers are in truth one, 
 since the wearing doAvn of the watershed has connected the 
 Ixio Negro and the Orinoco by a cross valley; and the same 
 process will, it is said, ultimately unite the Amazon and the 
 Plate River. In this region, therefore, the erosion of river 
 valleys is well illustrated, and Agassiz has observed that dis- 
 integration is especially raj)id since the heated surface of the 
 rocks is more readily disintegi'ated chemically and mechani- 
 cally by the abundant and warm atmospheric moisture 
 which the decay of a rank vegetation saturates with 
 carbonic acid. 
 
 The Amazon, 3,000 miles in length, and tidal for 500 miles, 
 is again, like the Mississippi, the common channel of streams 
 from very different parts of a region covering 2,000,000 
 square miles, whose flood times do not coincide, and thus 
 
EASTERN TRIBUTARIES OF THE NORTH ATLANTIC. 155 
 
 the river presents less ma,rkecl variations in volume than do 
 other streams, the dense vegetation having some share also 
 in this regulative influence. Having at its mouth a breadth 
 of 50 miles, it is still at 500 miles up the country one mile 
 broad; navigable for 2,000 miles, and receiving several large 
 and navigable rivers, it is well fitted to be the carrier of 
 a great trade, when the difficulties of climate have 
 been to some degree obviated by more extended clear- 
 ing of the forests, and the drainage which that operation 
 will entail. 
 
 The Orinoco, 1,560 miles in length, and tidal for 370 miles, 
 drains an area of nearly 400,000 square miles, and is navi- 
 gable for two-thirds of its length. Its waters, and those of 
 the Amazon, spread seawards with a velocity of about 3 miles 
 an hour, Vv^hich carries them visibly to a distance of, it is 
 said, nearly 300 miles, their muddy contents rendering them 
 easily recognisable. They are then swept into the Mexican 
 Gulf along with the St. Roque current. 
 
 The River Plate, with a drainage area of 900,000 square 
 miles, enters the Atlantic by an estuary of 180 miles in 
 length, the course of the river proper being between 1,700 
 and 1,800 miles. Its tributaries are also navigable, and arc 
 of considerable dimensions — the Salado being over 600 miles, 
 the Paraguay more than 1,000 miles in length. All these 
 South American rivers carry down large quantities of water, 
 the precipitation on the higher grounds being great; and 
 they are heavily laden with sediments, the soft materials of 
 the gTcat plains and terraces being easily disintegrated. 
 The waterparting about 12° S. lat. separates the N.E. 
 from the S.E. slopes, and expands eastwards in the Brazilian 
 Sierras, which consist of metamorphic rocks of palreozoic age. 
 Thus a mass of ancient strata establishes a coast Avatershed, 
 the streams of which are mostly small, the San Francisco 
 alone having any size; its leng-th is nearly 1,700 miles, and 
 its drainage area about 250,000 square miles. The Plate 
 River drains the eastern slopes of the Andes and the later 
 deposits of the Pampas, and thus forms a part of the area in 
 which the small streams draining the arid high grounds of 
 Patagonia are included. 
 . 123, Eastern Tributaries of the N. Atlantic— Rivers of 
 
156 PHYSICAL GEOGRAPHY. 
 
 N. Russia. — The rivers which drain the great plateaux on 
 either side of the Oural Mountains, from the North Cape to 
 Behring Straits, have an aspect very similar to the rivers of 
 Arctic America. The whole delta, bounded by a curved line 
 from the North Cape to 52° N. lat., 60° E. long., and thence 
 by Lake Baikal to Behring's Straits, forms to2:)ographically, 
 geologically, and by climate a single area, whose waters, 
 pouring northwards, flood the low grounds in winter to a 
 depth of from 30 to 50 feet. The Obi, the Yenisei, and the 
 Lena flow through 20° of latitude, and travel through so 
 much of low ground that the slope from the source of the 
 Lena to the sea is only 1 foot in 5 miles, obviously an ex- 
 aggeration due to the want of data for discriminating the 
 upper portion from the alluvial flats. The drifting of 
 Siberian timber westwards forms the chief evidence for 
 the Atlantic destination of these waters; but it is pi-obable 
 that some part travels eastwards to join the Pacific in- 
 draught. 
 
 123. West Coasts of Europe. — The Scandinavian peninsula 
 presents a simple series of transverse streams, of which only 
 the upper courses are above the sea level; the fiords repre- 
 senting their submerged lower valleys. The Elbe travels 
 550 miles from its Bohemian source, and drains G0,000 
 square miles of Central Euroj)e, bringing into the sea mineral 
 material suspended and dissolved to the amount of y^^^ ^th of 
 its volume. The Bhine, whose remote source is in the 
 Central Alps, has a course of 800 miles, and drains an area of 
 65,000 square miles. It carries through Holland a burden 
 of sediment amounting to y^^th of its volume (Hartsolker); 
 while at Bonn, Bischof found its solid contents to vary be- 
 tween ^gVfi^^ ^^y weight in flood, and -^yi-p ^th during a dry 
 season. The Rhine receiA'^es the Moselle and the Main, the 
 former passing through a district in which the soft loess 
 holds a prominent place. The Seine has a course of 400 
 miles, and from source to sea a slope of 1 in 1,628, the in- 
 clination being underrated, because its course below Paris is 
 very various. It is navigable for 300 miles. The Loire is 
 navigable for 460 miles of a course of 683 miles; but its 
 shallow, sandy bed renders navigation difiicult. The drain- 
 age area, 33,000 square miles in extent, is to a large extent 
 
BALTIC. 
 
 157 
 
 among the volcanic rocks of Central France. The Garonne 
 delivers to (j*^ ^7 volume of sediment into the Bay of Biscay, 
 gathered during a course of 360 miles, 260 of which are 
 navigable. The Adour is the last of the important tribu- 
 taries of the Bay of Biscay, all of which a^e navigable for 
 some distance inland, thus giving to France a commercial 
 prosperity which makes still more unintelligible to English- 
 m.en the saying that Paris means France. The Douro, Tagus, 
 Guadiana, and Guadalquiver drain Spain and Portugal to 
 the west. 
 
 124. British Islands. — No region shows more clearly than 
 do the British Islands the importance of the constant drain- 
 age at all points of the coast, as compared with that of 
 the rivers to which the attention of geogi-aphers is chiefly 
 directed. The subjoined table, borrowed from Professor 
 Geikie, will give a convenient standard of comparison for 
 other rivers, and a clear idea of the data on which the rate of 
 geological change is calculated: — 
 
 
 Lensfth. 
 
 Drainage Area. 
 Square Miles. 
 
 Annual Discharge of 
 Water in Millions 
 of Cubic Feet. 
 
 Annual Discharge 
 of Sediment. 
 Cubic Feet. 
 
 Thames, 
 
 Tay, 
 
 200 
 
 5,102 
 
 2,500 
 
 1,580 
 
 700 
 
 450 
 
 400 
 
 54,111 
 144,020 
 
 25,228 
 
 94,014 
 
 15,450 
 
 \ 
 I 
 
 1,805,003 
 49,000,000 
 
 8,099,000 
 30,022,000 
 
 5,328,000 
 1 lb. in 32 c. 
 f b. of water. 
 
 Clyde, 
 
 Boyne, 
 
 Forth, 
 
 Nith, 
 
 
 All the British rivers are tidal. The Thames, navigable for 
 110 miles above London, below which the estuary stretches 
 for 50 miles, receives tributaries, which are, in fact, a series 
 of canals for the conveyance of agi'icultural produce. The 
 short course of the British rivers and their comparatively 
 rapid fall, joined with their tidal character, foi'bid the forma- 
 tion of a delta, 
 
 , 125. Baltic. — The lake region of Sweden, which bears 
 comparison with America north of the latitude of Lake 
 Superior, pours its surplus waters into the Baltic, using that 
 
158 PHYSICAL GEOGRAPHY. 
 
 term for tlie whole body of water whicli enters the German 
 Ocean by the Skagerrack. The eastern side receives the 
 Neva, Dwina, and Niemen, in. addition to a number of 
 smaller streams; v/hile Prussia contributes the Vistula and 
 the Spree. Judging by the number of well known names, 
 the Baltic might be supposed to receive only a small amount 
 of fresh water; yet it is clearly established that, since its 
 communication with the ocean was reduced to what wo 
 now see, the shell fish, which yielded to the early in- 
 habitants of Denmark an abundant supply of food, are now 
 stunted in size in consequence of the brackish character of 
 the Avaters. 
 
 126. Mediterranean. — The Spanish coast is drained by a 
 number of streams, of which the Ebro is the best known. 
 The E-hone, which has a course of 500 miles, and drains an 
 area of 25,000 square miles, has a geological importance 
 which may be gathered from the fact that, though its course 
 is one-fifth of that of the Hoang-PIo, it removes, judging by 
 its sediment, one vertical foot from its drainage area of 
 25,000 square miles in 1,528 years, while the Chinese river 
 removes 1 foot in 1,4G4 years from an area 21 times larger. 
 On the other hand, the Po, with a course of 310 miles, and 
 a drainage area very little larger (30,000 square miles), does 
 the same work in 729 years. It is unnecessary to enumerate 
 all the streams which enter the Mediterranean on the north, 
 since all of them, with the exception of the Danube, the 
 Dnieper, and the Don, drain only the slopes close to the sea. 
 The Danube drains the great tract of Central Europe which 
 lies to the east of the sources of the Rhine, and between 
 the Carpathians and Balkans. The Dniester, Dniejoer, 
 and Don are the leading streams of a tract which is only, 
 so to speak, accidentally separated from the Casjoian, the 
 steppe region presenting a singular uniformity of aspect and 
 condition. 
 
 127. East and South Shores of Mediterranean. — If tlio 
 eastern boundary of the valley in which the longitudinal 
 rivers — the J ordan and Nahr el Avy (Orontes), flow respec- 
 tively southwards and northwards — were prolonged to the 40th 
 parallel of latitude, and carried thence eastwards to the 60th 
 parallel of longitude, the Ime thus trticed v/ould mark off all 
 
THE NILE. 159 
 
 tLe Asiatic drainage area wliicli contributes to tlie Atlantic. 
 The streams are numerous, but none of them are of great 
 size, and the greater number are transverse streams. It 
 follows from what has been said regarding the origin of 
 rivers, that such transverse streams indicate that they are 
 either of recent origin, excavated on emerging land,. or (as in 
 Scandinavia and Scotland) that the continuation of their 
 valleys — that is to say, the tracts in which longitudinal 
 valleys have been formed — are submerged. In either case the 
 transverse streams, corresponding to the so-called upper 
 courses of great rivers, are of small importance as regards the 
 commerce of the country. The in significance of all the North 
 African streams, except the Nile, is due partly to the small 
 size of the coast drainage area, partly to the dryness of the 
 region which, near the heated Sahara, and traversed by 
 southerly winds, has rainfall only in winter; and as this 
 is true also for the Syrian coast, the size of the streams there 
 depends on the elevation and breadth of the high grounds 
 where they take their rise. 
 
 128. The Nile, the problems of whose source and floods 
 have not yet been satisfactorily solved, has its most remote 
 sources to the south of the Equator, and then traverses 36 
 degrees of latitude, and may even have a longer course if 
 Lake Tanganyika should prove to be one of its feeders. Its 
 descent from the level of Victoria N'Yanza is not continuous, 
 being broken by a succession of cataracts, the first of which 
 are the INIurchison and the Ripon Falls, on the rivers issuing 
 from the Lakes Albert and Victoria N'Y'anza respectively, 
 while the lowest is 555 miles above Cairo, or about G55 
 miles from the Mediterranean. While the total descent is 
 over 4,000 feet, the inclination of the bed of the river 
 between the first cataract and Caii'o is G*5 inches per mile, 
 and belov/ Cairo is even less. The equatorial sources of the 
 river are still uncertain, but the Abyssinian branches, the 
 Atbara and the Blue Nile (Bahr-el-Azrek, long thought to 
 be the source of the river), are well known. The Nile is 
 alone among rivers in having no tributary for the last 
 1,500 miles of its course. The floods, therefore, which com- 
 mence in Lower Egypt about the summer solstice, and last 
 for 100 days — the river reaching its lowest state in April or 
 
160 PHYSICAL GEOGRAPHY. 
 
 May — must be due to the rainfall and melting of snow in 
 the equatorial regions. 
 
 129. West Coast of Africa. — The same remark holds true 
 for the West Coast of Africa, except in that region which 
 corresponds to the latitudinal band of mountains over which 
 the west wind blows from one side of the continent to the 
 other. The great rivers which pour into the Atlantic, the 
 Senegal and the Niger, or Joliba, are derived from the great 
 mountain masses; and the Zaire, or Congo, which drains the 
 central basin, may yet prove also to derive a part of its 
 water from the same lakes and mountains to which the Nile 
 is traced. In all African explorations a chief difficulty lies 
 in the seasonal changes of rivers and lakes, the wet season 
 filling a channel which had before been dry, and changing 
 first into swamp, and then into a lake, what had been an arid 
 desert. The continued residence necessary to recognize these 
 vicissitudes has not yet been possible for the European 
 traveller; but it is well to bear in mind that conflicting 
 reports regarding any one locality are sometimes reconcilable 
 by reference to this fact, as it has been proved to be in tho 
 case of Lake Tchad. The Orange River of South Africa 
 drains the whole breadth of the land from the eastern edge 
 of the plateau to the Atlantic. Its history is not well 
 known; but it contrasts by the length of its course, the con- 
 tinuity of which is interi-upted during dry seasons, -svith the 
 streams which drain the southern extremity of the continent, 
 among which the transverse and longitudinal valleys are 
 beautifully exposed. 
 
 130. Indian Ocean : East Coast of Africa.— The Zambesi 
 is the princijoal stream of this side of the continent ; it is the 
 counterpart of the Orange River as the drain of the central 
 plateau, but its importance, geologically speaking, is less than 
 that of the Lufiji, which reaches the coast after traversing a 
 range of mountains two months* journey in breadth. It is 
 not known whether this remarkable course has been formed 
 by the fusion of two streams flowing in oi:)posite directions on 
 either side of the lip of the central basin, or represents a 
 primitive stream whose route was marked out during the 
 emergence of the land. The Zambesi is a striking example 
 of river erosion pure and simple, unaided by atmospheric 
 
INDIAN ocean: east COAST OF AFRICA. 161 
 
 moistiu'e during large part of its course. The walls of the 
 ravine below the falls described by Livingstone are vertical, 
 like those of Colorado Canons, and for the same reason that 
 the dry atmosphere has no power of wasting the cliffs, and 
 giving them that sloping form characteristic of the valley 
 walls of temperate regions (Art. 45). The size and permanence 
 of the stream testifies to the abundant water supply at its 
 source, since evaporation during its course does not affect 
 its continuity. Abyssinia is drained chiefly towards the 
 Nile, the streams which descend the eastern slopes being 
 of the same character as that of the African coast lands 
 generally. 
 
 The Arabian shores are not sufficiently known to furnish 
 any information regarding the probable annual quantity of 
 water they shed seaward. The Abyssinian and Arabian 
 coasts are identical in type, the characteristic table-lands and 
 valleys of the African being repeated in miniature on the 
 Asiatic side. But the streams in the nidlahs are uncertain, 
 and the water is for the most part below the surface. The 
 Tigris and Euphrates, which enter the Persian Gulf, drain a 
 very extensive area southwards from the Armenian mountain 
 ^ chain which fronts the Black Sea, and bounded west and east 
 by the valley of the Jordan, and the most westerly belt of 
 Persia. No district shows more clearly that the area drained 
 by a river can be spoken of as its basin only with some 
 restriction. Tlie alternately transverse and longitudinal 
 streams which unite in the Euphrates indicate that several 
 
 sins, or the convergent slopes which shed their waters 
 ' -ito one channel, have been opened into each other. If 
 the north and south streams are the primitive courses formed 
 as the land was raised, then the basins of the east and west 
 tributaries are secondary valleys, and as such, portions of the 
 one great Euphrates basin ; if, on the other hand, the chains 
 of the Taurus and Anti-Taurus formed the waterpartings 
 which constituted the land they bound a water basin, then the 
 Euphrates, which cuts through the southern barrier, drains 
 at least two basins. This is certainly the case with the 
 Indus, the Ganges, and the Brahmapootra, all of which drain 
 valleys in the Himalayan range, or belonging to tlie Thibetan 
 plateau, and break through the southern rampart of the 
 23 L 
 
162 PHYSia\L GEOGRAPHY. 
 
 Himalayas to reach the low country. The upper and lower 
 courses of all these rivers contrast markedly. The upper 
 descend rapidly from the high grounds of their sources, while 
 in the plain they have lower slopes and a gentler flow. 
 The drainage areas of these rivers are together nearly equal 
 to that of the Mississippi. 
 
 The Irrawady and the Saluen, which fall into the Gulf of 
 Martaban, are the principal streams on the east side of the 
 Bay of Bengal. They have not been traced up to their 
 sources, but it is probable that they too descend from the 
 eastern portion of the Thibetan plateau. 
 
 131. Southern India: Homomorphic with Africa. — The 
 peninsula of Southern India resembles Africa south of the 
 Sahara, in consisting of an elevated table-land with steep 
 seaward faces, the edges of the plateau rising into hill 
 ranges. The plateau averages from 2,000 to 4,000 feet of 
 elevation, the greater height being in the south. The 
 Eastern Ghauts are lov/er than the Western, which reach to 
 5,000 or 6,000 feet on the Malabar coast, while the Nilgherry j 
 Hills even attain to 9,000 feet. The form of the mass is that of 
 a pyramid whose base fronts the Himalayas, and is separated 
 from them by the valley of the Ganges, into v/hich part of 
 the drainage flows, and which was once the sea channel that 
 isolated Southern India. But the smaller dimensions of the 
 ■pyramid have given to the periodic rainfalls a power which 
 has resulted in the formation of streams draining the whole 
 area, and flovang in valleys of denudation often of considerable 
 dimensions. The narrowness of the coast la.nd gives little 
 importance to these streams commercially. 
 
 132. Table of Relative Geological Importance of Princi- 
 pal Rivers. — The following table has been constructed with 
 a view to show the relative importance, geologically, of the 
 principal rivers which have been made the basis of investiga- 
 tion by those who seek for a standard of geological time. 
 The known proportions of sediment are assumed to be 
 those which the rivers would ofier if they all had the same 
 drainage area as the Mississi23pi. Of course the student 
 will bear in mind that enlarging the scale of natural 
 phenomena not unfrequently means altogether changing 
 their character. He will, therefore, look on this table as 
 
PACIFIC basin: eastern ASIA. 
 
 163 
 
 sliowing the relative, not the absohite, proportions for tlie 
 respective districts : — 
 
 Kame of River. 
 
 Equal Drainage Area 
 in square miles. 
 
 Equal Annual Discharge 
 
 of Sediment in 
 
 1,000,000 cubic feet. 
 
 Mississippi, 
 
 1,147,000 
 
 7,459 
 15,920 
 28,908 
 27,600 
 
 6,265 
 r)7.3SO 
 
 2,867 
 
 22,491 
 
 11,980 
 
 .12,740 
 
 5,808 
 26,268 
 
 Gano-es, 
 
 Hoanf-Ho, 
 
 Khone, 
 
 Danube, 
 
 Po, 
 
 Xith, 
 
 Tay, 
 
 Thames, 
 
 Portli, 
 
 Clyde, 
 
 Eo^'iie, 
 
 133. Pacific Basin: Eastern Asia. — The Philippine Archi- 
 pelago, lil?:e that ot the West Indies, lies to the east of the 
 barrier — here, hoAvever, incomplete — which separates t^vo 
 ocean areas. The drainage of the islands, insignificant for tlio 
 most part, pours into one or other ocean in varying quan- 
 tities. Thus, v/hile rain falls in Java well nisjli throuahout 
 the year, and is copious to the east of Timor, the area 
 between Timor and Java, including the eastern part of Kent 
 Island, is dry, the winds blowing over Australia acting 
 towards them as the Sahara to the north coast of Africa. 
 
 The eastern drainage area of the continent widens steadily 
 from the extreme north-east till the sources of the Hoan2:-ITo 
 are distant from the coast line 25° of longitude, while the 
 axis of the Malayan peninsula forms its boundary in the south. 
 As the area widens, tlie low grounds increase in extent, and are 
 traversed by numberless streams, which give the region great 
 fertility, and are themselves of commercial importance, from 
 the ease with which they can be opened into each other, the 
 canal system of the Chinese Empire enabling every inch of 
 that vast area to be productive. The Hoang-Ho, tlic Yang- 
 tze-Kiang, and the Amoor, are the three grand streams of 
 tills area. The principal facts concerning the Yellow liivcr 
 
164 PHYSICAL GEOGRAPHY. 
 
 have been already stated ; the Yang-tze-Kiang, or Blue River, 
 has a course of 3,600 miles, and is connected with the Hoang- 
 Ho by canals ; the Amoor has a very tortuous course of 2,600 
 miles from its source S.E. of Lake Baikal, and is navigable 
 by steamers for 2,200 miles. The floods of the Yellow River 
 show some peculiarities, of which no adequate explanation 
 has been given, the possibility being that while the excess 
 of water is in some places spread over swamps, the junction 
 of other streams, whose history is not known, supplies at the 
 flood season an equivalent for the v/ater thus lost from the 
 main channel. The interior of China is accessible for 900 
 miles by vessels — for the upper 300 miles only by those cf 
 small size; but far beyond that limit water transport is still 
 possible. The Song-ha which enters the Gulf of Tonquin, 
 and the Camboja, which enters the Gulf of Siam, have their 
 head waters connected by a canal. 
 
 134. Eastern Shores of the Pacific. — The narrow strip of 
 land which skirts the base of the longitudinal mountain chain 
 is intersected by many streams, the waters of which are more 
 copious than their length would lead us to expect. But the 
 catchment basins in the mountains are many and deep; and 
 the Pacific watershed is enlarged by the transverse valleys 
 vrhich drain the gTeat longitudinal valleys seaward. The 
 abundance of the water supply at the sources is testified, as 
 in Africa by the Zambesi, in North America by the Colorado, 
 which flows as a large stream through a more arid region, and 
 enters the Calif ornian Gulf, which is, in truth, the submerged 
 southern end of a longitudinal valley. The streams of 
 Southern America are short, rapid torrents. 
 
 135. Australia. — This fragment of a continent, homomor- 
 phic with Africa and Southern India, still resembles the 
 former in the fewness of its rivers to the north, and their 
 greater number to the south of the central dry plateau ; the 
 northern streams are, further, like those of Africa in their 
 variable character. The Mui-ray River, the largest of the 
 southern streams, receives the Macquarrie and Darling Rivers, 
 and thus drains the inland versant of the Blue Mountains 
 towards the west and south-west in such a way as to suggesb 
 a river that once flowed into the central basin of a continent 
 Ayhich is now submerged. 
 
BIVEES OP INLAND DRAINAGE. 165 
 
 TLe streams of New Zealand are simple; but those of tlie 
 South Island, whose sources are in the glacier grounds, are 
 slowly extending the fertile low grounds eastward; while the 
 dangers attending their floods promise to advance our know- 
 ledge of river phenomena by the engineering precautions 
 which the increasing value of the territory requires. 
 
 138. Continental Rivers, or Rivers of Inland Drainage. 
 — Hitherto we have spoken only of the rivers which enter 
 the sea; but a large quantity of the drainage of the land is 
 directed to the interior, forming lakes whose indefinite 
 gTOwth is arrested by evaporation ; nay, even their formation 
 is in some cases prevented, and the streams are dissipated in 
 the atmosphere without accumulating. These areas of in- 
 ternal drainage are the interior of Africa, north and south of 
 the equator ; the centre of Asia, north of the Himalayas ; the 
 Aralo-Caspian area on the confines of Europe and Asia, 
 though it would be perhaps more simple to regard as one the 
 great tract from the Black Sea to the confines of China. 
 But the area is divisible into two, Aralo-Caspian area noting 
 a tract only recently separated from the Mediterranean, while 
 the rest of Central Asia consists of plateaux, Avhich rise to 
 1 7,000 feet above the sea level. The Volga and Oural are the 
 largest rivers which enter the Caspian from the north; the 
 former having a course of more than 2,400 miles, and drain- 
 ing an area of 550,000 square miles on the west of the Oural 
 chain, while the Oural has a course of half that length in the 
 east. The Aras (Araxes) drains large part of the Caucasus, 
 while the southern shores receive the torrents from the high 
 lands which bound the Iranian desert to the north. On the 
 east the streams are few, and confined to the extreme north 
 and south ends of the lake, unless the overflow of the Oxus 
 reaches the Caspian about 39° lat. N. The northern and 
 southern extremities of the Aral Lake are the only points at 
 which streams enter — the Amu Daria (Oxus) being at the 
 southern, the Syr Daria at the northern end. 
 
 It is impossible in the present state of our knowledge to 
 give any details regarding the number and size of the streams 
 which lose themselves in the deserts, such as the Salt Plains 
 of Iran, in which the river disappears that passes Ispahan, 
 Their history is a subject of great interest, since the chanuell 
 
.166 PHYSICAL GEOGRAPHY. 
 
 in wliich water is always present on tlie margins of tliese 
 Y/astes exliibit an amount of erosion only explicable on the 
 supposition that they are the valleys fashioned under other 
 climatal and geographical conditions, the date of which may 
 vat be. ascertained. 
 
 BECTION II.— WATER IN THE INTERIOR 05' 
 THE EARTH. 
 
 Lakes : their Classificfttion — Lakes Formed by Irregular Surface Ac- 
 cumulations : by Overflow : by Obstructions : by Glacier Erosion : 
 by Subterranean Erosion : in Areas of Subsidence : by Elevation 
 -^Continental Lakes — Analysis of Lake Waters — ^Ancient Lakes 
 and Lacustrine Areas — Filling up of Lakes. 
 
 137. Lakes: their Classiiication. — It is impossible to 
 separate the rivers of inland drainage from lakes, smce the 
 conditions v^^hich determine the former are the same as those 
 which allow of the formation of the latter. 
 
 A lake is a body of fresh or salt water which has no outlet, 
 or whose su])eriicial area greatly exceeds that of its outilovr. 
 In these, as in all other natural phenomena, it is exceedingly 
 difficult to draw a sharp line of demarcation, and it may be 
 asked if the Sea of Azof or the Baltic are to be considered 
 as lakes. The only possible ansvv-er is that they hover on the 
 border between sea and lake, and tliat the existence of tidal 
 movements may be held to constitute a real difference. 
 Popular speech has acknowledged the difficulty; for the fiords 
 of the west of Scotland are called lochs, and the distinction 
 between gulf and lake is evanescent v.dien we find that Loch 
 Fyne has a shallow but broad mouth, that the Garelcch has 
 a shallow and narrow outlet, and that the shallow and nar- 
 Yow outlet of Loch Lomond is some miles from the cstuaiy, 
 a difference of elevation giving to each of these three their 
 distinctive characters. 
 
 Lakes may be classified as follows: — 
 
 1. Lakes which occupy hollows in supei'ficial accumula- 
 tions — 
 
 L Lakes formed by irregularity of surface accninulation, as in tho 
 
LAKES FORMED BY OBSTRUCTIONS. 167 
 
 hollows formed by confluent mounds of sand and gravel; by 
 contiguous morainic mounds; or by sand or shingle bars on 
 the coast. 
 Q, Lakes formed by obstruction, which may be (a) temporary when 
 a glacier blocks the mouth of a lateral valley; or (6) of greater 
 duration when alluvial gravel or a moraine chokes up a 
 valley, or when a lava stream crosses a river bed ; (c) lakes of 
 overflow. 
 
 II. Lakes wliich lie in rock basins — 
 
 3. Lakes formed, or at least deepened, by glacier erosion. 
 
 4. Lakes formed by subsidence : (a) when soluble strata have been 
 
 washed away, letting down the surface; (6) when subter- 
 ranean movements, as earthquakes, allow the surface to 
 subside. 
 
 5. Lakes formed by elevation, as may happen when the mouth of a 
 
 valley in a disturbed district is closed by upheaval. 
 
 6. Lakes formed on plateaux, as in equatorial Africa, Asia, and 
 
 Australia. 
 
 This classification cannot be regarded as perfect, since each 
 of these groups may merge into another, the last three being 
 perhaps reducible binder one. 
 
 138. Lakes formed by Irregularity of Surface Accumu- 
 lations. — The sand and gravel mounds known as eskers, 
 osar, or kames (kaims), frequently unite so as to surround 
 cups or grooves in which water may lodge, forming small tarns, 
 as among the great gravel mounds of Carstairs, in Lanark- 
 shire. This also happens among the close-set morainic mounds 
 of glacier valleys. Lagoons are formed by storm accumula- 
 tions of sand or gravel parallel to the coast, or by the irregu- 
 lar deposit of the materials of a delta, as in that of the Hhone, 
 the etangs being due to both of these processes. Lakes of 
 this class are few and insignificant. 
 
 139. Lakes Formed by Obstructions. — The parallel roads 
 of Glen Roy, which are the successive beach lines of a sinking 
 lake, illustrate the effects of a temporary obstruction, such as 
 is seen at the present day in the Marjelen See, the barrier of 
 •\vhich is a glacier moving across the mouth of the valley. 
 Such a barrier is temporary in this sense, that it depends on 
 climate, an increase of annual average temperature causing 
 its disappearance, whereas detrital barriers are not so re- 
 movable. Lakes in the course of rivers ai-e often due to 
 inequality in the texture of the strata through, wliich the 
 
1G8 PHYSICAL GEOGRAPHY. 
 
 cliannel is worn, in consequence of wliicli only a narrow 
 groove is cut with difficulty, while above, the softer strata 
 have allowed a wide valley to be formed. The waters held 
 back above the barrier widen the space in which they are 
 contained. But such lakes are obviously of recent origin, as 
 they are of necessity temporary, since the river sediments, 
 speedily deposited in the stiller water, tend to fill them up 
 and convert them into alluvial plains. Some of these lakes 
 are of considerable importance, and should perhaps be referred 
 to the third group, since their features have been, in many 
 cases, exaggerated in temperate regions by the travelling of 
 glaciers down their hollows. Shallower lakes arise by the 
 deposit across a valley of a mass of detritus by a tributary 
 stream; and the same result follows when a glacier pushes its 
 moraine so as to occlude the channel of an adjacent stream. 
 But the grandest examples of such obstruction are to be 
 seen in America, where the pre-glacial river channels have 
 been filled up with boulder clay, the moraine profonde of the 
 continental ice-sheet; and, on the disappearance of the ice, 
 the rivers resumed their previous courses. As has been 
 already stated, the great size of Lakes Superior, Michigan, 
 Huron, and Erie, is probably due to this cause, the oblitera- 
 tion of their southward channels ending in their overflow into 
 the St. Lawrence valley. Scrope describes a rare case of 
 the Lake of Aidat, in Auvergne, formed above a lava stream 
 which crossed the bed of the river. 
 
 To this group ought to be referred lakes of overflow, where 
 the bed is too small for the flood waters which pour over on 
 the adjacent lower ground. If from any cause the flats into 
 which they pour are limited, a lake may result which will be 
 alternately filled and emptied as the river rises and falls, till 
 at last the river banks may be raised so high that a permanent 
 sheet of water is formed, and this event may be secured by 
 the obstruction of the channel of overflow, as in Louisiana, 
 Tv^here rafts of wood sometimes stick at the mouth of the 
 breaches in the river bank. 
 
 140. Lakes of Glacier Erosion. — Tlie temperate regions, 
 America north of 45° N. lat., Sweden, Finland, Russia, 
 Scotland, and Cumberland and Wales in England, at tlio 
 Antipodes (New Zealand), abound in lakes, whose limita- 
 
LAKES OP GLACIER EROSION. 169 
 
 tion to tLe leading lines of tlie valleys of the country 
 is singularly close. Professor Ramsay has proved that to 
 the action of ice a very large number at least of these lakes 
 is directly due. The evidence in support of this view is, 
 first, that the lakes occur in the line of extinct glaciers; 
 second, that the deepest part of their basin is below the level 
 of the channel beyond the outlet; third, that at least in many 
 cases the outlet is through or over a barrier of rock, whose slope 
 down towards the deepest part of the lake is more rapid than 
 that of the upper portion of the lake floor. It is evident that 
 water, even running water, has no power to excavate its channel 
 deeper at one point than at another, more especially when 
 there is not merely a considerable difference of depth at 
 various points of the lake bottom, but, in a large number of 
 cases, the deepest part is below the level of the sea. A lake, 
 in fact, cannot make its own basin; it cannot deepen its own 
 basin in the same way that the Lake of Geneva or Loch 
 Lomond is deepened; its only power is to, widen its valley 
 by wearing away the banks at its margin. It has been 
 suggested that many lakes are in the position of former frac- 
 tures or dislocations of strata; but in very few cases, if in 
 any, is there discord between the strata on opposite sides of 
 the valley. There has been therefore no relative displace- 
 ment of the mass, such as constitutes a fault; and, even if 
 such difference exists in some cases, it would be necessary to 
 prove that faults as a rule coincide with the line of valleys. 
 But, though instances of such coincidence may be found, 
 faults, as a general rule, traverse a country very irregularly, 
 their dates of origin being anterior to, or subsequent to, the 
 formation of valleys, and their lines without the slightest 
 relation to these results of denudation. This question has 
 already been discussed in treating of the origin of valleys; 
 its special application to the case of lakes is required, because 
 it has even been suggested that lakes have been in some cases 
 determined by the fissure of a fiiult. But such disruptions of 
 solid rock do not occur as open fissures at the surface of the 
 ground, the mode of their formation (Art. 2G) rather making it 
 obvious that the close approximation of the fractured surfaces 
 is the rule; and, though it is quite true that such a fracture 
 line would be a line of weakness, and therefore sufficient to 
 
170 i?IIYSICAL GEOGRAPHT?. 
 
 determine the course of the stream along it, if the fracture 
 was parallel to the slope of a plain of denudation, it must 
 be remembered that such coincidence cannot be expected 
 to have. been very frequent, and, even when it did occur, 
 it would rarely happen that the line of the fault and the 
 course of the primary streamlets coincided. 
 
 If we take as an example the large lakes or lochs of the 
 West of Scotland, we shall find that in Loch Long the great- 
 est depth is not in the centre of the loch, ij3ut alternately 
 nearer to one or the other side. Especially is this the case 
 below the point at which Loch Goil joins the main valley on 
 the v/cst, v/here the deepest Avater is found below the cliffs 
 on the east side. From that point the water slowly shoals from 
 48 fiithoms to 34 near the mouth of the loch, the deepest 
 Avater being thus found about 6 or 7 miles from the Clyde. 
 In the Gareloch, a vestibule of shallow water exists outside 
 the barrier, so that the 21 fathoms found within the barrier 
 represents a greater depth than is met with in the Firth of 
 Clyde for 4 m.iles farther down. Near the mouth of Loch 
 Fyne, again, a depth of 104 fathoms has been ascertained; but 
 the opener water for some miles below the point of junction 
 v/ith the estuary does not show more than 85 fathoms. In 
 ail these cases, then, we have rock basins, as they have been 
 called — that is to say, lakes resting in hollows of the solid 
 rock — whose greatest depth exceeds that of the channels into 
 which their surplus Avaters are poured, and whose outflow is 
 over a barrier of solid rock, not of mere superficial detritus. 
 The same facts are illustrated by others of the West Highland 
 lochs, it being especially noteworthy that the line of deepest 
 soundings is deflected at the points of junction of two valleys, 
 and follows also the course which a glacier Avould have folloAved 
 in the main A^alley, approaching alternately to one or the 
 other shore. The example by Avhich Professor Ramsay 
 established his theory — that of the Lake of GcncA'-a — is that 
 of a lake 983 ft. in depth at its deepest part, Avhich is nearly 
 in the centre; a glacier entered the lake Avith the thickness 
 of not less than 3000 ft., and thus exerted on the com- 
 paratively soft rocks of the bottom an enormous vertical 
 pressure, which greatly increased the eroding poAver conferred 
 by the forward movement of the mass. But, from the point 
 
LAKES m AHEAS OP SUESIDENCE. ITl 
 
 wliere tLis giinding action was greatest, it must liave dimin- 
 islied ill proportion as the ice was melted, in tlio same way 
 that, at the present time, the lower extremities of glaciers are 
 thinner than the upper; hence the friction which, if continued 
 with equal force, would have suf&ced to grind dov.ai a whole 
 valley to the same depth, was arrested, and left a hollow, 
 v/hich, when the glacier retreated to its present limits, became 
 filled up v>^ith v/ater, and the Khone now floAvs out of the 
 lake ill an ordinary river valley. It must not be supposed 
 that this theory is applicable to all lakes; all that is contended 
 for is, that it solves the problem of the occurrence in so many 
 regions, where ice has formerly existed in considerable 
 quantity, of deeply excavated basins, whose floor is below the 
 level of the channel of outflow, and in many cases also below 
 ■ — it may be considerably below — the level of the sea. 
 
 141. Lakes Formed by Subterranean Erosion. — 'A few- 
 lakes, of small size, arise in some low-lying districts in con- 
 sequence of subsidence, after the removal of some of the 
 underlying bodies of rock; thus, in limestone districts, under- 
 ground channels frequently exist, the calcium carbonate 
 having been slowly washed away by v/ater containing de- 
 composing organic matter; and the subsidence consequent 
 upon tlie breaking- in of the roofs of these underground 
 channels not merely creates a depression, but, by blocking 
 up the subterranean watercourse, forces its contents up to 
 the surface. The removal of solid beds produces a similar 
 effect, whether the removal is natural, as in the limestone 
 districts and where beds of rock salt have been dissolved 
 away, or artificial, as Avhere subsidence follows mining 
 operations. 
 
 142. Lakes in Areas of Subsidence. — The Caspian Sea, or 
 lake, as it ought strictly to be called, covers an area as largo 
 as Spain, its length being 180 miles, its aver.age breadth 210 
 miles, and its maximum depth 3,000 feet. Not merely is its 
 surfiice 83 feet 6 inches belov/ the level of the Black Sea, 
 but if the barrier between the two were submerged, a vast 
 tract of land would be beneath the waters, including i^.stra- 
 khan and Orenburg. There can be no doubt that this great 
 basin is one of true depression, in which the Aral Lake has 
 to some extent shared. It is about one-nfth of tho area of 
 
172 PHYSICAL GEOGRAPHY. 
 
 tlie Caspian; it is 120 feet above tlie level of tliat lake, and 
 37 feet above that of the Black Sea. Its maximnm cleptli, 
 95 feet, is towards the western side, the eastern portion being 
 shallow. The Dead Sea occupies the southern end of a 
 longitudinal valley, the northern portion of which is drained 
 by a stream which cuts across the valley wall at Belfort, and 
 enters the Levant. The Jordan passes through the fresh 
 water Lake of Tiberias on its way to the salt Dead Sea, 
 whose surface is 1,300 feet below the level of the Mediter- 
 ranean. The deepest part of the lake is towards the north 
 end, where it is 1,308 feet; at the south end it is only 104 
 feet. Probably many of the lakes in the interior of conti- 
 nents owe their origin to slow subsidence; but, in defect of 
 evidence, it is safer to speak of their relations without specu- 
 lating on their origin. One of the best examples of lako 
 formation by rapid subsidence is that of Missouri, in which 
 State the earthquake of 1811 — 12, felt over an area of 300 
 miles north and south, let down a tract of 70 to 80 miles 
 north and south, by about 30 miles east and west. Thus the 
 Sunk Country, near New Madrid, is only one of several such 
 depressions in that region, and the cause is still visible in the 
 trenches and sink holes which were opened at the time of 
 the convulsion, water and fragments of rock being forcibly 
 ejected from the latter. The White Water pours into the 
 lake, as does the Mississippi when high in flood. 
 
 To this group may be referred a certain small number of 
 lakes, never of any very great size, which occupy the craters of 
 some extinct volcanoes. Their waters are either derived from 
 atmospheric sources, or they are filled from below by subter- 
 ranean channels. In Trinidad, a lake was formed on the 
 western coast by the subsidence of a patch of land, the hollow 
 becoming filled with a small lake of fluid bitumen; this pro- 
 cess having likewise probably originated the great Pitch 
 Lake of that island. "^ 
 
 143. Lakes Formed by Elevation. — Some of the lakes 
 found, for example, in the Andes at a height of 12,000 feet 
 may represent local depressions, or, as is possible in that 
 disturbed region, an upheaval of part of the valley floor may 
 have blocked the valley and converted it into a permanent 
 Jand-locked basin; but evidence is v/anting on this subject, 
 
CONTINENTAL LAKES. 173 
 
 and the explanation is suggested because the general features 
 of these valleys strongly resemble those of ordinary valleys 
 excavated by rain and rivers. To this group belong Lake 
 Titicaca and its analogues in the Andes, and Salt Lake in 
 the E-ocky Mountains. 
 
 144. Continental Lakes. — ^A large number of lakes cannot 
 be referred with certainty to any of the above-mentioned 
 groups ; the topographical details, as well as the geological 
 structure of the areas in which they occur, being unknown or 
 imperfectly known. The great lake region of Africa is 
 possibly in an area of subsidence, at least in part ; but the 
 origin of each particular lake is unknown, as their relations 
 to each other are still obscure. Lake Tchad is variable in 
 its dimensions, being in some seasons a shallow lake covering 
 upwards of 200,000 square miles, at other seasons reduced 
 to a marsh, connecting pools. The country in the midst of 
 which it lies is a wide, grassy plain, the monotony of which 
 is scarcely interrupted by a shi-ub. If the definition of a 
 lake involves the idea of a special basin, Tchad scarcely 
 fulfils this condition. The temporary lodgment of water on 
 the surface of a plain is a frequent occurrence on the plateaux 
 of Africa, Asia, and Australia, and the former existence of 
 such sheets in regions now too arid for the development 
 even of a marsh, is proved by the rait incrustations on 
 the surface left on the evaporation of the water. The 
 following are the principal lakes of Equatorial and Southern. 
 Africa: — 
 
 Height above Sea. 
 
 Victoria N'Yanza,... 3, 740, or 3,308, Speke. 
 
 Albert N'Yanza, 2,720, Baker; 2,500, Buclian (computed). 
 
 Tanganyika, 2,880, Livingstone; 2,800, Finlay (computed). 
 
 Liemba, 2,880, Livingstone. 
 
 Bemba, 4,000. 
 
 Nyassa, 1,522, Kirk. 
 
 Whether Tanganyika is in connection with Albert N'Yanza, 
 as Baker was informed by the natives, or is distinct from it, 
 as Livingstone believes, rests still (1873) on the accuracy of 
 the observations of height above the sea, in defect of the 
 practical solution by success or failure in the attempt to pass 
 from the one to the other. The conflicting statements bring 
 
174 
 
 PHYSICAL GEOGKArHY. 
 
 into prominence one fact of great interest, namely, that tlie 
 dimensions of Tanganyika vary with the season. That its 
 waters are fresh, points to its having some outflow free 
 enough to prevent the accumulation of salt in a region where 
 that is the most frequent result of evaporation. The lakes 
 of the Asiatic high grounds have probably very various 
 origins, some lying in glacier lines, others on the plateaux, 
 while others are perhaps due to movements of disturbance 
 which have altered the level of the valley bottoms. The 
 Australian continent has great sheets of w^ater in the interior, 
 their limits being variable, as in Africa ; but their extent, 
 comparative fixity, and association with lakes whose supply 
 is permanent, are points on which there is not yet precise 
 information. 
 
 145. Analysis of Lake Waters. — Ramsay gives {N'ature, 
 vol. vii, p. 312) the following table, illustrating the composi- 
 tion of the salts in the water of partially and completely land- 
 locked basins, and the last column shows the composition of 
 those obtained from the Old Well, Bath : — 
 
 rerccntag-e. 
 
 Mediter- 
 ranean. 
 
 Black 
 Sea. 
 
 Sea of 
 Azof. 
 
 Caspian. 
 
 Dead 
 
 Sea. 
 
 Old Well, 
 Bath. 
 Grains 
 per gal. 
 
 Chloride of Sodium, 
 
 „ Ma^nesiuia,... 
 
 „ Calcium, 
 
 ,, Potash, 
 
 Bromide of Magnesium, . . 
 Sulphate of Lime, 
 
 ,, Magnesium, .. 
 
 Bromide of Sodium, 
 
 Carbonate of Lime, 
 
 ,, Magnesia,... 
 
 Peroxide of Iron, 
 
 Bicarbonate of Soda, 
 
 Sulphate of Soda, 
 
 2-9480 
 0-3223 
 
 0505 
 
 0-1357 
 0-2430 
 0-0553 
 0-0113 
 
 0-0094 
 
 1-4020 
 0-1304 
 
 0-01S9 
 0-0005 
 0-0105 
 0-1470 
 
 0-0359 
 0-0209 
 
 0-9G5S 
 
 0-0887 
 
 0-0128 
 0-0004 
 0-0288 
 0-0764 
 
 0-0022 
 0-0129 
 
 0-3G73 
 0U32 
 
 0-0076 
 trace "^ 
 0-0190 
 0-1-239 
 
 0-0171 
 0-0013 
 — / 
 
 12-110 
 
 7-822 
 2-455 
 1-217 
 
 0-452 
 
 12-5 
 14-5 
 
 80 
 9 
 
 17 
 
 3-770 
 
 1-7661 
 
 1-1889 
 
 0-6294 
 
 24-056 
 
 1S3 
 
 From these figures it appea,rs that there is a diminution in 
 the proportion of common salt from the Mediterranean to the 
 Caspian, and that the waters of the Dead Sea contain a 
 relatively enormous quantity. The former fact is intelligible 
 when it is remembered that these seas receive a very largo 
 supply of fresh water from rivers, and the latter acquii-e^ 
 
ANCIENT LAKES AND LACUSTRINE AREAS. 175 
 
 interest from the circumstance tliat the sole aS.uent'of the 
 Dead Sea is the Jordan, which flows out of a fresh water 
 lake, that of Tiberias. 
 
 Lakes have been divided into groups — those which have an 
 outflow, and those which have not. This simple grouping 
 corresponds in some respects to the difference between salt and 
 fresh water lakes. But as the saline character va.rie3, the 
 Caspian and the Dead Seas being very unlike in this respect, 
 the lakes which have no outflow ma.y perhaps be divisible 
 into those which ha,ve an abundant supply of fresh water, 
 and those in which it is small relatively to the size of the 
 lake. Since salt is found crystallized or efflorescent wherever 
 sheets of fresh water have evaporated, as in the steppes of 
 E-ussia, the 'plains of North Persia, of Africa, America, and 
 Australia, the facts of the Jordan Vcilley justify the pro- 
 visional generalization that when a lake is fresh the pro- 
 bability is that it has a steady supply and discharge, even 
 though the latter may be secured by percola,tion through 
 fissures in its basin, llueh stress cannot be laid on the 
 apparently exceptional case of the Casj)ian, for though that 
 lake was probably Salter when its separation from the Black 
 Sea was accomplished than it now is, there is reason to 
 believe that its saltness is increasing, evaporation lowering 
 its level more rapidly than the fresh water is poured in. 
 
 146. Ancient Lakes and Lacustrine Areas. — Guided by 
 the analogy of existing formations, Mr. Godwin Austen first 
 suggested that the old red sandstone deposits represented 
 lacustrine formations on a large scale. Professor Kamsay 
 has applied this suggestion in the investigation of several 
 geological formations, and has succeeded in showing the pro- 
 bability that most of the red coloured strata have been formed 
 in lakes or inland seas. The red colour due to the presence 
 of salts of iron cannot be explained on the supposition that 
 these strata represent accumulations in the open sea, since 
 all marine strata with which we are acquainted are of various 
 hues, but never red. It is scarcely to be expected that a dej^osit 
 of peroxide of iron should take place in the ocean, since it3 
 diffusion would certainly be immediate. In many lakes, as 
 in those of Sweden, peroxide of iron is deposited ; and the 
 old red sandstone, at least of the Scottish type, is an exceed- 
 
176 PHYSICAL GEOGRAPHY. 
 
 ingly local formation. Tlie area whicli it covers was pro- 
 bably limited towards tbe south by a barrier of land, which 
 stretched across England from east to west, and which, if 
 prolonged, would pass to the south of Ireland. This barrier 
 was an important one in the carboniferous period, serving to 
 isolate two distinct types of coal formations, though these terres- 
 trial or swamp accumulations are not so distinct from each other 
 as the comparatively unfossiliferous old red sandstones to the 
 north of this line are from the truly marine and abundantly 
 fossiliferous deposits of the devonian series to the south of the 
 British Channel. The old red sandstone rests conformably upon 
 the upper Silurians, and the transition from one to the other 
 formation is effected by alternations of gray and red strata, 
 which gradually culminate in the entirely red sand'stones. The 
 difference in tint likewise corresponds to a difference in 
 organic contents, and with the gray beds the characteristic 
 Silurian fossils disappear. The life of the old red sandstono 
 consists of vegetable remains of terrestrial types, of shells 
 which are closely allied to fresh water families, and of ganoid 
 fishes, whose modern representatives are confined to tlie fresh 
 waters of the Nile, the North American rivers, the rivers of 
 Australia and of South Africa. The paucity of life on the 
 whole is remarkable ; the small size of such shells as do occur, 
 and the absence of characteristic marine remains, give pro- 
 bability to a hypothesis which rests to a considerable extent 
 upon physical and chemical grounds. The permian strata 
 present very many of the same characters; the shells are few 
 in number, the remains of amphibians or terrestrial reptiles 
 are abundant, their footprints being conspicuous where all 
 traces of the body have disappeared. Such plants as are met 
 with belong to terrestrial types; and, when all these are 
 considered, along with the fact that magnesian limestones 
 occur in large quantity — the magnesian carbonate being de- 
 posited by evaporation — that gypsum occurs in considerable 
 quantity, and that jpseudomorphous crystals of rock salt 
 indicate evaporation in confined areas, while peroxide of iron 
 gives a characteristic tinge to the beds, there is good reason 
 to believe that in the permian we find evidence of the exist- 
 ence of a confined basin, not necessarily a salt sea, but one 
 which, like the Caspian, may have gradually become mors 
 
FILLING UP OF LAKES. 177 
 
 saline. The same reasoning wonld apply to tlie cambrian 
 and other red-coloured pal£eozoic strata; hut the corrobora- 
 tive evidence derived from the character of the fossils cannot 
 be obtained, in consequence of the small amount of unaltered 
 Cambrian strata now exposed at the surface, and the great 
 metamorphism to which the deposits of that age, in England 
 and Ireland, have been subjected. 
 
 147. Filling up of Lakes. — The history of a lake basin 
 subsequent to its formation is very simple. Every tributary, 
 as well as the main stream, carries down its contribution of 
 detritus, and if the lake is of any size, the sediments have 
 time to subside, at least to a considerable extent, before the 
 waters reach the point of outflow. Delta formations com- 
 mence, and gradually carry the mouths of the streams farther 
 out towards the centre of the lake, till at last they join the 
 main stream as it flows through the swamp, which the shallow- 
 ing of the water by deposit tends to form. Thereafter we should 
 have a river flowing through a plain of alluvial materials, and 
 joined by tributaries at various angles. This is the explana- 
 tion of the frequent occurrence of alluvial plains at various 
 levels in the course of streams. The subsequent events are 
 identical with those that take place in the plains of open 
 grounds, the river gradually cutting its channel deeper into 
 the alluvial soil, shifting its position from time to time, and 
 thus at once deepening its o^Yll bed and slowly lowering the 
 general level of the plain. 
 
 SECTION III.— WATER IN THE INTERIOR OF 
 THE EARTH. 
 
 ' Conditions which permit its Presence — Springs due to Percolation, 
 or Connection with Volcanoes — Water Supply of London — Limit 
 of Saturation — Pollution of Water — Retentive Power of Strata 
 due to their Character, or to Subjacent Strata — Lodgment of 
 
 U Water — Leakage of Subterranean Reservoirs — Artesian ^Vells — 
 Yield of Springs — Contents of Spring Water — I^Iineral and Ther- 
 mal Springs — Springs in Yellowstone Valley — Thermal Waters 
 in England — Iceland : Theory of Intermittent Springs — Periodic 
 Springs — France and Spain — Bischof's Classification of Springs 
 — Examples of Springs Geologically Interesting — Petroleum 
 Springs — Underground Pavers and Caverns — Contents of Caves 
 — Landslips. 
 
 23 M 
 
173 PHYSICAL GEOGRAPHY. 
 
 148. The Conditions which permit its Presence.-— 
 
 Hitherto we have treated of water in the form of extensiva 
 sheets — oceans; of smaller sheets — lakes and inland seas; 
 and of the tributaries of these — rivers.. "VVe have now to 
 consider the v/^ater held in suspension in the atmosphere, and 
 the various forms under which it returns to the earth. In 
 the present section we shall speak of the circulation of water 
 in the interior of the earth, its movements, its properties, 
 and its results. The quantity of water which sinks into the 
 soil is obviously proportioned to the porosity of the surface, 
 and to the length of time during v/hich it is retained on any 
 particular point. Thus, if water falls on a rock surface, the 
 greater part of it flows off directly; if it falls on sand — sup- 
 posing that the sand is not overheated, as is that of the 
 desert — ^it sinks in, and descends vertically, until it is 
 arrested by some less porous stratum. 
 
 149. Springs exclusively due to Percolation, or Connec- 
 tion with Volcanoes. — ^Ifc may be said in general terms that 
 there is a subterranean circulation parallel to that on the 
 surface; and, if there v/ere no curvatures of the strata such 
 as have been described, v\^e should expect to find the largest 
 quantity of subterranean v^ater in the low grounds. The 
 exceptions, however, to this general statement give rise to a 
 large number of varieties of springs — varieties both as regards 
 their origin and their chemical properties. Those springs 
 v/hich are more or less intimately related v/ith volcanic 
 phenomena illustrate two different sets of conditions; in the 
 one case the springs are directly due to the volcanic action, 
 and in the other they are only modified by the vicinity of 
 volcanic activity. Water which sinks into the soil is either 
 retained by the strata into which it passes, or, if the quantity 
 of moisture supplied is in excess of their retentive power, it 
 escapes directly, but unseen, into rivers or the sea, or sinks 
 to still lov/er levels, or issues at the surface in the form of 
 springs. 
 
 150. Water Supply of London. — The history of the city of 
 London illustrates, perhaps better than any other case, at 
 once the economic importance of underground water supply, 
 and the principal geological features upon which tliat supply 
 depends. The valley in which London is situated is bounded j 
 
WATER SUPPLY OF LONDON. 179 
 
 to tlie nortli by the outcrop of the chalk in Hertfordshii-e, to 
 the south by the outcrop of the chalk in Surrey, and in the 
 hollow thus formed lies the London clay, a mass of lower 
 eocene strata. A vertical section beneath London passes 
 through the following strata : — 
 
 Bagshot Beds, 
 London Clay, 
 
 Woolwich and Reading Beds, 
 Thanet Sands, 
 and these rest upon Chalk. 
 
 Thus there is on the surface a permeable stratum of gravex 
 from 10 to 20 feet in thickness, resting upon tenacious clay 
 of from 100 to 200 feet in thickness, while beneath this 
 retentive mass 70 to 100 feet of sands and gravel are inter- 
 posed, and the subjacent chalk is itself also porous. The 
 growth of the City of London has been determined by this 
 upper gravel mass, for the ease of obtaining water supply 
 fixes the population necessarily on its surface. Where, to the 
 west and north-west of London, clay is exposed at the surface by 
 denudation, population for a long time was arrested, until 
 artificial water supply was obtained. Again, the river and 
 its tributaries are bordered by a lower lying bed of gravel, 
 and on this again a line of villages sprang up. The popula- 
 tion became continuous, and extension into the adjacent dis- 
 tricts was possible only when water was conveyed from a 
 distance by conduits. Again, the northern heights of Lon- 
 don are composed of sands from 30 to 80 feet thick, which 
 readily yielded a sufficient number of wells for the villages 
 which formerly occupied these sites. To the west, where the 
 Bagshot sands increase in thickness to 300 or 400 feet, the 
 depth to which it would have been necessary to sink for water 
 prevented these localities from being occupied. Finally, 
 where the sands which underlie the London clay crop out at 
 the surface, the line of outcrop determined another series of 
 villages. There is therefore a trouGjh in the centre of which 
 a large quantity of water accumulates, part of it bein^:^ovo 
 and part below the level of the river and of the sea. The 
 water sinking through the gravel which borders the river has 
 its level determined by that of the adjacent stream, and tho 
 surplus would overllow at tho surface if an opening wero pro- 
 
180 PHYSICAL GEOGRAPHY. 
 
 vided for its escape. But if, either by drouglit or by an 
 excessive withdrawal of water through these wells, the quan 
 tity contained in the porous strata is very much diminished, 
 then the level will sink correspondingly, and none will over- 
 flow. (Prestwich, Quart. Jour. Geol. Soc, 1872.) 
 
 151. Limit of Saturation. — The limit of saturation, there- 
 fore, has a very important effect upon the amount and per- 
 manence of water supply. In the centre basin the sands are 
 from 100 to 200 feet below the level of the Thames; at 
 Greenwich, the basin containing these strata is notched to a 
 depth of 100 feet, and below the level of the bottom of this 
 notch the London basin is constantly saturated. The chalk 
 is itself permeable by water, but it is comparatively retentive, 
 since a cubic foot has been found to retain 2 gallons of 
 water by mere capillary attraction, and the water is dis- 
 charged therefore very slowly. But the chalk is traversed by 
 fissures and broken up by lines of flints; and although these 
 above the limit of saturation are, so to speak, slow conduc- 
 tors of water, the movement below that level is very much 
 more rapid — there is, in fact, a tolerably free circulation. 
 The gradual drainage, by artificial wells, of the strata above 
 the chalk has compelled the penetration of the chalk itself, 
 which now supplies a large part of London south of the 
 Thames. The water is, as has been said, derived from the 
 surface, the rainfall being the most important source; but 
 some is also obtained from the adjacent high grounds, while 
 the river, aided by the rise and fall of the tides, dams back a 
 large body of water at and below its own level. 
 
 151. Pollution of Water. — It is obvious that, in addition 
 to the rain, any substances found upon the surface will like- 
 wise flow down into this subterranean reservoir; and 
 attention has been called to the fact that around London the 
 old habit of sinking cesspools has contributed a considerable 
 quantity of contaminating matter to the water suj-yply. 
 That these matters travel downwards with exceedinc: slow- 
 ness is true, and it must also be remembered that the more 
 distant is the source of the evil from the great body of the 
 water, the greater is the chance of injurious substances being 
 permanently kept back by filtration ; but the danger is one 
 deserving of consideration. 
 
ARTESIAN WELLS* 181 
 
 163. Retentive Power of Strata due to their Character, 
 or to Subjacent Strata. — The retentive power of strata 
 varies considerably ; the case of the chalk has already been 
 mentioned, and the sandstones of the older deposits are per- 
 liaps even more retentive than it. It may in fact be said, in 
 general terms, that the return of the water to the surface 
 would be restricted entirely to the escape between the planes 
 of stratification, but for the joints and fissures with which all 
 rocks are traversed. In the case of a conical hill, composed 
 of nearly horizontal strata of which the lowest above the 
 level of the valley is very hard, the quantity of water con- 
 tained in the hill above the hard bed will present a curve 
 which rises highest in the centre, and has its height deter- 
 mined by the annual rainfall. In such a case springs might 
 be expected to issue at the junction of the softer and the 
 harder rocks. 
 
 154. Lodgment of Water. — Where the strata are highly 
 inclined, and especially where they are finely laminated, even 
 though they themselves are not porous, surface moisture will 
 find its way down the stratification planes, and the succes- 
 sive layers of water become connected by the joints and fissures 
 of the strata, so that, following the inclination or dip of the 
 beds, we should expect to find at some point a reservoir, 
 whose place might be fixed with tolerable precision if the 
 strata curved in the opposite direction so as to form a 
 synclinal trough. But in the centre of the trough a fissure, 
 whether due to a fault or to the upward passage of a trap 
 dyke, might form a natural drain for the reservoii', and thus 
 the water might escape to still lower levels. 
 
 155. Leakage of Subterranean Reservoirs. — The general 
 proposition that a series of porous strata resting upon an im- 
 pervious one, more especially when a natural trough exists, 
 will yield a supply of water, is dependent on the contingency 
 of subterranean disturbances, the existence of which may not 
 be indicated by the surface features. But in a series of 
 water-conducting strata, a fault line or a trap dyke may have 
 the opposite efiect of damming back the water, and thus 
 forcing it to escape at the surface, and we have, in such a 
 case, a natural Artesian well. 
 
 166. Artesian Wells. — These v>^ells have derived their 
 
182 PHYSICAL GEOGRAPHY. 
 
 name from the place in wMch. they have been so long em- 
 ployed, Artois, in France. The method of sinking an Artesian 
 well is the same as that by which we endeavour to ascertain 
 the presence of underground minerals. Augers, iron instru- 
 ments, are gradually sunk down by the hand, or, when the 
 depth is great, by machinery, the hard rocks being pounded 
 and the debris withdrawn, and when the instrument at last 
 penetrates into the water-bearing bed, the fluid escapes to the 
 surface with varying force and persistence. If the boring is 
 arrested at this point, the supply may be permanent; if it is 
 carried still farther, the penetration of other reservoirs may 
 yield an increased quantity; but it has happened that the last 
 retentive stratum has been passed through, and the water, 
 instead -oi esca,ping at the surface, sinks to still lower 
 levels. The deepest wells in England range from 450 to 550 
 feet, and the water which escapes from them comes from the 
 chalk hills at least 15 or 20 miles from London. The well 
 of Grenelle, the bottom of which is more than 1,700 feet 
 below the sea level, drains a district a,bove 100 miles distant 
 from Paris, its lowest point being 1,798 feet; that of Passy 
 is sunk to 1,923 feet. The diameters of these bores are 
 various, but the Passy well is 4 feet at the surface and 2 feet 
 4 inches at the bottom. At La Chapelle, St. Denis, Mr. 
 Prestwich says the bore was commenced 157 feet above the 
 sea level; a shaft, 6 J feet in diameter, was sunk for 113 feet; 
 the bore thence started with a diameter of 5^ feet, and in 
 1872 had reached a depth of 2,034 feet, with a diameter of 
 4 feet 4:lj inches. It is expected that the lower greensands, 
 the stratum below the chalk, will be reached at a depth of 
 2,300 feet. Wliile the boring usually goes on uninterrup- 
 tedly, it happens occasionally that the auger suddenly sinks 
 some distance, and the inference is that the instrument has 
 penetrated into a hollow reservoir. From such a reservoir 
 at Tours were obtained land and fresh water snails and the 
 seeds of water plants, and in Westphalia, under similar cir- 
 cumstances, fish were obtained, these having doubtless come 
 from the nearest streams, situated several leagues away. 
 Desor verified Zickel's observation, that fish were obtained 
 from the Artesian well in the oasis of Ain Tala, in the north- 
 eastern part of the Sahara; and as these fish are found like- 
 
 'I. 
 
CONTENTS OP SPrJNG WATEIt. 183 
 
 Wise in tlie neighbouring pools, it is probable tliat tlie pools 
 and tlie wells are both supplied from a common subterranean 
 reservoir, to which distant streams have contributed by 
 canals wider than simple fissures. 
 
 157. The Yield of Springs. — The quantity of water which 
 issues from springs varies considerably. Ketentive rock, 
 parting slowly with its contents, will yield in temperate 
 regions a permanent spring, betraying fluctuations only at in- 
 tervals; thus it has been calculated by Mr, Beardmore that 
 water requires froui four to six months to pass from the sur- 
 face to the saturation level in the chalk, and that the effect 
 of the winter rainfall is therefore not apparent in deep springs 
 before summer. Hence the supply is practica,lly permanent 
 under existing conditions, since the extreme effect of a dry 
 summer and autumn v/ould not be appreciable fo]' sixteen 
 months — that is to say, the storage or reserve of water is 
 more than sufficient to maintain the stream till the effects of 
 the next rainfall are appreciable. More porous strata, 
 through which the transmission of water is rapid, are more 
 liable to fluctuations, and even to intermissions. On the 
 other hand, in limestone, where the water is confined to the 
 plains of stratification, and to the joints and fissures which 
 traverse the rock, the delivery is frequently as rapid as the 
 supply, and the spring therefore is dependent entirely on the 
 rainfall for its maintenance. The yield of the Artesian wells 
 furnishes interesting differences. Thus the well at Fulham, 
 .317 feet deep, yielded 50 gallons per minute; that of Tours 
 discharged 300 cubic yards of water in twenty-four hours. 
 But the force of discharge varies still more. In the Sheer- 
 T^css well the water rose from 328 to 189 feet rapidly; but it 
 required some hours to ascend to 8 feet above the ground. 
 The Chiswick well rose 4 feet above the ground from 620 
 feet; and one at Tooting did considerable damage, from the 
 force of its discharge, when it was pierced. 
 
 158. Contents of Spring Water. — The water seldom 
 escapes pure. Apart altogether from the contaminations 
 already mentioned as resulting from the ignorance, selfish- 
 ness, and carelessness of men, some wells regularly, others 
 occasionally, give forth the debris of vegetables, indicating 
 that surface matters have from time to time got access to 
 
184 
 
 PHYSICAL GEOGEAPHY. 
 
 their sources. In other cases, solid inorganic matters are 
 delivered in suspension and in solution. The following table, 
 drawn up by Dr. Frankland, illustrates the varying quantity 
 of materials by weight in 100,000 parts of water, the liist 
 column representing the total quantity of carbon and nitro- 
 gen in the organic matter. As the effects of filtration are 
 conspicuous in the case of the nitrates, these results must, as 
 Mr. Prestwich remarks, be accepted with caution as indica- 
 tive of the orioinal condition of the water : — 
 
 EOURCED OF V/ATEU. 
 
 Vv'ell, Royal Institution, London, 
 
 „ Eoyal Mint, 
 
 ,, Barclay's Brewery, , 
 
 Thames V/ater at Hampton, , 
 
 Springs, Head of Tliames, , 
 
 „ Moorparic, 
 
 ,, Otter, near Watford,.. . . 
 Well, Croydon, 
 
 ,, Cfiterham, 
 
 ,, Crenelle, 
 
 S 2 
 
 S^irface Gravel,.. 0-525 
 Tertiary Sands.... 220 
 
 „ .. 0-065 
 
 — I 0-2S4 
 
 Oolites....... .. ..! 0-023 
 
 Lower Greensand,' 0-040 
 
 Chalk, I 0-038 
 
 „ I 0-047 
 
 „ i 0-026 
 
 Lower Greensand, 021 
 
 
 et-t 
 
 
 c 
 
 O 
 
 n 
 
 ^ m 
 
 *? ei 
 
 r: 
 
 
 
 
 if^ 
 
 ?t 
 
 tn^ 
 
 p h 
 
 
 CM 
 
 !25^ 
 
 o 
 
 H 
 
 4-855 
 
 20-8 
 
 93-70 
 
 
 
 7-7 
 
 S3 -96 
 
 0-035 
 
 4-1 
 
 71-56 
 
 0-196 
 
 15-7 
 
 27-87 
 
 0-3f8 
 
 17-0 
 
 •28-25 
 
 0-034 
 
 
 
 4-55 
 
 0-422 
 
 21-0 
 
 32 -06 
 
 0-551 
 
 12-9 
 
 320 
 
 0-027 
 
 10-4 
 
 SI -08 
 
 » 
 
 6-8 
 
 14-09 
 
 It appears from this table that the freedom ot water from 
 organic matter is in proportion to the depth of the spring ; 
 but the facility with which organic matters scattered through 
 the soil are oxidated must be kept in mind; for the conclusion 
 that the source of deep wells is pure would be erroneous, 
 since the purity is only due to the prolonged filtration the 
 waters have been subjected to. 
 
 159. Mineral and Thermal Springs. — "While no springs 
 are absolutely pure, while the waters always contain a certain 
 amount of inorganic matter, the term mineral springs is 
 employed to designate those in which the mineral matters are 
 either in great excess, or of a kind not usually met in waters. 
 Thus, the silica of the Iceland geysers is not a normal con- 
 stituent, nor are the arseniates of the African springs. Car- 
 bonate of lime, again, is present in most waters, but its excess 
 in the springs of Auvergne and Tuscany, as well as in those 
 which arise in limestone regions, entitle these waters to 
 
SPRINGS IN YELLOTfSTONE VALLEY, NORTH AMERICA. 183 
 
 separate recognition. Again, while all springs of consideralDle 
 depth, have a temperature greater than that of the surface — 
 the increase ranging from 1" C. in 50 to 1° in 80 feet of 
 descent, the highest average of those not associated with 
 volcanoes being 27*7° C. — thermal springs are those w^iose 
 temperature is in excess of that proper to their depth. 
 Both mineral and thermal springs occur for the most 
 part in districts where there are either active or extinct 
 volcanoes, or in which earthquake phenomena are of fre- 
 quent occurrence, or, lastly, in localities which have been the 
 seats of very great disturbance. 
 
 160. Springs in Yellowstone Valley, North America. — 
 The Yellowstone National Park of the United States, a 
 district about half the size of Wales, presents a combination 
 ot phenomena as regards springs which will be more instruc- 
 tive than a series of systematic statements. In this area of 
 3,575 square miles, recently, geologically speaking, one of 
 volcanic activity, the sedimentary strata are carboniferous, 
 Jurassic, cretaceous, and tertiary, the whole resting on meta- 
 morphic rocks. Volcanic rocks occupy large part of the area, 
 and as these began to be deposited in tertiary times, the 
 present replacement of the igneous matter by springs 
 issuing from the old craters may be looked on rather as an 
 episode in the life ot a volcanic region than a sign of its 
 final quiescence. The floor of the valley in which Gardiner's 
 Biver (a branch of the Yellowstone) flows is covered with a 
 crust deposited by springs now no longer calcareous, and the 
 slopes are covered with a white crust 20 to 50 feet thick. 
 The water now issuing from beneath this surface layer forms 
 a stream 6 feet wide and 2 feet deep, whose temperature is 
 55 'S" C, but higher up the slope, in basins of 20 to 50 feet in 
 diameter, whose waters are 65*5° to 72° C. The calcareous 
 springs now active contain, in addition to the predominant 
 carbonate of lime, sulphuretted hydrogen, soda, alumina, and 
 magnesia. The siliceous springs are intermittent, boiling, and 
 quiet springs. The waters of the boiling springs are always 
 at 100° C. ; they graduate into the intermittent by those in 
 which the water is projected as much as 6 feet at regular 
 intervals. The intermittent are at 100° C. only during 
 activity, falling to 65° 0. in the intervals. The quiet springs, 
 
186 t'HYSICAL GEOGRAPHY. 
 
 once probably boiling, range from 64*4° C. to 26° C. ; and it is 
 interesting to note that tlie deposit of iron takes place at 
 lower temperatures than 65° C, the siliceous sinter showing 
 all gradations of colour, from deep red to the purest "white. 
 The mud springs vary in temperature and activity; the mud 
 contains a large quantity of alum, and is kept in movement 
 by jets of steam which toss up the pasty mass to the height 
 of 4 or 5 feet. There is overflow in these craters, and in one 
 where a jet of steam is constantly ascending to the height of 
 600 feet, the black mud is about 20 feet below the rim of 
 the basin. Sulphuretted hydrogen makes the atmosphere 
 near one spring oppressive. The geysers are most interesting, 
 from the variety of tint and form of the sinter with which 
 their orifice of outflow is surrounded. The column of one rises 
 for 200 feet, and after a quarter of an hour gradually subsides 
 to 2 feet below Ihe mouth of the orifice, the temperature 
 being then 6o° C. The waters of this, and the others of the 
 same group, contain as much as 85 per cent, of silica, 11 per 
 cent, of water, 4 per cent, chiefly of chloride of magnesium, 
 ■\^T.th a trace of lime. The association of such varied springs, 
 active and extinct, their generally pulsating character, 
 amounting in some cases to true intermittence, and their 
 association with earthquakes at the present time, in the midst 
 of a tertiary volcanic district — all these indicate the source of 
 the phenomena, but do not explain the chemical variety in 
 the waters, adjacent springs having unlike composition, and 
 afibrding dissimilar deposits. 
 
 In some of these thermal springs there is a large develop- 
 ment of confervseand similar vegetables, diatoms in abundance, 
 and that silky organic matter wdiich it is difficult to refer to 
 animals or plants with certainty. Chemically this haregin 
 varies in different springs, but its true nature is not knoAvn. 
 
 161. Thermal Waters in England. — The waters ol the Old 
 "Well at Bath, having a temperature of 48 '8° C, contain 144 
 grains of solid matter per gallon. The hot S2:)rings of Bristol 
 have a temperature of 20° or 22° C. At Buxton and Matlock 
 there are also warm springs. 
 
 The occurrence of calcareous springs in limestone districts 
 has already been referred to. Ferruginous springs are met 
 witli under similar conditions : thus the drainage of car- 
 
gPEINGS; FRANCE AND SPAIN. l87 
 
 Tboniferous districts sometimes yields oclireous springs, and 
 tlie water that lias drained silurian high grounds (in v/hich 
 there is an enormous quantity of diffused ii'on) often yields 
 a copious rusty precipitate. 
 
 162. Iceland: Theory of Intermittent Springs. — The vol- 
 canic region, of which Hecla is the active centre, contains 
 those geysers which were first studied by geologists. The 
 intermittence of these hot springs was long regarded as proof 
 of the existence of reservoirs in which water and gases v/ere 
 lieated till the expansion of both raised the fluid in the long 
 tube of the fountain to the level of overflow. The duct then 
 became a syphon, and an interval was necessary before the 
 reservoir was again full enough to permit the process to 
 recommence. In the chapter on Volcanoes it will be shown 
 that Tyndall's explanation of the gej^'sers as slowly but 
 steadily flowing springs, whose waters pass through portions 
 of rock at a high temperature, meets all the difficulties of the 
 case, and, moreover, renders it intelligible that intermittent 
 springs should be of such different chemical character. 
 
 163. Periodic Springs. — The alternations of a wet and dry 
 season give one kind of periodicity; but in those sj^ecially 
 included under the designation, the recurrence of the flow is 
 at intervals which cannot be thus accounted for. In some 
 cases the existence of a cavern is probable, the outflow being 
 little above the flow. As soon as the water attains a certain 
 level, its duct, bent and syphon-like, drains the cavern to the 
 level of outflow, and then ceases till fresh accumulation 
 repeats the process. In springs connected with volcanic 
 phenomena, the expansion of gases in such a reservoir is likely 
 to take place, ancl if it occurs is almost certain to furnish 
 periodic phenomena, supposing the supi^ly of water and gas 
 to be constant, and tlie orifice of discharge not too large. 
 
 164. France and Spain. — The volcanic areas in these two 
 countries furnish abundant examples of springs of various 
 kinds. Ansted points out that not all mineral and thermal 
 waters reach the surface of the ground; and his observation 
 accords with what was stated (Art. 27) regarding the courso 
 of lavas from the volcanic focus to the surface. In the older 
 strata, especially in the limestones of volcanic districts, -we 
 sometimes find that portions of tho limestone strata have 
 
ISS PHYSICAL GEOGRAPHY. 
 
 been silicifiecl from below, tbe alteration presenting a para- 
 bolic outline; and the only possible explanation is, that a 
 spring, probably thermal, sought to escape upwards by a 
 joint or fissure, which brought it in contact with a limestone, 
 and there exhausted its energy. The cases to which Ansted 
 refers occur in Anvergne, near Clermont, and in Cornwall — 
 the, so to speak, concealed springs in the latter locality 
 having only been discovered by boring. The calcareous 
 springs of Anvergne are well known, their waters being 
 employed there, as elsewhere, in the rapid production of 
 ornaments, by the deposit of lime in moulds over v/hich the 
 water is conducted. Carbonic acid, boracic acid, carbonates 
 of soda, and even arseniates of soda and potash, are found in 
 the waters of Vichy, the carbonates of soda forming by far 
 the largest ingredient. Sulj^hur in combination, as sulphur- 
 etted hydrogen or sulphate of lime, is of very frequent occur- 
 rence, especially in the Pyrennean springs. In Sj^ain the 
 presence of sulphurous springs seem more usually associated 
 with deposits of gypsum (sulphate of lime) than with the 
 jDresence of volcanic materials. 
 
 165. Bischof s Classification of Springs. — The substances 
 found in spring water are thus enumerated by Bischof 
 {Chemical Geology, i., p. 74), and, as they are usually im- 
 perfectly quoted, the list is here given entire : — 
 
 1. Saline Bases — Soda, potash, lithia, ammonia, lime, magnesia, 
 strontia, baryta, alumina, protoxides of iron and manganese, oxides 
 of zinc and copper, tin, lead, silver, antimony, arsenic, nickel, cobalt, 
 probably also as oxides. 
 
 2. Acids — Carbonic, sulphuric, sulphurous, nitric, phosphoric, 
 boracic, silicic, hydrosulpliuric. 
 
 3. Halogens and Metaloids — Chlorine, bromine, iodine, sulphur, 
 hydrogen. 
 
 4. Organic Substances — Extractive matter (baregin), crenic and 
 apocrenic acids. 
 
 And the chemical substances are obtained in different quan- 
 tities, according to the mode of origin of the spring. Bischof 's 
 classification according to origin is the most natural, from 
 the practical chemist's point of view. It is — 
 
 1. Springs which originate from rivers. 
 
 2, Springs originating in the water which sinks through the beds 
 of brooks and rivers, 
 
EXAMPLES OP SPEINGS GEOLOGICALLY INTERESTING. 189 
 
 3. Springs which originate from elevated lakes. 
 
 4. Springs formed by the melting of the snow and ice of glaciers. 
 
 5. Mountain springs. 
 
 C. Springs from great depths. 
 
 166. Examples of Springs Geologically Interesting. — 
 Besides those already mentioned, the following are the best 
 known examples of mineral and thermal waters : — The Carls- 
 bad waters contain 462 grains per gallon of solid matter, and 
 have potash and soda as their dominant ingTedients, but they 
 contain also a considerable quantity of metallic compounds. 
 The waters of Aix la Chapelle emit large quantities of nitro- 
 gen, probably derived from decomposition; and nitrogen in 
 one combination or another is found in many of the springs 
 in the central plain of Germany. The calcareous w^aters of 
 France have been mentioned; the source of their lime is 
 obscure, since few of the rocks iii that region contain it in 
 large quantity, and the disprop)ortion between the deposits 
 and the apparent source is even more striking in Tuscany, 
 where the deposits are rapid, continuous, and extensive. 
 Italy, rich in regions of volcanic activity, contains many 
 springs whose composition is as various as usual in volcanic 
 regions. The carbonic acid exhalations of the Campagna di 
 Roma, of the famed Grotto del Cane, near Naples, are of the 
 same kind as those of the Limagne d'Auvergne, and among 
 the extinct volcanoes of the Rhine. The enormous quantity 
 of this acid thus disengaged at the surface probably does not 
 represent all that is generated. Considering its power to 
 decompose many rocks, either as a gas or dissoh-ed in 
 ■water, it is obvious that this gas, derived from below" ground, 
 must have an enormous share in the alteration of strata in 
 addition to the work it does wiien carried down by rain from 
 the atmos})here to the soil. The deposit of silica from hot 
 springs has been mentioned, and the action of w^arm alkaline 
 w^aters in maintainincj silica in solution is intelliorible. But 
 its deposit from cold water is less clear. It appears that such 
 cold springs as form sinters either discharge little water, 
 which forms a stream so shallow as to admit of total evapora- 
 tion — as perhaps w^as the case when the veins of rocks were 
 lined, and ultimately filled, with quartz; or, as in the cold 
 siliceous spring of th.e Azores, the mineral is in the form of a 
 
190 PHYSICAL GEOGRAPHY, 
 
 double silicate of iron and alumina. It must always be hoYiiQ 
 in mind that our interpretation of wbat cold springs do is open 
 to fallacy, since the warm springs of Bagneres de Bigorre 
 became cold after the earthquake of 1660, while that of 
 Bagneres de Luchow, in the Pyrenees, formerly cold, has had a 
 temperature of 50° C. since the earthquake of Lisbon, 1755 — 
 two interesting examples of the connection between thermal 
 waters and subterranean disturbances, or, to put it in another 
 form, of the relation which still exists between the volcanic 
 centres and these areas of former disturbance which have lonsf 
 been quiescent. The geological interest of siliceous springs 
 lies in this, that fossilization, as of the wood in the Azores, 
 West Indies, and Australia, is effected on a comparatively 
 large scale by the agency of these underground waters. 
 
 167. Petroleum Springs. — The oil wells of America, 
 chiefly found in connection with devonian rocks, owe their 
 origin to percolation, the hydrocarbon compounds being de- 
 rived from the decomposition of animal and vegetable remains. 
 The borings by which the rock oil reaches the surface either 
 provide a channel for the fluid which travels along the fissures 
 of the strata, or tap reservoirs, just as the Artesian wells 
 reveal the presence of water caverns. It is not always clear 
 why the oil reaches the surface with impetus. The first dis- 
 charge from a boring is sometimes a quantity of gas with 
 more or less brine, the oil thereafter running pure, so that in 
 some cases, at least, the expansion of gases may be efficient. 
 But the whole history of these springs is still obscure. The 
 upper miocene strata of Trinidad yield the asphalt which 
 gives to the pitch lakes of that island their name. The 
 volcanic rocks of the carboniferous series in Scotland, on the 
 other hand, sometimes retain the products of decomposition, 
 as in the oil shale district of Bathgate, the bituminous matter 
 having doubtless slowly filtered out of the shales. Equally 
 retentive are the fetid limestones, that of the upper Silurians 
 at Niagara yielding its bitumen when the limestone is burnt 
 in kilns. 
 
 168. Underground Rivers and Caverns. — Underground 
 channels of considerable magnitude are sometimes formed, 
 and their formation is closely connected with the phenomena 
 of springa. The ingulfment of streams is a frequent event, 
 
UNDERGROUND RIVERS AND CAVERNS. 191 
 
 botli on tlie large'and small scales, and tlie phenomenon takes 
 place chiefly in limestone districts, where the percolation of 
 water charged with carbonic acid dissolves away the lime 
 along the lines of fissures. Fractures of strata undoubtedly 
 may, probably often do, take part in fixing the line; but the 
 solvent power of water is sufficient to do the v/ork without 
 much assistance. Bearing in mind what was said (Art. 151) 
 regarding the rate of flow of water through fissures — that 
 capillarity retarded it so long as the fissure was above the level 
 of saturation, but that below that level the water flowed freely, 
 if not rapidly — the genesis of an underground stream is not 
 difficult to follow. The rivers of the IMorea, v^hich subside 
 into " swallow holes," or Katavothra, to use the Greek name, 
 emerge near the sea in streams of remarkable uniformity, 
 showing that during the eight months' drought which inter- 
 venes between the seasons of rain the subterranean channel 
 is largely supplied from other sources than the swaJlow holes; 
 or that, as is more likely, the reservoir is slowly drained by 
 narrow apertures, the closure of which, by sediment or by 
 earthquake disturbance, might give rise to a lake in the posi- 
 tion of the swallow hole. Similar examples are met with 
 alon^' the north shores of the Mediterranean in the cretaceous 
 limestone, in v^hich the Morean channels are excavated. 
 But the most interesting district is that of Carniola on the 
 Adriatic. There the Timavo rushes at once from the rocks 
 a navigable stream, being fed by underground tributaries 
 from the interior. The caves of Planina, Lueg, and Adels- 
 bcrg are well known, the principal channel of the last being 
 more than a mile and a half in length, so far as it has been 
 followed. The swallow holes or "dolinas" of this district, 
 tlie cavities into which the streams sink, form, according to 
 the patency of the mountain streams, and the size of their 
 channels, empty chasms, wells, or lakes — the Lake of Zirknitz 
 {Lacus Lugens of the Bomans) varying in area from 40 to 
 80 square miles. This complicated network of channels 
 above and below ground finds its counterpart in other regions. 
 The Humrersee of the Harz Mountains is a dolina like that 
 of Zirknitz ; but the great devonian limestone district near 
 Liege is historically and geologically the most important of 
 the cavern systems. Dr. Schmerling examined forty caverns, 
 
192 PHYSICAL GEOGRAPHY. 
 
 among tliem the now celebrated Engis and Engihoul Caves, 
 and these, with the series found in the Dordogne and Perigord, 
 have furnished most important evidence as to the former 
 fauna of Southern and Central Europe, as to its arctic 
 character, and the association of man with animal forms now 
 extinct or restricted to the far north of Europe and Asia. 
 In England — where, as in Yorkshire, ingulfed streams are 
 also found, important tributaries issuing directly from the 
 rocks bordering the main river — caverns are also met with 
 whose contents are imjDortant contributions to the early 
 liistory of man. The caves of Settle, Ingleborongh, Kirkdale, 
 Denbighshire, of Wokey Hole, Kent's Hole, Brixham, may 
 be mentioned as among those from which remains have been 
 obtained, proving their human occupation at one and, for the 
 most part, at several periods between the time when the 
 reindeer, mammoth, and rhinoceros flourished and the time 
 when men sheltered in them from the Hanoverian troops, 
 or hid therein the smuggled stores which were to cheat the 
 king's revenue. The papers of Euckland, Pengelly, Boyd 
 Dawkins, and others, have given details of those caves 
 hitherto discovered, but their number will be greatly increased 
 as chance reveals openings which debris has concealed. The 
 miammoth Cave of Kentucky, whose main passage is 10 miles 
 long, while the length of the lateral passages amounts to 240 
 miles, contains a curious assemblage of animals, the imperfec- 
 tion of whose organs of sight raises interesting questions for 
 the zoologist. A long time was, of course, needed for the 
 erosion of such enormous subterranean caves, but there is no 
 ground for believing that the inhabitants of the Mammoth, 
 the Adelsberg, or the other Carniolian caves have an earlier 
 date than the latest tertiary times. 
 
 169. Contents of Caves. — The floors of caves are covered 
 with stalagmite, fine mud more or less calcareous, and breccia, 
 consisting of rounded as well as angular fragments, among 
 which, as well as through the other sediments also, bones and 
 implements may be scattered. One or all of these materials 
 may be present, and each layer may contain organic remains 
 belonging to a diflerent period, though disturbance of the 
 layers may confuse the chronology. In observing such da- 
 posits, it is important to bear in mind that the sources of the 
 
LANDSLIPS. 193 
 
 bones may be various; the occupants of tbe cave dying, 
 would leave their remains to be covered over by calcareous 
 droppings from the roof, or by the sediments of the stream 
 which still flowed through the cave. An alteration of the 
 level of the cave mouth, or a rise of the stream on whose 
 banks it opened, might allow the flood-borne bodies of ani- 
 mals to be swept in and left there. Fissures might allow 
 the bones of animals to drop in from the surface of the 
 ground, and to become mingled with the older species. Lastly, 
 the Morean rivers show that bones and conglomerate may 
 enter simultaneously, and the cavern will then yield evi- 
 dence as to the animals which flourished on the surface of 
 the country at periods which cannot be defined. 
 
 The hematite deposits of Cumberland are laid down by 
 rivers which have flowed through limestone caverns, tho 
 segregation of the iron being due to the action of decaying 
 vegetable matter. 
 
 170. Landslips. — Fissures in loose soil or solid rock are 
 traversed by water which may displace more or less of tho 
 surface by its excess in a rainy season, by its expansion on 
 conversion into ice, or by the coincidence of either of theso 
 with underground erosion, whereby a part of a hill slope or 
 clift may be made to topple over. The amount and form of 
 the slipped mass depends on a variety of circumstances; but 
 it may be said, in general terms, that those slips which, are 
 due to underground erosion have steeper faces than where 
 only surface joints are concerned; and that the harder the 
 rock, and more definite its system of jointing, the more will 
 the characteristic features of such a rock be preserved after 
 the catastrophe. The oblique joints of masses of tough 
 boulder clay furnish an admirable study in the mechanism of 
 springs. 
 
 The walls of a valley which a glacier has fashioned describe 
 an elegant and even curve. Those of a valley in which atmo- 
 spheric waste has been at work in soft, superficial deposits, 
 consist of two parts — the steeper, rockier part, and the more 
 sloping, softer portion; but, where landslips have occurred, 
 the ascent is by a series of lines at difiei-ent angles, to the 
 more vertical rocky portion. The landslip leaves its mark iu 
 a change of slope. 
 
 23 N 
 
CKAPTEK V. 
 
 SECTION I.— FORMS OF WATER IN ATMOSPHERE. 
 
 Atmosplieric Circulation of "Water — Atmospliere every^^'here Humid 
 — Aqueous Vapour: Evaporation — Amount of Evaporation from 
 Soils and Plants — Condensation — Saturation — Dew, Mists, 
 and Fogs — Fogs of Cities — Height of Fogs: Fogbanners of 
 Hills — Clouds — Condition of Water in Clouds — Velocity of 
 Clouds — Distinction between Dew and Fog — Rainbow — Colour 
 of Clouds — Transport of Aqueous Vapour — Conditions of Rain- 
 fall—Relation of Rain to Prevalent Winds — Influence of Higli 
 Grounds: Ilomomorpliism— Wliy Rainfall is not Incessant — 
 Dryness of Interior of Continents — Angle of Wind to Land — ■ 
 Polar and Equatorial Winds — Influence of Vegetation — Rain 
 from Clear Sky — Amount of Rain wbich Flows off the Surface 
 ■ — Periodic, Variable, and Constant Rains — Periodicity of Raiu- 
 faU— Table of Rainfall. 
 
 171. Circulation of Water in Atmosphere. — Hain is the 
 form in which, over the largest portion of the world, mois- 
 ture, is restored to the earth, after it has been lifted into the 
 atmosjjhere from land and sea. In the extreme north and 
 south, precipitation takes place only in the form of snow. In 
 the temperate regions v/e find seasonal differences, so that snow 
 and rain alternate, while on either side of the equator snow is 
 never encountered except as a precipitation upon the summits 
 of the highest mountain ranges. If we were to connect by 
 a curved line the limits of snowfall in the northern and 
 southern hemispheres, with tlie limit of snowfall ujDon the 
 mountains of the equator, we should find that the figure 
 thus described v/ould be that of an oblate spheroid, the poles 
 of which would coincide generally with the poles of the 
 earth; thus a shell of temperature above the freezing point 
 would surround the earth, not parallel to its surface, but 
 rather exaggerating the equatorial protuberance and polar 
 flattening of the globe. 
 
AQUEOUS vapour: evaporation^. 195 
 
 172. Atmosphere everywhere Humid. — A certain amount 
 of moisture is present in the atmosphere even in the hottest 
 regions, with the exception of a few very limited areas known 
 as rainless, and even there an absolutely dry atmosphere 
 probably does not extend to any great distance above the 
 surface. In the Arctic regions themselves, part of the snow- 
 fall is simply the restoration, to the surface, of moisture 
 derived from it. 
 
 173. Aqueous Vapour: Evaporation. — Atmospheric mois- 
 ture is in the form of aqueous vapour. Evaporation means 
 the removal of water into the atmosphere in this invisible 
 form, and as the vapour seldom remains in contact with the 
 sui'face from which it has risen, there is constantly ojopor- 
 tunity for the formation of fresh vapour, the volume of the 
 water correspondingly diminishing. Evaporation is limited 
 by diminution of temperature, or by the confinement of the 
 space into which the vajDour is thrown off. The pressure of 
 a very small quantity of aqueous vapour upon the surface of 
 v/ater checks evaporation, and thus fmnishes a difficulty not 
 yet solved, since a very considerable atmospheric pressure 
 does not interfere with the process. In cold weather, at the 
 temperature of 0°C. (32° F.), the pressure is '1811 inches 
 of mercury; e.t 20° C. it is '6850 inches; at 50° C. it is 
 3-622 inches. Trifling as are these pressures, they are enough 
 to arrest the process, though a much greater atmospheric 
 pressure has little or no effect in retarding it. The same 
 effect is produced when evai:»oration takes place into a con- 
 fined space, the intervals between the particles of air being 
 occupied by particles of vapour; in other words, the resist- 
 ance of the vapour being purely mechanical, the atmosphere 
 becomes very speedily saturated, and the water thereafter 
 ceases to diminish in volume, its particles being impeded in 
 their ascent. By the process of evaporation the temperature 
 of water is diminished. Thus the cooling effect of a breeze 
 on the body does not in reality result from the contact of 
 ail', but is due to the removal of moistnro, to the hastening 
 of evaporation from the surface of the skin. Applying this 
 upon a most extended scale, we find that to evaporation, or 
 rather to the great extent of surface from which CA^aporation 
 took place in former times, was due that extreme amount of 
 
196 PHYSICAL GEOGRAPHY. 
 
 cold wliich rendered the winters of this area more severe, 
 and which prolonged the glacial cold into, geologically speak- 
 ing, recent times. 
 
 174. Amount of Evaporation from Soils and Plants. — 
 It appears from Professor Elliot's experiments, that the rate 
 of evaporation from different soils, depends, in the first 
 instance, on the extent of surface their particles present; 
 and, next, on the compactness of their structure, which in- 
 creases or diminishes the capillary flow of water from below 
 upwards. While sand lost -^ of its moisture in a given 
 time, clay lost -f, and peat moss -|; the amount has thus a 
 direct ratio to the incoherence of the materials. Moreover, 
 the capillarity of earth is greater than that of moss, hence 
 the evaporation continues longer after the surface has become 
 somewhat dry. The observations of Von Pettenkofer on a 
 growing oak tree show that there is an increase of evapora- 
 tion from May till July, a decrease thereafter till October; 
 and that the amount of evaporation is 8*33 times greater 
 than that of the rainfall. Hence it follows that a consider- 
 able amount of moisture is poured into the atmosphere by 
 vegetation, and that the water is drawn through the roots of 
 the plants from the subsoil. Vegetation thus not merely 
 retards the surface evaporation, but also restores the mois- 
 ture which some months before had sunk into the deeper 
 parts of the soil. 
 
 175. Condensation. — During evaporation, heat is inces- 
 santly communicated to the vapour, or, in other words, the 
 separation of the particles which constitute the vapour stores 
 up force. When the heat is withdrawn, when the force with 
 wliich the particles are kept asunder is overcome, the particles 
 of the water return into contact with each other, liquefaction 
 taking place, or, as it is called, condensation, and the elastic 
 force of the vapour manifests itself as heat. When water 
 is boiled, if the steam is brought in contact with the skin, 
 or with cold water, it is condensed, and the heat becomes 
 apparent by scalding the skin, or by boiling the water into 
 which it is introduced. 
 
 176. Saturation of Air. — The saturation point of the 
 atmosphere is that point at which it ceases to be capable of 
 containing more vapour o^ water. If the atmosphere i-emains 
 
DEW. 197 
 
 at tlie same temperature, there is no apparent cliange, even 
 when the air can receive no more moisture. But if, by lowering 
 the temperature, the elastic force of the vapour is diminished; 
 if, that is to say, the force with which the particles repel each 
 other is lessened, condensation takes place; and the point at 
 which this occurs is known as the dew point. -♦-' 
 
 177. Dew. — For the formation of dew, three things are 
 necessary: air saturated with moisture, a clear sky, and 
 depression of temperature. Any object on the ground which 
 is a rapid radiator parts with its heat into space, and, chilling 
 the ail' in immediate contact with it, causes the precipitation 
 of the moisture in that portion of air. As the phenomenon 
 is thus due to radiation, it will be affected by anything which 
 modifies that process. In the case of grass, the blades part 
 with their heat, while the moisture in the lower part of their 
 stems prevents its replacement by radiation, as the feeble con- 
 ducting power of the vegetable prevents its transmission in 
 that way. The passage of clouds across the sky, diminishing the 
 amount of radiation by reflecting the heat, stops the process, 
 as does shelter of any kind, the thermometer beneath Dr. 
 Wells' experimental sheds being 1'8°C. higher than one out- 
 side. The necessity for calmness of the air arises from this, that 
 circulation would equalize the loss by radiation, and prevent 
 any one portion from sinking sufficiently low. The tempera- 
 ture of the grass is often as much as 10° to 18°C. below that 
 of the air a few feet above it; and this difference is explaiued 
 by a convective movement, whereby successive layers become 
 lowered so as to m^ntain an interval of 2°C. between their 
 temperature and that of the surrounding air. Thus, \inder 
 favourable circumstances, the dei:)Osit of dew would go on 
 steadily, and the injurious effects of low temperature be from 
 hour to hour increased. Hence the necessity of guarding 
 tender plants for some distance above groimd from the effects 
 of excessive radiation. In tropical countries, the artificial 
 formation of ice in shallow pans, kej^t off the ground by dry 
 straw, takes place under the same conditions which determine 
 the fiill of dew, and the dry straw, bemg a bad conductor, 
 prevents the transmission of heat from the soil to replace 
 that lost by radiation; for the process is arrested when the 
 straw becomes wet, and is converted into a good conductor. 
 
19& PHYSICAL GEOGRAPHY. 
 
 It belongs to meteorology fco determine tlie precise tempera- 
 ture at wliicli dew is deposited; it is sufficient here to indicate 
 that the temperature of the dew point varies with that of the 
 atmosphere 
 
 178. Mist: Fogs. — The formation of dew only takes placo 
 when the aqueous vapour is invisible; but fogs and mists are 
 visible under other conditions. Inequality of surface and 
 movement of the air are, together or apart, essential to bring 
 into contact masses of air, one or both of which are charged 
 with moisture. In a valley in which dew formation might go 
 on if the superincumbent air remained clear, the cold air on 
 the slojDOS gradually moving downwards turns the balance. 
 It was stated that the atmosphere above the dewy surface is 
 often considerably below the dew point. If other cold, moist 
 air is brought in contact with it, further condensation at once 
 takes pla.ce, and fogs are formed such as may be seen creej^ing 
 from either side of a valley till the v,^hole flat ground is 
 covered. The colour of these fogs varies from the delicate 
 blue, almost transparent^ veil, to the dense white mass which 
 saturates the clothes of the traveller. In a very wide plain, 
 like that of Biggar, between the Clyde and Tweed, such a 
 dense mist forms a sea-like surface about three feet deep, above 
 which mounds project like islands, and on these dew may be 
 formed within a few feet above the fog. 
 
 The position of lakes and rivers is marked by fogs whenever 
 a considerable difference of temperature exists between the' 
 air over them and that on the banks. It is imm.aterial whether 
 the land or river atmosphere is the coldei'; whether it is the 
 glacier stream v;hich traverses the warmer low grounds, or 
 the river which flov/s from warm high grounds to cooler 
 plains. Conversely, when a long strip of land projects into 
 the sea, the inequality of temperatm*e gives rise to fog, which 
 may fringe the bank of land or entii'ely obscure it. The 
 British Islands repeat this phenomenon on a large scale, with 
 this addition, that the seas they separate are of unequal 
 temperature ; in winter and spring, when this difference 
 is at its maximum, fogs prevail. 
 
 The direct contact of two masses of warm and cold air is 
 Illustrated off the Newfoundland banks, the Labrador current 
 and the Gulf Stream being accompanied by humid atmo- 
 
HEIGHT OP fogs: FOGBANNEES OP HILLS. 199 
 
 splieres whose meeting gives rise to almost constant fogs; and 
 the difference of teuiperature is as great farther to the sonth, 
 the waters of the Gulf Stream being proportionally Y>-armer 
 than those of the adjacent sea. In the Pacific Ocean, where 
 a coral bank lies across a current of warm water, the boil up 
 and flow over of the interrupted warm stream maintains a 
 constant temperature in excess of that proper to the latitude, 
 and a correspondingly constant fogbank and rainfall. High 
 grounds at right angles to the prevailing v/inds, more espe- 
 cially where these travel over a considerable ocean area, are 
 for the . same reason the seat of fogs, the coasts of Norway 
 and Peru being notable examj^les. 
 
 179. Fogs of Cities. — London fogs are good examples of 
 interference with natural processes. The artificial heat of the 
 city, and the smoke which, though scarcely obvious to the 
 resident, marks its position at a great distance, combine with 
 the humidity consequent on its proximity to the river to make 
 the coldest months periods of long-continued and dense fogs. 
 Professor J. Thomson has described similar phenomena at 
 Belfast. In all large cities dew and hoar-frost are infrequent 
 as compared with equal a,reas in smaller tovms, or in the 
 country; and the reason of the difference has been put to 
 practical use by the vine-growers, who maintain smoky fires 
 on clear nights to windward of their vineyards, the artificial 
 clouds checking radiation. 
 
 180. Height of Fogs : Fogbanners of Hills. — It is diffi- 
 cult to fix the upper limits of fogs. In general, Avhen of local 
 origin, they form a layer which conforms to the undulations 
 of the surface. The fogs which come with the east winds of 
 spring on the shores of Britain are probably of considerable 
 depth. The flat summit of the Campsie hills is the last 
 resting place of these fogs towards the west, and when the low 
 grounds are clear, the fog mass gradually declines from a 
 vertical thickness of about 800 feet to a thin layer, which 
 ultimately disappears. As the height of the range averages 
 1100 feet above the sea, this would give about 2000 feet, or 
 less than half a mile, as the upper surface of the fog. There 
 is, one may almost say mth certainty, an interval between 
 fog and cloud. The connecting link between the two is 
 the fogbanner, which sometimes hangs over the ridgo 
 
200 PHYSICAL GEOGRAPHY*; 
 
 dividing two valleys of unequal slope, as may be seen every 
 autumn in tlie highlands of south and north Scotland; some- 
 times streaming out from a prominent peak, changing its 
 corm but not shifting its place. These furnish only another 
 phase of what has been already described as the consequence of 
 the contact of two masses of moist air at unequal tempera- 
 tures. The chill mountain top condenses the atmospheric 
 moisture to leeward of its peak, and as the chilling is 
 increased by the wind the streamer maintains its place, its 
 extremity, however, being dissipated by the wind. The mass 
 is, in fact, constantly regenerated at the peak as it is wasted 
 to leeward. But if the air falls in temperature, the condensa- 
 tion extends till at last a sheet of fog cloud covers the adjacent 
 summits, and creeps downwards over the slopes. But before 
 this general covering is developed, fog masses may become 
 detached and float away on the wind; these "packmen," as 
 they are called in south Scotland, being solitary travellers 
 which surely foretell rainfall. In all hilly countries the 
 evening fog masses take the direction, and often follow 
 closely the form, of the ridges below, and they may be seen 
 by watching, to grow from above downwards till they rest 
 on and finally cap the hills, the clear interval between them 
 and the hills being occupied by moist air, which the slowly- 
 moving breezes (for this is only seen on quiet evenings) 
 gradually condense. These clouds are, therefore, due to 
 terrestrial radiation. 
 
 181. Clouds. — The distinction is not always attended to 
 between these hill fogs due to terrestrial radiation, and those 
 clouds, properly so called, which result from a cooling process 
 that commences in the upper regions of the air itself. Clouds 
 have been made the subject of very various classifications, 
 but the simplest is that which divides them into three prim- 
 ary groups : — 
 
 1. Stratus, the horizontal layers due to the cooling, by radiation, 
 
 of a mass of air hi situ. 
 
 2. Cumulus, the massive foam-like clouds which form the sum- 
 
 mits of ascending columns of moist air, condensation com- 
 mencing with a loss of electric tension. 
 
 3. Cirrus, which owes its existence and its light curd-like form 
 
 to the contact and cooling of two masses of air. 
 
 Between these leading forms an endless variety of inter- 
 
 1 
 
VELOCITY OF CLOUDS. ^01 
 
 ]3fiecllate steps may be recognised. The stratus, vrliich in 
 Britain is well called the cloud of night, is the dominant 
 form in the central plains of Germany. It is the lowest of 
 the cloud masses, and represents the cooling by radiation of 
 moist air which very slowly ascends. As the sun's rays 
 decrease in power with their obliquity, the aqueous va]Dour 
 is less and less dissipated; its tension diminishes as the 
 temperature falls, and radiation gradually condenses the 
 upper layer from the east towards the west. The cumulus 
 again is the cooling of a mass of moist air which is, so to 
 speak, poured into the colder upper atmosphere, its form 
 being, as Saussure first suggested, exactly comparable to that 
 of a coloured fluid poured into clear water, and the analogy 
 is often very close when the cumulus masses seem to roll 
 over each other upwards. The cirrus belongs to the highest 
 of the cloud-bearing regions, and its form and movements 
 are a sure index of the direction of the wind within the next 
 few hours at the surface. For the details of the form and 
 significance of clouds, the student must consult special treatises 
 on meteorology, Buchan giving in his Handbook an excel- 
 lent summary of the leading points. 
 
 182. Condition of Water in Clouds. — It has been sup- 
 posed that the cumulus is a frozen mass; but the diificulty, 
 already sufficiently great, of understanding how clouds are 
 supported in the air, is thereby needlessly increased. The 
 movements of masses of cloud, even of the seemingly fixed 
 cumulus of summer, shows that if they approach solidity it 
 can only be by their assuming the state of snow. The edges 
 of the masses are irregular; they are constantly changing, 
 and when precipitation takes place the sudden change of 
 form in the mass immediately above indicates a very great 
 amount of mobility. If they were in the state of ice, we 
 should expect that evaporation imder the sun's heat would 
 coat them with a fog layer, which is never the case. 
 
 183. Velocity of Clouds. — It is probable that the clouds 
 move much more rapidly than do the lower strata of the 
 atmosphere, even when both are travelling in the same 
 direction. Buchan has observed a velocity at the rate of 72 
 miles an hour, and quotes Mr. Stephens as having, from 
 twenty observations, calculated 109 miles as the rate. If the 
 
202 PHYSICAL GEOGRAPHIC. 
 
 higliest cirri, the isolated masses of which are most easily 
 made subjects of observation by following their often sharply- 
 defined shadows on the gTonnd, are at 10 miles elevation, 
 even if they are at 5 miles height, theii^ apparent motion may 
 be very much less than their real motion; but the high velo- 
 cities above quoted have an important bearing on the move- 
 ments of the upper air currents, as demonstrating in them a 
 to and fro movement, depending probably on the radiation of 
 vapour masses; just as slight surface movements are seen on 
 the surface of the Gulf Stream. 
 
 184. Distinction between Dew and Fo^. — The clearness 
 of the atmosphere above the surface on which dew is con- 
 densed is the consequence of the unequal rate at which the 
 tension of the vajjour and of the air diminishes. Before the 
 dew point in the air itself can be reached, the gas must have 
 fallen to the same temperature as the vapour. The formation 
 of fog commences when this happens, as a consequence of the 
 mixing of tAVO masses of air. Clouds begin to appear when the 
 cooling which accompanies expansion reduces the air to the 
 same temperature as the vapour. Neither dew nor fog forms 
 on the thermometer suspended above the ground, the air 
 cooled by its radiation sinking down to the ground as fresh 
 portions take its place, just as the cool air flows down the hill 
 side into the valley. 
 
 In the preceding articles an attempt has been made to 
 separate the two kinds of condensation, the one giving rise to 
 dew and fogs in consequence of the cooling by radiation of 
 the earth's surface, and of the air in contact with it; the 
 other giving rise to clouds, properly so called, which result 
 from radiation of vapour suspended in the air, whose tem- 
 perature is lowered along with its rarefaction. :."-".:" 
 
 185. Rainbow: Colour of Clouds. — The spectrum pro- 
 duced by refraction in the drops of condensed vapour is 
 usually double, the inner and the outer having the reds 
 adjacent, the violets being at the extremes. The form of 
 the bow is that of a semicircle when the sun is on .the horizon, 
 but the a-rc becomes less the higher the sun rises, and at 45° 
 the bow is not formed. The occurrence of rainbows against a 
 clear sky, though not frequent, proves the possibility of a con- 
 siderable amount of condensation taking place in the form of 
 
 . 1 
 
TitANSPORT OF AQUEOUS VAPOUR. 203 
 
 / 
 
 a thill layer not thick enough to obscure objects behind it. 
 The refractive power of the atmosphere gives rise to in- 
 cr'^asing chromatic changes towards sunset. As the sun de- 
 chnes westward the amount of the atmospheric layer which 
 his rays traverse increases : at noon the vertical thickness is 
 pierced, towards sunset the oblique, almost horizontal, rays 
 undergo greater diffusion and refraction. The absorptive 
 power of the atmospheric vapour depends on its quantity 
 and form, according as it is in the finest state of division, 
 or its particles are aggregated into sj)^^^^6S. The succession 
 of tints at dawn and sunset is due to the changing absorptive 
 power relative to each tint; but this, as well as the details 
 of the refraction and reflection within drops of water, belongs 
 to the department of Physics. Equally beyond the scope of 
 this volume is the discussion of Mock Suns or Parhelia, 
 Mock Moons or Paraselen^e, Halos, Coronas, or Broughs as 
 they are called in Scotland, and Lunar rainbows. These are 
 all due to refraction and reflection on and in masses of 
 vapour; but whether that vapour is in the gaseous, or the 
 vesicular state, or is crystallized, is not certain. In leaving 
 this subject it may be added that the electric and magnetic 
 states of the atmosphere are stiH comparatively unknown, 
 and that thus influences may be at work whose action and 
 power we cannot at present estimate. 
 
 186. Transport of Aqueous Vapour. — The moisture lifted 
 from land and sea does not remain where it has been gathered, 
 but is carried away by the currents of air, which will be 
 described in the next section, and distributed over large 
 areas. In those regions characterised by the prevalence of 
 steady winds, we have this conveyance of moisture regular 
 in certain ' directions, and thus we have a constant circula- 
 tion or transfer of water from one region to another, the 
 result of which is the maintenance of equilibrium, the circle 
 being completed by the return of water to sea and land, to 
 make up for the loss by evaporation. Condensation of 
 atmospheric moisture has already been spoken ofl", but as yet 
 the quantities thus restored to the ground have been small, 
 though since they are incessantly being lifted up and laid 
 down, they are important geological agents, as well as essen- 
 tial to the Avellbeing of plants and animals. 
 
204 PHYSICAL GEOGRAPHIC, 
 
 187. Conditions of Rainfall. — The conditions uncler wliich 
 rain is precipitated, are stated by M. E. Renou, as quoted by 
 Buclian."^ 1. Two layers of cloud at least: an upper layer, 
 the cirrus, which, being at a great height, is composed of 
 minute ice particles at a very low temperature, probably not 
 higher than -40°C. ; and a lower layer, the cumulus or 
 cumulo-stratus, which has its density increased and its tem- 
 perature diminished by the descent of the ice crystals of the 
 cirrus. 2. The temperature of the air at the earth's surface 
 as high as possible. 3. The atmospheric pressure notably 
 lower than in surrounding regions. 4. Regular horizontal 
 currents of air allomng the atmosphere to remain a sufficiently 
 long time in a state of unstable equilibrium. 5. A rapid 
 movement of the air tending to re-establish the equilibrium 
 of pressure and temperature, by mixmg together the different 
 layers of the atmosphere. 
 
 188. Rainfall Greatest near Ground„ — A curious imper- 
 fection of rain gauges is illustrated in the following tables, 
 constructed by Colonel Ward from observations extending 
 over four years, 1864-7, and Mr. Chrimes, during 18G6-7. 
 The amount of rainfall is relative. 
 
 ETeight above Belative Rainfalt,. 
 
 Ground. 'Ward. Clirimes 
 
 0. 107 
 
 2 inclies 1'05 
 
 6 „ 101 
 
 12 „ 100 1-00 
 
 24 „ -99 
 
 36 „ -98 
 
 60 „ -96 -94 
 
 120 ,, -95 -91 
 
 180 ,, -90 
 
 240 „ -94 -89 
 
 800 „ -88 
 
 The explanation of the anomaly, which must be borne in 
 mind if exact comparison is attempted of observations at 
 many different localities, is, that the rain drops enlarge by 
 attracting vapour particles as they approach the ground; 
 that, especially in heavy rains, the water rebounds from the 
 ground and forms a fine spray over it; that tlie gauge causes 
 
 * Ilandhoolc, p. 18G. 
 
INFLUENCE OP HIGH GROUNDS: HOMOMOJIPHISM. 205 
 
 eddies in tlie air, and thus becomes the centre of convergence 
 for a large amount of water particles. 
 
 189. fielation of Eain to Prevalent Winds. — Obviously 
 the rainfall and the prevalent winds of a country go together, 
 the direction whence the rain conies being that of the winds. 
 In the British Islands the most abundant rains are found 
 upon the west coast, the westerly and south-westerly winds 
 crossing the Atlantic, and in their way becoming charged 
 with moisture. As the temperature over a great expanse of 
 water is more uniform than that on land, the transfer of a 
 volume of humid atmosphere from the warmer ocean to the 
 colder land area results in precipitation. Other illustrations 
 will be found in the rainfall of Southern India, which comes 
 with the S.W. monsoon; in that of the Peruvian coast, and 
 others. But in these cases the features of the land have 
 something to do Avith the amount of rainfall. 
 
 190. Influence of High Grounds: Homomorphism. — 
 Homomorphism, already referred to in a previous chapter, is 
 vv^cll illustrated by the distribution of the rainfall. On the 
 coast of Norway the mean yearly rainfall is 82*12 inches; 
 at Portree, in Skye, 12|- inches fell in thirteen hours in 
 December, 1863. At Coimbra, 118 inches of annual rainfall 
 is recorded, and the quantity diminishes as we pass onwards, 
 the greatest amount being upon the western side of such 
 mountain ranges as project prominently from the plains. In 
 America, on the western side, 89-9 inches are recorded at 
 Sitka; 65 on the west side of Vancouver's Island; 45 at 
 Fort Vancouver on the Columbia river; and only 5 over a 
 large part of the great inland basin. In South America the 
 same excess is found upon the western coast ; the same 
 increase as we advance from the equator southwards, and 
 the same diminution of rainfall upon the eastern portion of 
 the land. In Asia the greatest amount of rainfall is at the 
 foot of the outstanding high grounds : thus the Malabar coast 
 intercepts a considerable quantity of rain; the Himalayas 
 arrest an enormous quantity, which is returned to the plains 
 of Bengal. But the total quantity is greatest, and tlie sift- 
 ing process exerted by hills on moist air is clearest, in the 
 Khasia district. At Darjeeling, the rainfixll between June 
 and September amountcci to 120 inches in one year, while a3 
 
206 PHYSICAL GEOGRAPHY. 
 
 mucli as 264 inclies fell in August, 1841, and" the total 
 annual fall has been known to exceed 600 inches. "When it 
 is remembered that by the rainfall of a district is meant a 
 layer of water which would cover the district uniformly, 
 supposing it neither to run away nor to become absorbed, 
 nor to evaporate, the meaning of these measurements will be 
 intelligible by stating this last-mentioned quantity as a layer 
 60 feet in thickness covering a district. And as 1 inch of 
 rain corresponds to 100 tons of water per acre, the quantity 
 of change of the surface which may be credited to rainfall 
 alone is very large. From all these cases it would appear 
 that mountains have a powerful effect in causing precipita- 
 tion; on the one hand, by arresting the air in its movement, 
 and thus subjecting it to a certain amount of cooling and 
 compression, the result of which is downfall; on the other 
 hand, the air thus arrested rushes upwards, being forced 
 from behind, and passes into a cooler, more rarefied stratum, 
 where precipitation at once takes place. But that this latter 
 ascent of the air is more important than mere cooling by 
 contact, is shown by cases in all hill districts, where, if the 
 hill is low, the rainfall is on the lee side of it. 
 
 191. Number of Rainy Days. — In temperate regions it is 
 difficult to say what constitutes a rainy day: -01 inch in 
 twenty-four hours is that suggested by Symons, and made 
 the basis of Buchan's table. ^'' 
 
 Latitude, 
 
 60°— 50" 
 
 161 dcays per annum. 
 
 }} 
 
 50°— 46'' 
 
 134 
 
 }) 
 
 46°— 43° 
 
 103 
 
 j> 
 
 43°— 12° 
 
 78 
 
 This table is quoted for the sake of the general relations it 
 suggests; but the total quantity of rainfall may be small or 
 great in proportion to the estimated number of rainy days. 
 
 192. Why Rainfall is not Incessant. — The simple fact of 
 the air containing a large amount of moisture does not 
 necessarily involve its downfall, else the trade winds, which 
 are always heavily charged with moisture, would coincide 
 with regions of well-nigh constant rain. The moA^ement of 
 the atmosphere being constant and uninterrupted enables 
 the air to carry its burden, the capacity for moisture being 
 * Handbook, p. 191, 
 
INFLUENCE OF VEGETATION^ 207 
 
 increased tlie farther it travels. But where, as iii what 
 might be called the atmospheric backwater of the Doldrums, 
 the movement of the winds is arrested, precipitation is seen 
 on an enormous scale. 
 
 193. Dryness of Interior of Continents. — This stoppage 
 of the wind and constant draining of the atmosphere by con- 
 tact with elevated ground explains the dryness of the interior 
 of continents. Immediately to the north of the Himalayan 
 chain we have the dry table-lands of Central Asia : the 
 Kocky Mountains border a region in which only 5 inches 
 fall yearly, although a little way off the annual average is as 
 much as 89 inches. Upon the west coast of the British 
 Islands the mean is about 40 inches, rising at some places 
 to 70 inches; upon the east coast 25 inches forms the average. 
 The mean for Bussia in Europe is 15, and the effect of the 
 Scandinavian chain of mountains is seen in the 20 inches 
 recorded for Sweden as contrasted with the 82 inches of the 
 Norwegian coast. The same holds true for South America, 
 Western Patagonia being tolerably wet, wliile Eastern Pata- 
 gonia suffers at times from excessive and protracted droughts. 
 The central areas of Australia and Africa owe their dryness 
 to the same cause. 
 
 194. Influence of Angle of Wind. — ^\Yhile it is true in 
 general terms that the points of greatest rainfall are upon 
 the windward side of high lands, on which prevailing winds 
 blow, a slight modification is traceable to the angle which 
 the wind forms with the trend of the high ground. The 
 greatest rainfall will be where the wind blows at right 
 angles to the coast ; but the quantity will decrease in pro- 
 portion to the obliquity, so that the arrest of the prevailing 
 wind by the high ground may be very slight, and the rainfall 
 will then be due rather to the friction retarding the margins 
 of the moving current. 
 
 195. Polar and Equatorial Winds. — It is also a sound 
 general proposition that winds blowing from the poles are, 
 as a rule, drier than those blowing from the equator; and as 
 the westerly winds prevail increasingly from the region of 
 the trades towards either pole, the rainfall upon the western 
 shores is naturall}'' greater than on the eastern. 
 
 J96, Influence of Vegetation. — Vegetation has a consider- 
 
208 PHYSICAL GEOGRAPHY. 
 
 able power in affecting tlie rainfall ; and it is now a well 
 established fact that luxuriant forests have a larger amount 
 of precipitation, other things being equal, than other parts 
 of the same region. The cutting down of the timber on the 
 island of Mauritius was a very important one among the 
 influences which suddenly increased the unhealthiness of the 
 island. Sir John Herschel's observation at the Cape of 
 Good Hope illustrates the influence of trees. The fog clouds 
 of Table Mountain frequently hang for some time without 
 any rainfall. But Sir John remarks that in a forest there 
 v/as heavy rain, though outside the air was simply moist. 
 The explanation is very similar to that of dewfall on grass, 
 the extended radiating siu'face of the leaves lowering the 
 temperature and causing precipitation. 
 
 197. Rain from a Clear Sky. — Rainbows have been 
 mentioned as occurring under this anomalous condition. 
 The fine rain, or serein as it is called, is probably due to 
 local refrigeration of the air, caused by the interference of 
 one current with another, the arrest of motion giving a short 
 time of condensation before the air, takes a new or resumes 
 its old course. 
 
 198. Amount of Rain which flows off the Surface. — It 
 has been calculated that one-third or one-fourth of the rain 
 which falls on the surface flows off it, the remainder being 
 absorbed by the soil, or given back by evaporation to the 
 atmosphere. 
 
 199. Periodic, Variable, Constant Rains. — The equa- 
 torial zone of constant rain is that in which the atmospheric 
 currents are most variable, and where at the same time the 
 results of evaporation are most abundant. The frequent 
 changes of direction cause correspondingly frequent conden- 
 sation by arrest of movement. 
 
 The periodic rains of the trades and the monsoons corre- 
 spond to the passage of the sun to north and south of the 
 equator, and though great quantities fall within short periods, 
 wliilo tlic dry season is one of little or no rainfall, these must 
 not be regarded as dry seasons, since the dews are very heavy. 
 
 The variable rains characterise the regions to the north 
 and south of the trades and monsoons, including, therefore, 
 all the temperate and polar regions. 
 
PERIODICITY OP RAINFALL. 
 
 209 
 
 200. Periodicity of Rainfall. — J. JSTorman Lockyer lias 
 tabulated tlie evidence in favour of an eleven years periodi- 
 city of rainfall, coincident with the periodicity sun spots and 
 cyclones. Mr. G. J. Symons gives the following table'*" : — 
 
 Maximum Sun Spot years, 1837 1848 18G0 1871? 
 
 Heavy Rainfall, 1836 1848 1860 1872 
 
 Amount of Eainfall, 33-49 35-98 33-34 ?34 
 
 Per cent, above average, 19 28 18 20 
 
 Minimum Sun Spot years, 
 Small Rainfall, 
 Amount of Rainfall, 
 Per cent, below average, 
 
 TABLE OF RAINFALL. 
 
 England and Wales. ,r' 
 
 1833 
 
 1844 
 
 1856 
 
 1867 
 
 1834 
 
 1844 
 
 1858 
 
 1868 
 
 24-52 
 
 23-72 
 
 22-79 
 
 ?28-8 
 
 13 
 
 16 
 
 19 
 
 4-2 
 
 West Coast. 
 
 South Coast. 
 
 East Coast. 
 
 Cumberland — 
 
 Cornwall, 
 
 22-47 
 
 Norfolk, 21 
 
 Cockermouth, 22 
 
 Devon — 
 
 
 York (High 
 
 Seathwaite, 113 
 
 Sidmouth, 
 
 16-64 
 
 Grounds), 40-50 
 
 Lancashire — 
 
 Dartmoor, 
 
 52-33 
 
 Doncaster, 21 
 
 Manchester, 30 
 
 Plymouth, 
 
 45-100 
 
 Durham — 
 
 Bolton, - 40 
 
 Dorset — 
 
 
 Bishop Wear- 
 
 Coniston, 64 
 
 Abbotsbury, 
 
 18-45 
 
 mouth, 17 
 
 Liverpool, 24-25 
 
 Blandford, 
 
 29 
 
 Northumber- 
 
 Anglesea, 34-5 
 
 Hampshire — 
 
 
 land — 
 
 Caernarvon, 54 
 
 Aldershot, 
 
 16-51 
 
 Newcastle, 24 
 
 Montgomery 
 
 Woolmer 
 
 
 Shields, , 23 
 
 & Merioneth, 54 
 
 Forest, 
 
 26-90 
 
 TKTiAN^T) 
 
 Cardigan, 37 "5 
 
 Sussex — 
 
 
 Staffordshire, 23 
 
 Pembroke, 31-40 
 
 Hastings, 
 
 1818 
 
 Leicester, 19-26 
 
 Caermarthen, 
 
 Chichester, 
 
 32-79 
 
 "Rf>rlfnrr1 1 (» 
 
 Glamorgan, 42 
 
 Kent and 
 
 
 Middlesex — 
 
 Somerset — 
 Taunton, 19-06 
 
 Surrey — 
 Margate, 
 
 16 38 
 
 Hampstead, 16-22 
 Winchmore 
 
 W. Harptree, 36 76 
 
 Cranbrook, 
 
 28-90 
 
 Hill, 23-11 
 Wiltshire — 
 Chippenham, 18-14 
 
 
 
 
 Salisbury, 25-25 
 
 23 
 
 Average of 14 stations on W. Coast, 43*33. 
 12 ' „ S. Coast, 30-26. 
 
 6 „ E. Coast, 25-16. 
 
 7 „ Inland, 2014. 
 
 * Nature, Dec 26, 1872. 
 
 
210 PHYSICAL GEOGRAPHY. 
 
 Scotland. 
 
 West Coast. 
 
 Isles, 19-59 
 
 Mull, 74-5 
 
 East Coast. 
 
 Eife, 18-25 
 
 Midlothian, . . . 16-27 
 
 Haddington, . . . 17-23 
 
 Average of 2 stations on W. Coast, 56-75. 
 „ 3 „ E. Coast, 17-25. 
 
 Ireland. 
 
 Waterford, 39-5 
 
 Sligo, 38-5 
 
 Dublin, 21-75 
 
 W. Germany, .... 20 
 Sweden, 20 
 
 Mr. S3nnons riglitly doubts tlie importance of the coin- 
 cidence v/ith cyclonic periodicity, since greater energy of 
 cyclones can hardly be expected to influence the rainfall 
 over the whole globe. 
 
 But from a very extended comparison of observations, the 
 conclusion seems justified that the maximum and minimum 
 sun spot years have respectively a larger and smaller num- 
 ber of atmospheric disturbances, ranging from 9 to 12 on the 
 table, Art. 310, and that the rainfall likewise varies. But 
 such annual variation cannot take place without correspond- 
 ing variations of temperature, and as this would depend on 
 unequal solar radiation, the coincidence empirically ascer- 
 tained may yet prove to include electric and magnetic dis- 
 turbances, and to refer all to a common cause. 
 
 The follov/ing lines, on the adjoining table, taken from Mr. 
 Symon's Abstract of liawfall, 1832-68, arranged according 
 to sun spot years, will illustrate this special relation, and 
 give a fair comparison of tlie rainfiiU oyer the globe, 
 
TABLE OF RAINFALL. 
 
 211 
 
 1844 
 1848 
 1856 
 18G0 
 
 Year 
 
 ie>. CO >&• to 
 
 00 >?>. C» -I 
 
 o ni o -V 
 
 Guernsey. 
 
 rr.' 
 
 
 CO 
 
 CO to CO to 
 to to O 1*^ 
 
 o to to o 
 
 Greenwich. 
 
 CO to OS CO 
 
 CO -r -r to 
 
 o rf^ cb K' 
 
 Sandwick, Orkney. 
 
 to t^. to CO 
 O rf^ O' o 
 CO O ~-I CO 
 
 Tarn, Bacsin de 
 Saint Ferriol. 
 
 
 
 to CO OS CO 
 CT I-" O tf^ 
 
 CO O O M 
 
 Halite Garonne, 
 Toulouse. 
 
 Ife. >*^ CO Ut 
 
 O CO o to 
 o 00 di o 
 
 Basses Pyrenees, 
 Bages Beost. 
 
 OS to to to 
 --r --T o-i o 
 
 O )ji. rfl C/) 
 
 Conrcon, Cliarente 
 Inferieure. 
 
 to to to to 
 
 t^ V< tr- ~^I 
 
 >^ o « o 
 
 Paris. 
 
 Its- 1^. CO CO 
 t-J l-J CO )(- 
 
 CO o tri tji- 
 
 Geneva. 
 
 
 CT >;^ ci C5 
 
 O >f^ lO o 
 CO t-J l^. I-J 
 
 Great Saint Bernard. 
 
 to to CO 
 ; CO Ci o 
 
 CO rfi- o 
 
 Rome. 
 
 
 o to 
 
 Jerusalem. 
 
 
 O C-T ~1 
 
 ; >(- CO CO 
 
 to -^ «5 
 
 Calcutta. 
 
 &r? 
 
 lO to (^ *" 
 O CO O t-" 
 CJ< -i OT to 
 
 Algiers. 
 
 > 
 
 I-" to to to 
 CO O CO o 
 >f^ CJl M CO 
 
 Oran. 
 
 lO t-l 
 
 OS en ; ; 
 
 Constantino. 
 
 to to 
 
 CO CO ; ; 
 
 o >-^ * 
 
 Toronto. 
 
 ^9 
 
 CO 
 00 
 
 Pluhulelnhia. 
 
 a 
 
 P- 
 
 re 
 
 CO Ci 
 ; ; era O 
 ' ■ * ob 4I 
 
 New York, Fort 
 Cohuubus. 
 
 OS OS 
 
 10 to 
 
 Ma.9sacliusetts. 
 
 eg 1 JLunmto, i n.>^^ 
 rfl 1 N. Gr.inada. \ r a ' 
 
 2 ««>• - 1 
 10 lu 
 
 4^ itl CO C-i 
 
 Carbadoes. 
 
 1?^! 
 
 w 
 o 
 
 > 
 
212 PHYSICAL GEOGRAPHY. 
 
 SECTION II.-SNOW AND ICE. 
 
 Sno"w: Form of its Crystals — Hoar-frost — Snow Flakes — Sleet— Tex- 
 ture and Colour of Snow— Snow as compared with Rain — Limit 
 of Snowfall — Snow Line, or Limit of Perpetual Snow — Height of 
 Snow Line in different Latitudes — Geological Importance of Ice 
 — Temperature and Density of Freezing Water — Density of Salt 
 Water at Freezing Point — Lowering of Freezing Point by Pres- 
 sure — Influence of Forces applied to Ice — Plasticity of Ice : 
 Glacier Motion — Expansion of Frozen Water : its Geological 
 Effects — Ice Formed by Compression of Snow — Genesis of a 
 Glacier — Unequal Movement of Parts of a Glacier — Structure of 
 Glacier Ice — Comparison of Glacier and Lake Ice — Daily Motion 
 of Glaciers — Curves of Glacier Valley : their Influence on Erosion 
 of Valley — Mean Daily Motion : Seasonal Variations — Varia- 
 tion of Movement at Surface — Variation of Movement below 
 Surface — Retardation due to Compression — Difference of River 
 and Glacier — Bifurcation of Glacier — Crevasses : Bergschrund 
 — Dirt Bands — Diminution of Glacier by Superficial and Ter- 
 minal Waste — Dimensions of Glaciers — Diminution of Feeding 
 Ground : Surface Waste — Avalanches — Position of Morainic 
 Detritus on Glacier — Part of the Detritus sinks into the 
 Glacier — Moraines of Deposit — Subglacial Stream : Notch in 
 Terminal Moraine — Glaciers at Sea Level — Striation of Glacier 
 Bed — Characteristic Features of Glaciated District — Extent of 
 Ice over Different Regions — Development of Ice Sheet of N. 
 Hemisphere during Glacial Period — Definition of Glacier and 
 Ice Sheet — Ice Sheet formed by Fusion of Local Glaciers— Lower 
 Boulder Clay : the Moraine Prof onde of the Ice Sheet — Upper 
 Boulder Clay — Relation of these Two Deposits — Erratics — Pack 
 Ice : Ice Foot — Coast Ice — Ground Ice — Iceberg — Ice Floe — 
 Travelling of Icebergs — Geographical Effects of Icebergs : 
 Striation — Hail — Structure of Hailstones — Relation of Hail to 
 Storms. 
 
 201. Snow. — The particles of aqueous vapour in the atmo- 
 sphere are frozen when the temperature falls below 0°C. The 
 crystalline figures then formed are hexameral. The simplest 
 forms are six-sided rods; more complex combinations are 
 offered by stars, the rays of which consist of simple rods, 
 whose extremities are bevelled into six-sided pyramids. The 
 angles of these rays are 60°, and if secondary rods project 
 from them, these also have an angular divergence of 60°. 
 But this regularity is lost sight of in the secondary ornamen- 
 tation when that ceases to be rod-like. Petaloid figures may 
 be reduced to the simple six-rayed type, but it is scarcely 
 
SNOW AS COMPARED WITH RAIIT. 213 
 
 possible to trace the rectilinear foundation of many of the 
 
 secondaiy patterns. Mr. Glashier's figures of snow crystals 
 
 might easily be mistaken for drawings of the siliceous 
 
 skeletons of radiolarians, or microscopic protozoa, Avhose 
 
 ' homogeneous body substance is associated with frameworks 
 
 of the most exquisite beauty, and marvellous regularity. 
 
 , The crystals sometimes form regular hexagonal plates, which 
 
 I may be deduced from the six-rayed stars by increasing the 
 
 secondary raylets. 
 
 202. Hoar-frost. — Crystals are formed after dewfall if 
 j the temperature continues to sink; but they are less regular, 
 
 and they adhere so as to form, not a continuous layer, but a 
 fur of minute ])yi'amids. 
 
 203. Snow Flakes: Sleet. — Snow flakes are formed by 
 aggregation of the crystals into masses varying from an inch 
 to a quarter of an inch in diameter. Their adhesion is less 
 perfect the lower the temperature; and boys are well aware 
 of the fact that snow which will not work into balls is always 
 small flaked. Sleet appears to be snow flakes partially melted 
 in their descent, and accompanied by moisture condensed on 
 the surface of the irregular masses. 
 
 204. Texture and Colour of Snow. — The variety of surface 
 which the bevelled spicules of a snow crystal ofier to the light, 
 yields an infinite play of prismatic colours, which combine 
 into white ; while the reflection from the crystals in the walls 
 cf cavities, formed by air entangled among the crystals and 
 flakes, contributes to this effecfc in the same way that salt or 
 sulphate of magnesia is whiter in mass than when in a thin 
 layer. The red and green tint of snow witnessed in the Alps 
 is due to the presence oi Protococcus nivalis, a microscopic alga. 
 
 205. Snow as compared with Rain. — Snow is to water as 
 1 to 10 by weight on an average; but the small size of the 
 flakes sometimes diminishes the ratio to 1:8. In general it 
 may be held that 1 inch of snow is equal to '1 inch of water. 
 But the efiects of snow are not to be thus estimated. Whereas 
 evaporation reduces rapidly the temperature of the soil on 
 which '1 inch of water has fallen as rain, an inch of snow 
 checks terrestrial radiation because of its low conductivity, 
 and of the air entangled in its mass; and this same fact like- 
 wise prevents the evaporation from the suiface of the »snow 
 
^14 i?HYSICAL GEOGRAPHY. 
 
 affecting the temperature of the soil. Hence a difference of 
 20° C, or mucli more, may exist between the soil under snow 
 and the air above it. 
 
 206. Limits of Snowfall.— To the south of 30° IST. lat. 
 snow never falls in Europe; but the line which marks its 
 most southerly extension is not a straight one. It is, in fact, 
 the isotherm of 11'1° C, and this, like all other lines of equal 
 temperature, passes into lower latitudes over continents ; into 
 higher latitudes over oceans. In the Atlantic it recedes to 45°, 
 and over N. America descends to 33°ISr. lat.; in the southern 
 hemisphere the limit shows similar but less extensive curves, 
 the flexures northwards corresponding to the southern apices 
 of Australia, Africa, and America. 
 
 207. Snow Line, or Limit of Perpetual Snow. — The 
 snow over the greater part of the area thus marked ofi melts 
 after it falls ; but as we advance towards the poles the length 
 of time during which it lies, that is, remains unmelted, 
 increases, till, in 78°ISr. lat., 54-5°S. la-t., the heat of summer 
 is unable to remove the winter's accumulation, and the snow 
 is there said to be perpetual. While this is the horizontal 
 limit of perpetual snow, the vertical is at an increasing height 
 above sea level, till, at the equator, it is on the Andes of 
 Quito 15,800 feet above the sea. But the lines connecting the 
 equatorial with the polar limits are not regular any more 
 than are the isotherms; in other words, the shell of air at a 
 higher temperature than 1'6°C. is not of luiiform thickness. 
 In the first place, it is nearly at the same height from the 
 equator to 20**^ on either side of it; thence it declines slowly 
 towards the poles. But the area is unequal in the two hemi- 
 spheres, since it reaches 23° farther towards the pole in the 
 north than in the south. Further, its height varies under 
 local conditions : thus, it is 4000 feet lower on the south face 
 of the Himalayas than on the north. On the east side of the 
 Andes it is more than 2000 feet lower than on the west. 
 The quantity of moisture determines the amount of precipi- 
 tation, and it is greater over the plains of Bengal than over 
 the dry, heated Thibetan plateau; greatest over the track of 
 the S.E. trades blowing from the Atlantic; while the steeper, 
 barer slopes towards the north of Asia and towards the 
 Pacific retain less moisture and absorb more radiant heat. 
 
GEOLOGICAL IMrORTANCE OF ICE. 
 
 215 
 
 HEIGHT OF SNOW LINE IN DIFFERENT LATITUDES. 
 
 
 Lat. 
 
 • 
 
 Ileiglit ill Feet. 
 
 Spitzbergen, ..... 
 
 N. 80° 
 
 jE. 
 l^Y. 2500 
 
 Norway, 
 
 71° 
 
 2500 
 
 ••■••* • • . . . • 
 
 70" 
 
 (2900 
 \ 3350 
 
 Sulitelma, .... 
 
 
 3835 
 
 
 G0° 
 
 { C* 4450 
 
 i . . . . . 
 
 (I.f 5500 
 
 Kamtschatka, ..... 
 
 56-30° 
 
 5249 
 
 Oonalaska, Aleutian Isles, 
 
 53-30° 
 
 8510 
 
 Aldan, 
 
 or 
 
 4476 
 
 Altai, 
 
 50° 
 
 7010 
 
 Alps, ..... 
 
 46° 
 
 (N. 8500 
 I S. 8885 
 
 Caucasus, . . 
 
 43° 
 
 11,063 
 
 JWBlY^Vy •••••• 
 
 39-40° 
 
 14,170 
 
 Pyrenees, 
 
 42-75° 
 
 8,950 
 
 Rocky Mountains, .... 
 
 40°-43° 
 
 12,500 
 
 Etna, 
 
 37-75° 
 
 9,500 
 
 
 37° 
 
 11,200 
 
 Himalayas, ..... 
 
 28°-29° 
 
 ^N. 19, 560 
 : S. 15,500 
 
 
 1.3° 
 
 14,000 
 
 Andes of Quito, .... 
 
 0° 
 
 15,800 
 
 Bolivia, .... 
 
 S. 16° 
 
 17,700 
 
 ... ... .... 
 
 18° 
 
 20,000 
 
 ... ... . • . . 
 
 27° 
 
 13,800 
 
 Chili, .... 
 
 33° 
 
 (E. 12,700 
 ( W. 14,700 
 
 ... .... . • * . 
 
 42-30° 
 
 6,010 
 
 
 43° 
 
 6,000 
 
 Mount Cook, New Zealand, . • 
 
 44-25° 
 
 ( S.E. 7,800 
 1 W. 6,900 
 
 Straits of Magellan, .... 
 
 53 30" 
 
 3,707 
 
 South Georgia, 
 
 54-30° 
 
 
 
 * Coast, t Interior. 
 
 208. Geological Importance of Ice. — The geographical dis- 
 tribution of ice, as it is popularly understood, in the shape, 
 that is to say, of solid masses of considerable size, is perhaps, 
 from the geological point of view, the least important of all 
 the facts concerning this form of water. Largo as are the 
 glaciers of tropical and sub-tropical lands, thcii' effects are 
 
216 PHYSICAL GEOGRAPHY. 
 
 trifling in comparison witli the modifications of the features 
 of a country effected at every point of the temperate regions 
 by the conversion of watery vapour into ice. 
 
 209. Temperature and Density of Freezing Water. — 
 Fresh water attains its maximum density at 4*^0.; if the 
 temperature sinks below that point fresh water expands 
 gradually as the temperature falls, till the freezing point is 
 attained, at which there is an abrupt increase of volume 
 caused by abstraction of heat during solidification, but with- 
 out any lowering of temperature. 
 
 The lowering of the temperature is efiected by convection. 
 When the surface layer has reached its extreme density at 
 4°C., it sinks to the bottom, and there is a vertical circula- 
 tion which only comes to an end when, all being of the 
 same density, this vertical motion is no longer possible. If 
 the water now remains still, it may continue liquid even 
 though the temperature sinks considerably below 0°C. But 
 if such lowering takes place, a very slight disturbance will 
 convert the whole mass into ice. But this process is not an 
 indefinite one in water of any depth, and having a consider- 
 able extent of surface. Evaporation accelerates the cool- 
 ing of the surface, and with the continued sinking of the 
 temperature to 0°C., the point of maximum density is passed; 
 thereafter expansion occurs and the chilled water floats. If 
 ice is formed, that also floats, being of less density than the 
 chilled water. The change of behaviour, at a point short of 
 freezing, prevents the fresh Avaters of the globe being frozen 
 throughout their mass. After a cake of ice is formed its 
 increase is slow, the cake, among other effects, checking cool- 
 ing by radiation. 
 
 210. Density of Salt Water. — But the ocean is differ- 
 ently affected by cold. As has been already said (Art. 
 74), salt water continues to contract to its freezing jDoint, 
 -3-G7°C. (25-4° F.) if kept still, -2-6^0. (27-2^ F.) if 
 
 disturbed. 
 
 211. Lowering of Freezing Point. — Professor James 
 Thomson inferred, from the mechanical theory of heat, that 
 the temperature at which water freezes is not a fixed point, 
 but that it must vary with the pressure ai^plied to the water, 
 and comnumicatcd by it to the ice in process of freezing or 
 
fLASTlCITlT OP ICE: MOTION OF GLACIERS. 217 
 
 melting. In fact, he found that tlie freezing point must vary 
 with pressure, just as the boiling point was already well 
 known to do. From experimental data of various kinds he 
 deduced, by theoretical considerations, the result that the 
 freezing point must be lowered by "0075^0. for one additional 
 atmosphere of pressure applied, and twice as much for two, 
 thrice as much for three, and so on for many additional 
 atmosj^heres. This deduction was subsequently confirmed 
 exiDcrimentally by Sir "William Thomson. 
 
 212. Influence of Forces Applied to Ice. — Professor 
 Thomson also deduced afterwards, by other theoretical con- 
 siderations, that any force whatever Avhich tends to alter the 
 form of ice wet Avith ice-cold water, whether these forces 
 apply to the ice pressures or tensions, that is, pushes or pulls, 
 whether they are twisting or cross-bending forces, must 
 impart to the ice a tendency to melt and to give out its cold, 
 which will tend to generate, from the surrounding water, a 
 corresjjonding quantity of ice free from the applied forces. 
 This second result, it is to be observed, is quite distinct from 
 the former one, which related to the lowering of the freezing 
 point by pressure applied to the water and communicated by 
 it to the ice; here the forces are applied to and transmitted 
 through the ice alone, and are not communicated to the water 
 at all. 
 
 213. Plasticity of Ice; Motion of Glaciers. — From these 
 two principles, and especially from the later of them, he has 
 offered a theory to account for the plasticity of ice, as mani- 
 fested by the motion of glaciers down their valleys, past all 
 kinds of obstructions and sinuosities. He has pointed out 
 that, whatever part of the ice may be subject to forces tend- 
 ing to change its form, that i:)art must proceed to melt away, 
 and to give out its cold to the surrounding liquid. Each 
 such melting away, and transfer of forces to newly frozen 
 ice, must entail a change in the general dimensions of the 
 mass of ice as a whole, which will constitute a flow of the 
 glacier down its valley. The yielding by melting entails also 
 successions of fractures, either as small fissures or great 
 'crevasses, which allow a more general rapid movement tlian 
 would occur in virtue of the melting and refreezing alone. 
 The fractured masses reunite when pressed together again in 
 
^1^ PHYSICAL GEOGRAPHY. 
 
 the subsequent progress of tlie glacier. The principles 
 brought forward by Professor James Thomson relative to 
 simultaneous melting and freezing under forces which tend 
 to change the form of the ice, appear to afford an explana- 
 tion of the process of ''regelation," discovered by Faraday, 
 when he turned attention to the fact that two pieces of melt- 
 ing ice will, even in hot summer weather, unite firmly 
 together if left pressing against each other. ^ 
 
 214. Expansion of Frozen Water: its Geological Effects. 
 — Water, when it passes into ice, changes its volume from 1 
 to 1*099. In the well known experiment, water in a corked 
 bottle is frozen, and the bottle bursts ; but if the cork is left out, 
 a plug of ice projects from the neck. As the surface of the 
 earth in temperate regions — and these represent the greater 
 part of the area of the northern hemispheres — is constantly 
 charged with moisture, the freezing and expansion of the 
 water with which the ground is saturated has the effect of 
 loosening its particles, and although no apparent change may 
 be obvious during the frost, when the thaw comes the loose 
 granular condition of the soil is very apparent; nay, we can 
 sometimes even detect an appreciable elevation of the surface 
 to the extent of an inch or two inches above its former level. 
 Such a loosening of the particles prepares them for removal, 
 and thus the agency of frost is one of the most important 
 in atmospheric denudation. In glacier valleys, the surface 
 moraine is derived from the sides of the v?lley, but this 
 debris is very seldom obtained by the undercutting of the 
 cliffs, the unsupported face of which would then tumble, as 
 haj^pens in river vallej^s; the rocky fragments are in reality 
 cast off by the rending action of the ice formed in their 
 interstices, which splits them wedge -like, producing ever 
 fresh surfaces. 
 
 215. Ice Formed by Compression of Snow. — But the 
 
 conversion of Avater directly into ice is only one of the 
 methods of its production. The enormous masses of this 
 
 * J. Thomson. Theoretical Considerations on the Effect of Pressure 
 in Lowering the Freezing Point of Water. Trans. Roy, Soc, Edin., 
 xvi., 1849. On the Plasticity of Ice. Proc. Roy. Soc, viii., 1S5G-7. 
 Pecent Theories and Experiments on Ice at its Melting Point. Ibid., x., 
 1859. On Crystallization and Lujnef action. Ibid., xi., ISGl. 
 
Genesis op A GLACiEit. 219 
 
 inateiial wLicL. constitute a glacier, are obtained by the 
 gradual compression of snow till it lias jDassed from tbe 
 crystalline form into that of solid, transparent ice. The 
 colour of snow is due, as has been said, to the presence 
 between the crystals — entangled amongst the crystals — of 
 air which intensifies the white formed by the blending of the 
 prismatic rays from the crystalline faces; and from the dull 
 opaque appearance of snow to that of clear ice, several 
 transition stages may be observed. The imperfectly consoli- 
 dated snow constitutes neve or firn, the friable mass cf rotten 
 ice of Alpine travellers — found for the most part above the 
 snow line. 
 
 216. Genesis of a Glacier. — The passage of snow into ice 
 takes place in this v>^ay : the snow which falls upon the 
 mountain summit is increased from time to time by fresh 
 precipitation; but in the interval it undergoes diminution by 
 evaporation, which takes place to a very great extent even 
 in the most remote Arctic regions. The supply of snow, 
 however, is in excess of the removal by this process, and thus 
 we have from year to year an increasing residue, which grad- 
 ually rises higher and higher above the surface of deposit. 
 If the area upon which this increasing mass alights sinks 
 to the low grounds by vertical cliffs, accumulation may go 
 on for a considerable time before any of the mass leaves the 
 surface ; but as the depth of snow increases, the pressure 
 upon the lower strata likewise increases, and unless the area 
 is confined, that pressure produces lateral displacement of the 
 lower portion, which then falls over the cliffs. In such an 
 imaginary locality no great amount of ice may be produced, 
 this lateral displacement relieving the vertical pressure. 
 But, in general, the snow alights upon the summits of hills 
 from which gradual slopes descend, and the vertical accumu- 
 lation, pressing vertically upon the lower strata, displaces them 
 downwards along the line of slope, and thus we have at once 
 pressure acting in two directions — vertically through the 
 mass, and parallel to the surface of the slope, causing the 
 extruded portions to descend to lower levels. But as the 
 vertical accumulation goes on steadily increasing, the portion 
 that descends the slope likewise increases; and by the double 
 pressure, vertically and from behind, the snow is gradually 
 
220 PHYSICAL GEOGRAPHY. 
 
 made to pass tliroiigli a neve stage, and to acquire that of 
 pure solid ice. Tliis descending mass is not, liowever, uni- 
 form in composition tliroughoutj tlie surface is covered with 
 fresh fallen snow, and we have, therefore, transitions verti- 
 cally from snow, through neve, into the characteristic glacier 
 ice. The simple push from the feeding ground is speedily 
 exhausted, the motion of the glacier being due to that plas- 
 ticity which was explained in Arts. 212, 213. The moving 
 mass descends the valleys in exactly the same way that a 
 rivulet descends from the summit of a waterparting, and the 
 glacier therefore has a superficial resemblance to a river, but 
 the conditions of its motion are in contrast with those of the 
 flowing water. 
 
 , 217. Unequal Movement of the parts of a Glacier. — But 
 
 the surface of the moving mass exhibits features similar to 
 those on the surface of a river, and we can map the appear- 
 ance in the one case by the movement of the debris which 
 tumbles upon the surface of the ice; in the other case by the 
 foam or drifted material carried forward by the water. In 
 both we have the friction of the moving mass greater at the 
 bottom than at the sides of the stream, and least in the centre 
 at the surface. The motion, therefore, is most rapid at the 
 last-named point, and in consequence we have a series of 
 curves, the convexity of which points down the stream. 
 Hence the detritus in mid-stream arrives soonest at the end 
 of the river, be it of ice or of water. 
 
 218. Structure of Glacier Ice. — Glaciers exhibit a strati- 
 fied appearance, which is due to certain subordinate processes 
 that their materials pass through. In the intervals of snow- 
 fall the sun acts upon the surface, and Avith great intensity 
 at the higher parts of the mountains. It melts the particles 
 somewhat, and forms an imperfectly consolidated layer of 
 greater or less thickness, according as the interval between 
 the successive snow showers is greater or less; the danger of 
 glacier travelling is, therefore, in proportion to the frequency 
 of the snow showers, for the less the interval the softer is the 
 surface. The vertical pressure, already spoken of, thus 
 operates not upon snow alone, but likewise upon imperfectly 
 frozen layers, formed by the consolidation of water — that is, 
 pf melted snow particles. 
 
DAILY MOTION OF GLACIERS. 221 
 
 219. Comparison of Glacier and Lake Ice. — The origin 
 and subsequent history of lake and glacier ice being unlike, 
 their structure is likewise dissimilar. The lake ice represents 
 the slow crystallization of the water, and shows a beautiful in- 
 ternal structure when examined in strong light. The crystals 
 of which it is built up are identical with those of snow, and lie 
 in the planes of freezing. The glacier ice, on the other hand, 
 is formed by comj^ression of snow; and in the process the 
 crystalline character is annihilated, yielding idtimately a 
 transparent substance. Moreover, in addition to vertical, 
 there is lateral pressure and motion: the ice mass is broken 
 up internally by the motion, and comes, in fact, to present a 
 granular aspect, not a crystalline one. They stand to each 
 other, as Helmholtz puts it, in the same relation as calc spar 
 and marble, both of which consist of carbonate of lime; but 
 in the former, as in lake ice, the material is regularly 
 crystallized; in the latter, as in the glacier, it is in irregular 
 crystalline grains. 
 
 220. Daily Motion of Glaciers. — It is well known that 
 glaciers descend the valleys in which they lie. Their lower 
 end remains at the same place for many years, notwithstand- 
 ing the incessant melting to which it is subjected, and this 
 fixity can only be secured by a constant supply from above. 
 The fact that the glacier extends beyond the limit of per- 
 petual snow, necessarily presupposes motion. But apart 
 from theoretical considerations, the motion has been observed 
 and registered by recording the movements of the debris on 
 its surface, and by the insertion of jDOsts in lines across the 
 ice. On the Mer de Glace, below Montanvert, the daily 
 motion of pegs inserted in a line from west to east, was found 
 by Tyndall, 12, 17, 23, 26, 25, 26, 27, 33 inches. Higher 
 Tip, opposite Les Fonts, the posts from west to east showed 
 a movement of 7, 13, 16, 20, 21, 23, 22, 15. Without 
 following all the details of the experiment, suffice it that the 
 shifting of the maximum speed from one side to the other 
 corresponds with bends of the valley; so that glacier and 
 river alike impinge most forcibly on the concavities of the 
 curves in their course. To this extent, therefore, the general 
 statement that the ice in the centre of the surface of tho 
 glacier moves fastest, must be modified. 
 
222 PHYSICAL GEOGRAPHY. 
 
 221. Curves of G-lacier Valley; their Influence on 
 Erosion of Channel. — The significance of the facts just 
 mentioned lies in this, that in some valleys or fioi'ds, such as 
 that of Loch Long, in Argyleshire, the deepest soundings 
 are nearer alternately to one or other side, and that the 
 deviation is greatest below the point at which a tributary 
 glacier entered. Of course this depth cannot fairly be 
 assigned to glacier erosion, unless its amount exceeds that 
 which a river of water is capable of scooping out for itself. 
 
 222. Mean Daily Motion: Seasonal Variations. — The 
 maximum speed of the Mer de Glace, observed in these 
 experiments in 1857, ranged from 20 to 33 or 36 inches 
 daily; while the movement at the margins varied from 7 to 
 15 inches daily. On the tributary glaciers the movement is 
 less rapid; that of the Glacier du Geant showing 11, 13, 
 5 inches as its lateral and maximum movement along the 
 transverse line; that of the Glacier de Lechand showing 5, 
 10, 6 inches. The winter motion of the Mer de Glace in 
 1859 shovv'ed, from observations near Montanvert, a speed 
 only half that of the summer months. 
 
 223. Variations of Movement at Surface. — It appears 
 from Tyndall's observation,''^ that not only does the maximum 
 velocity shift towards the concavity of the valley curves, but 
 that the movement is alternately faster and slower along 
 the same line. These irregularities are due to inequalities 
 of the channel in which the ice flows, inequalities such as 
 to make obvious at the surface the frictional resistance to 
 which the obstacle gives rise at the bottom. 
 
 224. Variation of Movement Below the Surface. — The 
 diminution of frictional resistance is progressive from below 
 lip wards; but the difference of rate of movement in the 
 highest and lowest portions, is pro2:>ortional to the depth of 
 the glacier. It has been stated that the surface velocity 
 varies, a reduction taking place where, probably, inequalities 
 of the channel diminish the vertical thickness of the ice. 
 An observation of Tyndall's on the Glacier de Geant showed 
 that the motion at 4 feet from the bottom, at 35 feet from 
 tlie bottom, and at the top vras 2| inches, 4 J inches, 6 inches 
 in twenty-four hours. 
 
 * Forms of Water, pp. 67-97. 
 
BIFURCATION OP GLACIERS. 223 
 
 S25. Retardation Due to Compression. — The Mer de 
 
 Glace is the joint stream formed by the glaciers de Geant, de 
 Lechaud, and de Talefre. The width of these before their 
 union is 2597 yards; but afterwards the three pass through 
 a valley, which at Trelaporte is only 893 yards wide. Below 
 this point the maximum velocities, already referred to, were 
 20, 23, 34, 25, 27 inches; but in the Glacier de Geant, thi^ee 
 points in the length of the stream moved 20*55, 15-43, and 
 12-75 inches daily, and as the extreme points were 1032 
 yards apart, the inference seems fair that the compression of 
 the ice in the narrows, and the consequent deepening of its 
 mass, have to do with the retardation, the counterpart of 
 which would be found in a river under similar conditions. 
 
 226. Difference of River and Glacier. — Similar as are 
 the movements of fluid and solidified water, one geologically 
 important difference exists. The river is arrested by an 
 obstacle and flows roimd it, the ice surmounts the obstacle, 
 retaining to a considerable extent its average thickness. The 
 ice is not rigid, neither is it viscous, but it has a certain 
 power of adaptation to the ineqiialities of its channel, 
 conferred by that plasticity which has been spoken of in 
 Alii. 213. 
 
 227. Bifurcation of Glaciers. — In one very important 
 particular the glacier differs from the river; for while both 
 are enlarged by the convergence of tributaries, the river only 
 divides in the delta, if it forms one. But the glacier, by 
 virtue of its plasticity, is not arrested, is not always diverted 
 from its course by elevations of its channel. An inequali'jy 
 of surface which would constitute a waterparting between 
 two streams of water, may present no obstacle to an ice 
 stream. The glacier may, however, divide, and its moieties 
 may descend valleys which lead ultimately to different seas. 
 Between the source of the Tweed and St. Mary's Loch, two 
 examples of such bifurcation are recorded in the still fresh 
 moraines. The glacier which occupied the seat of Loch 
 Skene divided on the opposite hill, and the curved terminal 
 moraine shows that one portion went towards the Grey 
 ]\Lxre's Tail to reach the Solway, the other descended Winter- 
 hope Burn to join the Megget. On the opposite side of the 
 waterparting, the Upper Talla glacier similarly divided, one 
 
22^ PHYSICAL GEOGRAPHY. 
 
 portion passlnf^ to the left, tlie otlier passing by Megget to 
 St. Mary's Loch.^ 
 
 228. Crevasses: Bergsclirund. — When the angle of the 
 bed of a glacier changes: a, if the bed rises, the ice of the 
 Tipper stratum passes through a greater space in the same 
 time than that of the lowest layer; h, if the bed slopes 
 downwards, the bottom moves through a greater arc in the 
 '^ame time. In the one case the surface is compressed; in 
 the other it is stretched. 
 
 The cross fractures, known as crevasses, occur where a 
 change of level takes place, as in the cascade of the Glacier 
 du Geant, where the bed suddenly slopes, and in the Gorner 
 glacier at Zermatt, where the more rapid motion of the mid 
 stream strains the lateral portions, which break when the 
 pull exceeds a certain amount, in other words, when it is 
 greatly in excess of the yielding by melting (Art. 213). 
 The ridges between these transverse fissures may give way 
 under local strain, and originate new fissures connecting the 
 crevasses, the portions of ice thus isolated being known as 
 seracs. The bergschrund is a crevasse formed where the 
 neve adheres firmly to an outstanding precipice, while its 
 lower part is carried forward by the stream; in this, as in 
 the other case, the line of fracture is at right angles to 
 that of the stream. 
 
 229. Dirt Bands. — But below the point at which the cre- 
 vasses are formed, regelation unites the broken surfaces; 
 not uniformly, however, for the portions descend so as to 
 present a series of terraces, just as a faulted mass of rocks 
 presents inequalities of level of its parts. Solar radiation 
 begins to act, and, as the vertical face of the terrace melts at 
 a different rate from the plane surface, alternate bands of 
 pure and of debris-strewn ice result, the latter soon showing 
 the characteristic curves of a stream in which friction retards 
 the lateral portions. 
 
 230. Diminution of the Glacier by Superficial and Ter- 
 minal Loss. — The upper surface of the glacier undergoes 
 diminution by evaporation, and the lower by melting. The 
 svm's heat during the day melts the surface, and the water 
 pours down into the centre of the glacier through the ore- 
 
 ^' quart. Jour. Geol Soc, 1864. 
 
riMINUTION OF FEEDING GROUND: SURFACE WASH. 225 
 
 vasses. The limit of forward movement of the glacier is 
 determined by the temjDerature of the valley into which it 
 flows, and variations of "temperature are recorded by the 
 position of the detritus which falls from the extremity. If 
 the heat of the valley is such that the melting of the ice is 
 equal to its forward movement, the extremity of the glacier 
 is stationary; if the melting is greater than the supply from 
 behind, the extremity recedes; if, on the other hand, the 
 forward movement is in excess of the melting, the glacier 
 advances farther into the valley, and these three stages are 
 illustrated from time to time in the Alps. The rate of 
 movement has been determined by Agassiz, Forbes, Tyndall, 
 and others, and varies from a few inches daily to a few feet. 
 The backward or forward movement of the glacier, or, as it 
 may be more strictly called, the advance or retreat of its 
 lower limit, according to seasonal variations or climatal 
 changes during long periods, is estimated by the position, at 
 the extremity of every glacier, of the detritus which it 
 conveys. 
 
 231. Dimensions of Glaciers. — The length and depth of 
 a glacier depend on the area of the feeding ground, the 
 number and size of tributaries, the amount of snowfall, and 
 the temperature of the valley into which the ice descends. 
 The Alpine glaciers are never more than thirty miles in 
 length at the present day, but the signs of their former 
 greater size are traceable for more than sixty miles from the 
 present end, and their scorings are seen 2000 feet above the 
 bottom of the valley. Nowhere in low latitudes are such 
 dimensions to be found now. Higher average annual tem- 
 perature, due to secular changes; a drier atmosphere, conse- 
 quent on man's agricultural operations; hot winds from the 
 Sahara sand desert, in place of the cooler moist winds 
 which crossed the former water surface of that region — all 
 these have tended to diminish the size of the Alpine glaciers. 
 In the Himalayas, a glacier of 36 miles long has been 
 measured, and in New Zealand the great Tasman glacier is 
 12 miles in length. The Humboldt glacier, in Greenland, 
 measures at the coast 60 miles across, and 300 feet tliick. 
 
 232. Diminution of Feeding Ground: Surface Wash. — 
 In addition to these climatal changes, it must be remembered 
 
 23 P 
 
226 
 
 PHYSICAL GEOGRAPHY. 
 
 that the downward passage, even of neve, denudes the rock 
 over which it travels. In all hill^ districts on which snow 
 lies for some part of the year, the higher slopes are covered 
 with a quantity of irregular debris, the fragments of which 
 are often irregularly scratched. The snow, slipping from 
 time to time, tumbles these fragments over each other, and 
 they are gradually carried down hill. This surface wash is 
 also met with in glacier districts, on spots too steep for the 
 formation of a glacier. During the long periods that the 
 incipient glaciers have slowly ground their beds, reduction 
 of the snow field or feeding ground must have resulted, and 
 this must be added to the causes of the diminution of 
 glaciers. 
 
 MORAINES. 
 
 233. Avalanches. — If the slope is too rapid for the snow 
 to accumulate and slowly compress its lower strata into ice, 
 the snow, or it may even be the neve, glides down, when the 
 equilibrium is overthrown. These avalanches are often of 
 enormous size. But ice avalanches also occur when a glacier 
 forms in a hollow, which terminates not in a gentle declivity, 
 but either in a cliff or in a slope too steep for it to rest. 
 The end of the glacier breaks off: if it were in water, it 
 would float away as an iceberg; being on land, it forms aji 
 avalanche, 
 
POSITION OP DETRITUS ON GLACIER. 
 
 227 
 
 234. Position of Detritus on Glacier. — Moraine is tlie 
 term used for the rubbish while it is still being borne by the 
 ice. The larger blocks, and the more conspicuous debris, are 
 along the sides of the glacier, being derived from the walls 
 of the valley; and if the glacier remains single, the two 
 marginal lines run parallel nearly to the end. But if two 
 glaciers unite into one stream, the adjacent lateral moraines 
 coalesce, and form a single median moraine. But this union 
 is accompanied by a change of speed; for whereas both lateral 
 moraines were retarded by friction, united they descend 
 with the speed of the central stream. The finer detritus, 
 and the small isolated stones, sink into the ice, being heated 
 by the sun. But all debris in large enough blocks, or in thick 
 enough layers, to permit the slow conductivity of stonB to 
 come into play, remains on the surface. As the ice melts, 
 that portion on which the objects rest is protected from 
 the sun's heat; it does not melt, and thus comes to form 
 a pillar or a ridge, with a more or less extensive cap. In 
 the same way, small pebbles jn'otect little cones of sand from 
 rainfall, and thus come to rest on peaks, the intervening 
 hollows beino- denuded. 
 
 ~<"^ ^^ ' -L'»,aiJ 
 
 CLACIEr» TABLE. 
 
223 
 
 PHYSICAL GEOGRAPHY. 
 
 235. Part of Debris Sinks into Glacier. — But not all 
 
 the rubbish remains on the surface; part drops in between 
 the ice and the valley wall, part is engulphed in the crevasses. 
 Received into the spongy, slushy ice forming the lowest 
 stratum, the coarse and fine materials are carried forward, 
 grinding and being ground. The sediments resulting from 
 this process are carried out at the end of the glacier by the 
 stream thence issuing, the water of which is derived from 
 the melting of the upper and lower surface of the ice. The 
 coarser materials are detained in the immediate vicinity of 
 the glacier; the finer are carried away and deposited along 
 the course of the river, the distance to which the consequent 
 muddiness extends depending on the speed of the river, or 
 the occurrence of lakes in its course. The loess (loss), which 
 forms so important a deposit in the tertiary series of the 
 Continent, is derived from glacier erosion. 
 
 236. Moraines of Deposit. — Two kinds of material are 
 found at the lower end of the glacier, the one distributed 
 mostly in mounds, and consisting of angular fragments of 
 rock in the same state as when they fell from the slopes 
 
 bordering the glacier ; 
 the other consisting of 
 fine mud, of gravel. 
 
 and of larger blocks, 
 all giving evidence in 
 their rounded edges and 
 smooth surfaces of prolonged friction, llie superficial mor- 
 aine, as the piles of angular rubbish are called, may have the 
 form of a series of cones, each cone corresponding to the line 
 of detritus that has travelled down on the ice ; or it may 
 appear as a continuous ridge parallel to the face of the 
 glacier, and sometimes so regular as to look like an artificial 
 embankment. The melting at the extremity of the glacier 
 is not confined merely to its face, but the sides likewise, for 
 a short way up, melt at the same rate, and the rubbish borne 
 at these parts, falling over, constitutes a lateral moraine, 
 which, uniting with the terminal moraine, may form a 
 semicircular barrier round the ice. These piles of angular 
 materials, the size of the fragments in which is often very 
 considerable, rest upon a quantity of deep morainic matter, 
 
SUBGLACIAL STREAM. 229 
 
 As no English word exactly corresponcls to the French term, 
 it is perhaps preferable to use it, and speak of the mud and 
 polished stones as constituting the moraine profonde. This 
 moraine profonde derives its materials both from the iipper 
 and the under surface of the glacier. The liea-sy mass of ice, 
 travelling slowly along its channel, grinds the surface of the 
 rock beneath, and carries away a certain amount of finely 
 pulverized material. The droppuigs from the sides of. the 
 valley tumble either upon the top of the glacier, or fall into 
 the groove between the ice and the valley wall. Some of 
 the superficial moraine drops into the ice through the cre- 
 vasses, and ultimately reaches the lower surface; while that 
 which has fallen in at the side of the glacier very speedily 
 gets to the bottom, and each particle, caught up and carried 
 forward by the ice, becomes a powerful agent in friction, 
 disintegTating the sui'face over which it travels. But the 
 fragments ha^'e not the free motion of pebbles in water ; 
 and while the latter rolling over and over acquire a rounded 
 form, being smoothed at all points, the glacier -formed 
 stones are usually angular, the angles being sciiiewhat 
 rounded, and one or more of the surfaces uniformly smoothed. 
 The dii-ection of the scratches or strise upon these polished 
 surfaces tells whether the stone has moved continuously 
 onwards in the same line, or has moved alternately in diffe- 
 rent lines; in the one case the scorings are parallel, in the 
 other they cross each other. The mud and very fine sand 
 form an impalpable powder, which takes a very long time to 
 settle in water, and this thinner material issues from beneath 
 the ice in a turbid stream. 
 
 237. Subglacial Stream — Notch in Terminal Moraine. 
 — There is constant melting going on between the ice and its 
 channel, due to heat developed by friction, to pressure, and 
 the passage of water from the surface into the deeper parts, 
 the evidence of this passage being the fact that a g^icier 
 stream is smaller during the night than during the day. In 
 consequence of the permanence of this stream the terminal 
 moraine never forms an uninterrupted mound, but is breached 
 at some point or another, and the finer sediment is carried 
 away to lower points in the valley, the coarser being left 
 Ibehiud. If the glacier recedes rapidly, the moraine profonde 
 
236 PHYSICAL GEOGRAPHY. 
 
 is sp:f^ead out as a sheet -witliiii tlie termmal moraine, and lias 
 all the appearance of a newly-drained lake bottom. If the 
 recession is slow, rows of terminal moraine form semicircles 
 across the valley, and it is by the distance of the older moraine 
 mounds, from the ice at the present time, that we infer the 
 former extension of the glacier. 
 
 238. Glaciers at Sea Level.— To the north of 70° kt. 
 glaciers descend to the coast line. When the ice reaches the 
 sea the detritus it carries is spread out over the' adjacent sea 
 bottom. If the ice is in sufficient mass to push out beyond 
 the shore line, it may travel seawards for some distance in 
 contact with the bottom before it gets into water deep enough 
 to float it. As the specific gravity of ice is such, that for 
 every foot of ice above the surface of water there are about 
 9 feet below, it is obvious that in shallow seas a glacier may 
 extend very considerably beyond the limits of the land; nay, 
 if the ice is very thick, it may spread over the surface of the 
 sea as a considerable cake before the movements of the water 
 have sufiicient power to break it through. Kane speaks of 
 the ice sheet in the Arctic seas as moving up and down with 
 the action of the tide like a door upon its hinges. At some 
 point or other of the floating sheet portions are detached by 
 the formation of crevasses, and these float away as icebergs, 
 carrying with them, adherent to the under surface, such 
 detritus as they may have picked up. Upon the upper sur- 
 face, animals that have by accident been unable to escape are 
 not unfrequently carried ; and in this way fragments of the 
 rocks of one region may be scattered over the ocean floor of 
 another region, and the bones of animals may be found in 
 latitudes to which they are not native. 
 
 239. Striation of Glacier Bed. — The friction exerted by 
 a glacier is greatest, obviously, at the bottom of its channel; 
 but the sides likewise are grooved and scratched in exactly 
 the same way, although with diminishing force, as far as the 
 last point of contact of ice and rock. The former greater 
 thickness of glaciers is frequently recognisable from the height 
 at which these lateral longitudinal markings are found above 
 the limits of the present glacier. Glacier denudation is 
 unlike that of rivers, inasmuch as, when an obstacle occurs 
 in its course, the results of friction are entirely confined to 
 
GUCIAL PERIOD. 231 
 
 tlie upper side of the obstacle. " TliiiS; we may distinguisli tlie 
 lee side of a projecting mass of rock by its angularity; whereas, 
 in a river, both the upper and the lee side are smoothed, 
 although, perhaps, not equally so. 
 
 240. Characteristic Features of a Glaciated District. — 
 The uniform abrading power of the ice gives a characteristic 
 roundness to the inequalities of surface over which it has 
 passed; and thus the centre of Scotland is distinguished by 
 the evenly curved outlines of the hill summits, particularly 
 in the southern districts. The smaller dome-shaped bosses of 
 rock are known as roches moutonnees, a term which, however, 
 is equally applicable to the hills. 
 
 241. Extent of Ice over Different Regions. — The amoimt 
 of ice covering particular districts varies considerably^ In 
 the Alps, snow and ice cover the central summits and the 
 heads of the valleys, the peaks and ridges separating valleys 
 being frequently entirely free. In Greenland, on the other 
 hand, no mountain peaks are seen projecting; but the whole 
 country is covered with a sheet of ice and snow, which seems 
 to rise inland, and to conceal even the most prominent features 
 of the land. But it must be remembered that little of the 
 interior has been explored, and that fog and haze render 
 unreliable the observations made from a distance. It is 
 possible that Greenland is not a continent, but a. series of 
 low-lying islands. 
 
 242. Development of Ice Sheet of N. Hemisphere in the 
 Glacial Period. — During the glacial period a sheet of ice 
 covered the whole, even of the more prominent features, of 
 the land in the northern temperate regions. The growth and 
 diminution of this great polar ice -cap was gradual. The 
 increasing rigour of the climate, dependent upon astronomical 
 changes which will be discussed hereafter, permitted the for- 
 mation of glaciers at first round the central peaks of mountain 
 ranges. As the cold increased these extended into the plains, 
 and, filling the valleys up to their summits, flowed over them, 
 so that adjacent glaciers became united. The land — in some 
 areas at least — went down, and thus the ice covering became 
 universal; enormous tracts presenting the appearance of Green- 
 land. As the climate improved again, and as the land rose 
 above tho level of the sea, the ice covering shrank, till at last 
 
233 tHYglCAL GHOGRAPfiY. 
 
 only limited glaciers remained in tlie higli grounds j and iii 
 some parts of the northern hemisphere these also disappeared. 
 Thus, in Britain, the hill districts of Wales, Cumberland, 
 south and north Scotland, still contain the remains of glaciers 
 in the form of moraines, as perfect as if they had been shed 
 yesterday. 
 
 243. Definition of Glacier, and Ice Sheet. — It is conve- 
 nient to retain the distinction between glacier and ice sheet, 
 notwithstanding the fact that they merge into each other. 
 The former refers to a mass of ice whose movements are con- 
 trolled by the minor features of a country; the latter to a 
 mass which has overtopped these boundaries, and is no longer 
 controlled by them. 
 
 244. Ice Sheet of Glacial Period formed by Fusion of 
 Local Glaciers. — The theory that the ice coverings of all the 
 northern continents originated, each in its own district, and 
 that, as for example in Britain, each separate hill district 
 became a centre for a separate mass of ice, depends for its 
 proof upon the now well-ascertained fact, that the scratch- 
 ings left by the ice follow the lines of ,, the great valleys, 
 and radiate from the highest points of the country. Not 
 merely do these stride follow the valleys, but they curve in the 
 valleys in such a way as to suggest a solid body which has 
 been deflected from side to side like a stream of water, only 
 with less sharp curvatures : they repeat, in fact, the deflec- 
 tions indicated (Art. 220) by the observations on the Mer de 
 Glace. The mass of the Scandinavian peninsula shows very 
 beautifully this radiation from the higher grounds. In 
 Britain we find indications that the ice sheet, as soon as it 
 had become continuous, moved without regard to minor 
 inequalities, though still retaining a certain relation to the 
 leading features of the country. 
 
 245. Lower Boulder Clay, the Moraine Profonde of the 
 Ice Sheet. — The northern part of the British Islands is 
 covered with a series of accumulations, the lowest of which 
 is a clayey mass, more or less tenacious, according to the 
 amount of slialey strata whose distintegration has contri- 
 buted materials to it. In this clay are included fragments 
 of rock showing the rounding of the edges, the smoothing 
 and scratching of the flat surfaces, and the frequent limita- 
 
ttELAflON OP BOULDER CLAYS. 233 
 
 tion of tlie scratching to one face only; in fact, all tlie 
 characters which are presented by the galets of the modern 
 glacier. Tliese materials, of Tarious sizes and shapes, are 
 thrown together Avithout any order; they are scattered irre- 
 gularly through the mass, and this, as well as their positions, 
 indicates that they have not been assorted by water. The 
 absence of angular fragments in the mass, though they are 
 frequent on its surface, or to put it more plainly, the fact 
 that all the fragments, large and small, have undergone fric- 
 tion, shows that the till is not the remains of the superficial 
 moraine, but of the moraine profonde. The absence of such 
 unworn fragments as have travelled along the surface of 
 the modern glacier, might have been expected, since, in the 
 first place, the ice sheet at the period of its greatest develop- 
 ment covered all the peaks from which fragments might 
 have dropped on its surface; and, in the second place, the 
 shrinking and re-extension of the ice sheet, of which there 
 is abundant evidence, must have given even to unworn frag- 
 Inents their galet form. 
 
 246. Upper Boulder Clay, or Stratified Till.— But this 
 Tinstratified mass is covered in the north of England by a 
 deposit into which it gradually passes, or from which it is 
 separated by a layer of sands and gravels containing marine 
 shells. The stones in this till have evidently the same 
 characters as those of the upper boulder clayj but they show 
 signs of longer friction; and the more or less regular strati- 
 fication indicates that water has had some share in their 
 reassortment. 
 
 247. Relation of the Upper and Lower Boulder Clays. — 
 The accompanying diagram shows how the lower till was 
 deposited by the ice sheet, and how the stratification came to 
 pass. The enormously thick land ice could not have allowed 
 the accumulation of 90 or 100 feet of till beneath it, so long 
 as the whole country was above sea level. But bearing in 
 mind that the ice sheet actually passed far beyond the limits 
 of the land, and that its specific gravity kept it in contact 
 with the bottom for a considerable distance, it is evident that 
 when the depth of water was sufficient to float the ice, an 
 angle was left between its lower surface and the sea bottom, 
 in which the rubbish pushed down by the ice lodged. As 
 
234 PHYSICAL GEOGRAPii^. 
 
 the land went down this angle increased, the detritus was 
 deposited in greater quantity, until afc last there was suffi- 
 cient depth of water to allow currents to affect the uppermost 
 part, and to reassert it in a rude fashion. The lower and 
 upper tills are inverse to each other in quantity, as might be 
 expected from the above explanation. It follows, then, that 
 the lower till was not deposited on the land above sea level 
 but below it, and its accumulation in quantity was only 
 possible when the ice was lifted off the sea bottom in conse- 
 quence of continued submergence. 
 
 -^-, BERa GLACIER 
 
 Lav el 
 
 S. B. C— Stratified Boiilder Clay. B. C— Lowel' or Unstratified Boulder Clay. 
 
 The diagram represents the glacier from the point where it 
 quits the land. 
 
 248. Erratics.— Over all glaciated regions large blocks 
 are found, whose size indicates that they could not have been 
 carried by water, while their shape shoAvs tha,t they have not 
 been subjected to water friction. Some of these have been 
 identified as fragments of rocks, whose locality is so distant 
 that no agent but ice could have transported them to where 
 they are now found. Some of these were doubtless carried 
 on glaciers and left when the ice melted; others must have 
 been carried on icebergs, and dropped when the berg melted, 
 or, as frequently happens, was overturned. Since icebergs 
 at the present time travel in the North Atla„ntic as far south 
 as the Azores, in the South Atlantic to within 36° of the 
 equator, it is not difficult to understand how, in former times, 
 erratics may have wandered far from the parent rock. These 
 blocks are sometimes found perched on the summits of rocky 
 ridges, in positions where water could not have left them. 
 But it must be remembered that some rocks, as granite, are 
 decomposed by the atmosphere in such a way as to leave 
 blocs percheSf exactly simila<r to erratics, though they have 
 
icebergs: ice floe. 235 
 
 never ti-avelled from tlie spot. Many of tlie Logan stones, 
 or rocking stones, of the tors in Cornwall, have this local 
 origin. 
 
 249 Pack Ice : Ice Foot. — In the Arctic regions, and in 
 such seas as the Gulf of Bothnia, the surface of the sea itself 
 is frozen for a part of the year. The pack ice, extending 
 seawards, commences at first with the ice foot, a narrow coast 
 belt of ice, sometimes increased to 30 feet in thickness. The 
 sheet becomes broken up by channels, and, under the influ- 
 ence of storms, portions become piled upon each other, consti- 
 tuting masses dangerous to navigation. Geologically, the 
 pack ice is productive of very little change; but the detached 
 fragments may, like the icebergs, become the means for con- 
 veyance of animals into lower latitudes. 
 
 250. Coast Ice. — Coas,t ice is rarely formed within the 
 British area, and is, for the most part, met with only where 
 fresh waters are poured into the sea. 
 
 251. Ground Ice. — Ground ice, or ice formed at the bottom 
 of rivers, is, in some of the American streams, of considerable 
 importance as a. geological agent. The sheet, whose forma- 
 tion is determined by the contact of water with stones at the 
 bottom, which have become chilled by radiation in clear 
 water under a cloudless sky, rises and carries with it the 
 fragments among which it was formed, conveying them down 
 the stream into the open sea, and, as the ice melts, distri- 
 buting them over the floor of the ocean. 
 
 252. Iceberg : Ice Floe. — By ground ice, by icebergs, and, 
 under certain circumstances, by coast ice, very large quan- 
 tities of sedimentary materials are transferred from one place 
 to another, in addition to the more conspicuous erratics 
 already mentioned. The iceberg is the mass detached from 
 a glacier at the sea level, or broken off" from its end at the 
 top of a coast cliff"; the ice floe is a part of the pack ice set 
 free by being broken up under the influence of storm waves 
 and tidal movement. The bergs are often of enormous dimen- 
 sions : even after they have travelled far, a height of more 
 than 200 feet has been seen above water, indicating, if the 
 berg were of equal cii'cumference at all parts, a total height 
 of at least 1800 feet. It has been sufrsrested that the dimen- 
 eions of icebergs have been exaggerated; perhaps what Mr* 
 
236 PflYSICAL GEOGRAAHY. 
 
 Kuskin calls excitement, when it leads to distorted drawing 
 of hill slopes, is excusable in the circumstances under which 
 the largest reported icebergs were encountered; but if there 
 has been error, the optical phenomena attending their appear- 
 ance may have had some share in it. 
 
 253. Travelling of Icebergs. — Bergs reach the South 
 Atlantic, seemingly across the easterly drift; others traverse 
 the Gulf Stream diagonally. It has been suggested that their 
 mass above water puts them in the position of a ship with 
 sails set, which must sail on the wind. Laughton, without 
 denying this in all cases, suggests that the Antarctic icebergs 
 are carried north by local currents, determined, perhaps, by 
 projections of the unknown Antarctic land; and that the 
 Arctic bergs acquire sufficient " way" to traverse the Gulf 
 Stream, aided, perhaps, if they are. deep enough to reach it, 
 by the cold stream which dips under the warm one. 
 
 254. Geographical Effects of Icebergs: Striation.— The 
 distribution of detritus and of animal remains has been 
 referred to. The number and frequency of bergs off New- 
 foundland is believed to have materially increased the height 
 of the Bank, and correspondingly shallowed the water during 
 long ages. The striae made by bergs were once thought to 
 explain the phenomena of glaciation, now assigned to the 
 action of land ice. The continuity of the strise under the 
 boulder clay, however irregular the surface may be, could 
 not have been effected by floating ice, which must pass round, 
 but cannot override an obstacle. The desolation of a district 
 by the chilling of the atmosphere during the melting of a 
 large stranded berg, as happened in Shetland, is fortunately 
 a rare occurrence. >■■■■ 
 
 265. Hail. — Though hail is one of the forms of water which 
 has been frozen, the circumstances which invariably attend 
 its occurrence entitle it to separate consideration. Hail 
 S.torms are events of summer and of the day, not happening 
 at night or in winter. They are preceded by slight depres- 
 sion of the barometer, and usually great heat: immediately 
 before the fall a shivering sound is often audible; the baro- 
 meter rises, and the air is cooler afterwards. They are most 
 frequent within the tropics, diminishing northwards. Moun- 
 tain ridges seem to influence their severity and frequency. 
 
RELATION OP HAIL TO STORMS.' 237 
 
 Obviously the freezing of atmospheric moisture yields the 
 hailstones; but the rapidity with which they have been 
 frozen, judging from the regularity of the crystals, makes it 
 difficult to explain their formation, since the dry condition 
 in which they usually fall indicates that they have been 
 subjected to a temperature far below 0°C. 
 
 256. Structure of Hailstones. — They are more or less 
 spherical or oval; and on fracture often very strikingly re- 
 semble the zeolites or radiating crystalline minerals found 
 in trap rocks. The crystalline lines are often crossed without 
 being interrupted by concentric lines, such as would result 
 from the freezing of successive layers of water. The size 
 varies from small shot up to masses of an inch in diameter; 
 larger stones — and some of very great size are reported — are 
 formed by adhesion of several, and probably in the extreme 
 cases the regelation took place after they had reached the 
 ground. The freezing probably affects a sphere of water on 
 which more moisture is deposited and freezes, so that when 
 conical pieces are formed these are probably fragments of 
 broken sjiheres. But in the hailstorm of July 1872, in 
 Glasgow, fragments were caught in large quantities before 
 they reached the gi'ound, and these were conical, with spherical 
 bases, but no trace of concentric arrangement. Their dis- 
 ruj)tion must have taken place in the atmosphere. 
 
 267. Relation of Hail to Storms. — Hail in Europe usually 
 comes with a S.W. wind, though local features may shift 
 the direction. The suddenness of the storms, the speed with 
 which they travel, and the restricted area they cover, coupled 
 with the fact that their direction coincides with the usual 
 wind storms of a country, make it probable that, taking place 
 most frequently in tropical regions, and in the hot weather 
 of temperate regions, they are associated with alterations in 
 the electric tension of the atmosphere even when they are not 
 accompanied, as often happens, by a thunderstorm. Whether 
 the cold is that of the upper air into which warm moist air 
 has ascended, or is the result of the meeting of two currents 
 at une(^ual temperatures; is uncertfiin, 
 
CHAPTER VI. 
 
 SECTION I.— THE ATMOSPHERE. 
 
 Composition of Air : its Density — Height of the Atmospheric Column: 
 Ether — Variations of Pressure : Height and Temperature : Aque- 
 ous Vapour — Areas of High and Low Pressure — Annual Varia- 
 tions of Pressure — Influence of Aqueous Vapour on Temperature 
 - — Absorptive Power of Gases — Influence of Ozone — Temperature 
 of Atmosphere — Decrease of Temperature with Height — Eftects 
 of Heat — Analogies of Light, Heat, and Sound — Transmission of 
 Light through the Atmosphere — Transparency and Colour of 
 Atmosphere — Reflection: Refraction: Absorption — Twilight — 
 Absorption : Diminution of Light by Distance — Polarization of 
 Atmosphere — Transmissson of Sound — Intensity of Sound — 
 Sounds louder by Night — Refraction and Reflection of Sound — ■ 
 Resonance — Transmission of Heat — Reflection of Heat — Diminu- 
 tion by Distance. 
 
 258. Composition of Air. — Having hitherto considered 
 what may be called the proper mass of the earth, all, that is 
 to say, which is essential to the planet as a body moving in 
 space, v/e have now to inquire into the composition, proper- 
 ties, and movements of the environments of the spheroid. 
 Disregarding, for the time, speculations as to the existence 
 of an ether occupying an apparently vacant space, through 
 which the terrestrial bodies move, we shall confine our 
 attention to the atmosphere, that invisible, elastic layer of 
 variable and uncertain thickness, which surrounds the 
 globe. Perfectly pure air consists of oxygen and nitrogen, in 
 the proportion of 21 to 79 by volume; but this theoretical 
 atmosphere so constantly contains other substances that one 
 is almost in doubt whether the term impurity is legitimately 
 applicable to the compound. Carbonic acid is almost always 
 present; other gases are likewise present in small quantities; 
 and there are, in addition, solid matters of various kinds, 
 organic as well as inorganic. The mixture of the different 
 
 I 
 
HEIGHT OF THE ATMOSPHERIC COLUMN: ETHER. 239 
 
 gaseous components of the atmospliere is not one which 
 depends upon motion for its completeness, since, if the gases 
 were left in contact without disturbance, a compound would 
 very soon be formed by diffusion. The particles of which 
 this elastic layer is composed are not stationary, either rela- 
 tively to each other or to the solid globe, and the various 
 kinds of movement, the existence of which we know with 
 certainty, fall to be considered by the meteorologist, who 
 tests the scientific value of his conclusions by the accuracy 
 v\^ith which they enable him to foretell atmospheric change. 
 
 259. Density of Atmosphere. — If we ascend from the 
 level of the sea to the summit of a mountain, we find that 
 the density of the air is diminished, the pressure, that is to 
 say, is less ; and if the height to which we have ascended is 
 12,000 feet, the atmosphere, which at the sea-level would 
 have occupied a certain cubic space, will, at the greater 
 elevation, occupy double that space. The mode of estimating 
 this difference of pressure is by the barometer, and if the 
 mercury stood at 30 inches at the lower level, it would re- 
 quire 15 inches at the upper level. At the level of the sea 
 the pressure of a column of atmosphere, the height of which 
 is unkno^vn, equals the pressure of a column of mercury 30 
 inches in height; and as this represents the pressure of about 1 5 
 pounds (14*7 lbs.) on each square inch of siu'face, the pressure 
 at the top of the mountain of 12,000 feet would be about 7^- 
 (7*35 lbs.) pounds on each square inch. It has been calcu- 
 lated that at the altitude of between 40 and 50 miles, no 
 appreciable pressure would be detected. 
 
 260. Height of the Atmospheric Column: Ether. — If 
 the atmosphere were of equal density throughout, it would 
 be, judging from the mercurial column, five miles in height. 
 But we have seen that the density diminishes with height, 
 and that at 40 or 50 miles no appreciable pressure Avould be 
 detected. It has been calculated from observations on meteors, 
 which become visible when they penetrate the terrestrial 
 atmosphere, that the upper limit is about 200 miles, while 
 M. Liais, from the phenomena of polarization, fixed it at 212 
 miles. Within and beyond this limit space is occupied by an 
 elastic medium or ether, which is capable of transmitting the 
 vibrations of light, and of retarding b^ friction, though the 
 
240 PHYSICAL GEOGRAPHY. 
 
 influence be very small, the motion of the celestial bodies 
 through it. 
 
 261. Variations of Pressure ; Height and Temperature. 
 
 — The pressure expressed in inches of mercury, that is to 
 ,say, the height of the column which the atmosphere can sus- 
 tain, varies under several influences. The diminished pressure 
 felt in ascending to a height is not equal in all parts of the 
 globe. For, as the force of gravity diminishes as the square 
 of the distance from the centre of the earth to any point on its 
 surface, the diminution will be slowest at the equator, most 
 rapid at the poles, the mean being about 45° N. lat. The 
 correction for height is, in Britain, '001 inch for every 400 
 feet of ascent ; the vertical column of the atmosphere is less 
 by that amount, and '001 expresses its diminished sustaining 
 power in inches of mercury. To equalize the difierence in 
 the rate at which the force of gravity diminishes, the mean 
 point, 45° N. lat., is the zero point; but, at the equator, "003 
 inch requires to be subtracted from the observed height ; at 
 the poles that amount requires to be added, so as to obtain 
 an average for pressure at the sea level at all points. The 
 temperature of the air modifies its pressure, which is less at 
 high than at low temperatures; but as temperature falls 
 •55^ C. for every 300 feet of ascent, the correction for height 
 is the addition to the observed temperature of a number of 
 degrees corresponding to the elevation. 
 
 262. Varieties of Pressure: Aqueous Vapour. — The quan- 
 tity of aqueous vapour in the atmosphere varies under condi- 
 tions whose recurrence is determined by the variations of solar 
 influence. In discussing this subject, the student must bear 
 in mind that he will find in these paragraphs only such a 
 general summary as will make him t© understand the compli- 
 cated character of the phenomena on which climate depends, 
 and the consequent difiiculty of deciding on the influences to 
 which plants and animals are subjected. The fuller discussion 
 of the phenomena will be found in special treatises, such as 
 Buchan's Handbook of Meteorology. Buchan tabulates the 
 diurnal variations of pressure at Calcutta, from which it 
 appears that the greatest pressure, varying from "039 to '076, 
 occurs at 9-30 a.m.; the next, at 10*30 p.m., varies from -008 to' 
 •026: while the minima range from - '017 to - -027 at 3-30a.m,; 
 
AREAS OF HIGH AND LOW PRESSURE. 241 
 
 from - '048 to - -071 at 4*30 p.m. As the tidal wave is later 
 than the time at which the sun and moon cross the meridian 
 of any place, these barometric variations follow periods at 
 which, judged by the sun's position, the daily maxima and 
 minima should occur. The sun crosses the meridian afc 
 noon; but as the earth is not rapidly heated, the highest 
 temperature is about three hours later; and the lowest tem- 
 perature, about 3 A.M., is correspondingly later than the hour 
 at which the sun crosses the antipodal meridian. The highest 
 and lowest pressure correspond to the hours at which the 
 atmosphere contains the greatest and least amount of vapour 
 of water. But though evaporation is greatest at the hottest 
 hours in the afternoon, the pressure is greatest in the fore- 
 noon, when the atmosphere has not yet been heated up by 
 solar radiation, and the elastic vapour is therefore retained in 
 its lowest stratum. The afternoon heat relieves the pressure 
 by expansion; the evening cold increases it till dew falls, and 
 thereafter the pressure diminishes till about 3 a.m., when, 
 though the air is coldest, it is also driest. It is obvious that 
 the more nearly equal the daily temperature, or the lower 
 the temperature, the less will be the daily variation of pres- 
 sures, and the difference is also less marked in uniformly moist 
 climates. But if the moisture is well-nigh equal throughout 
 the year, while the hot and cold of summer and winter are 
 extreme, the barometric variations will be great, as in Siberia, 
 and westwards, into the central plain of Europe. 
 
 263. Areas of High and of Low Pressure. — Apart from 
 the influences already mentioned, pressure is modified by 
 some other conditions. Thus, whoever has walked beside a 
 wall on an exposed hill top, while the wind was blowing at 
 right angles to the wall, has seen the dust blown against the 
 wall and obliquely upwards and leewards, or, if the wind was 
 very strong, vertically upwards, while, on the lee side, a slower 
 movement carries dust towards the wall, and upwards along 
 its face to join the main current. This atmospheric backwater 
 is more due to the action of the wind as it passes the top of the 
 wall, tearing off a portion of the leeward aii*, and thus drawing 
 on a current, than to the curve downward of the air behind 
 the obstacle. To windward the air is jammed against the 
 wall, to leeward it is rarefied : to windward atmospheric 
 
242 PHYSICAL GEOGRAPHY. 
 
 pressure is increased, to leeward diminislied. By analogy, the 
 weather side of mountain chains might be expected to exhibit 
 increased pressure. This, doubtless, helps to increase the 
 pressure in Siberia, the maximum for that area not being at 
 the place of greatest cold, but to the westAvard, the Altai 
 and other high lands, whose axis is E.N.E., stopping the west 
 winds and heaping them up; while the area in which Yakutsk 
 lies is, like Eastern Patagonia, a leeward area of low pressure. 
 264. Annual Variations of Pressure. — The annual varia- 
 tions of pressure are determined by the heat and cold of 
 summer and winter, the former causing an increase of evapo- 
 ration, thus overcharging the lower strata with vapour; the 
 latter checks evaporation, while the tension is diminished by 
 the more copious precipitation. From the isobaric maps, or 
 maps on v/hich the lines of equal barometric pressures are 
 recorded, it appears that the pressure diminishes towards 
 the poles, but that the area over which pressures less than 
 29-9 are observed, advances and recedes with the season. 
 Thus, while in July the line of 29*9 is limited by the parallel 
 cf 40° S. lat., and the lines of lower j)ressure are curiously 
 parallel to it, in January the line curves northwards from 
 New Zealand, and gradually reaching lower latitudes in the 
 S. Atlantic, approaches nearest to the equator in S. America 
 about 25*^ S. lat. The 29*9 line in the northern or land hemi- 
 sphere is about 50^ N. lat. over tlie Atlantic in January, 
 55*** in July. In the summer this isobaric line curves south- 
 wards, over the American continent, nearly to the Gulf of 
 Mexico; over the eastern continent it reaches as far south as 
 the equator; but in both cases it recedes northward as it 
 approaches the Pacific. In winter (January), the line recedes 
 from its lowest latitude over the Atlantic, so as to form a 
 curve whose convexity is towards the poles, while the high 
 pressure follows the great continents. The seasonal influ- 
 ence is therefore well marked. Taking the mean of the year, 
 it appears that maximum pressure (30'1) exists in the north 
 and south Atlantic, in the former stretching across the 
 ocean between 25° and 40° N. lat,, in the latter between 15° 
 and 25° S. lat., and that similar areas occur in the Pacific, 
 between 120° and 160^ W. Ion., and 25° and 40° N. lat., 
 tliough no equally definite space can be indicated to the south 
 
INPLUE!TCE OP AQUEOUS VAPOUR ON TEMPERATUEE. 243 
 
 of the equator. A mean annual pressure of 30 inclies ranges 
 over JSTorth America between 20° and 50° N. lat., while in 
 the Old World it passes obliquely from the south of Europe 
 to the north-east of Siberia. In the southern or water hemi- 
 sphere, the belt is more nearly latitudinal between 10° and 
 35° S. lat. ^A low pressure zone, 29*9, swells out north- 
 wards towards the Himalayas, while the same pressure 
 characterises the Arctic area north of 55° lat., the Antarctic 
 area south of 40° lat. The greatest superficial area of high 
 pressure is over the land, of low pressure over the sea. 
 
 265. Influence of Aqueous Vapour on Temperature. — 
 The sha,re which the vapour of water takes in the phenomena 
 of atmospheric pressure is scarcely inferior to that which it 
 takes in modifying temperature. Perfectly pure air, consist- 
 ing only of oxygen and nitrogen, permits all the waves of 
 the sun, luminous and non-luminous, those of light and 
 heat, to pass through; its diathevinancy, which is to heat- 
 waves what transparency is to light-waves, is perfect; and 
 this is true also for oxygen, hydrogen, and nitrogen. A 
 sun-beam concentrated on ice, after passing through a globe 
 of water, is not apparently affected, yet it has no longer the 
 power of melting the ice; and if the positions of the two 
 "forms of water" v/ere reversed, the same result would 
 ensue. The want of diathermancy is therefore due to the 
 presence of water, not to the mode in which its particles are 
 aggregated. Several very impoiiiant conseqiiences follow 
 from this fact. The temperature of dry air is not affected 
 by the passage of the sun's rays through it; it is at once 
 transparent and diathermanous. While, therefore, the face 
 is blistered by the direct rays of the sun on an Alj^ine 
 glacier, the traveller has only to step into the shadow of a 
 rock to realize the fact that the temj)erature of the air is in 
 reality at the freezing point. This, of course, is only true 
 where the air is greatly rarefied, expansion lowering the tem- 
 perature. But even at great heights the air is comparatively, 
 not absolutely, dry in glacier districts, evaporation taking 
 place freely from the surface of the ice. The capacity of the 
 atmosphere for heat is therefore in proportion to the aqueous 
 vapour it contains. Air is, however, heated by contact 
 with warm surfaces, and, expanding, rises, colder and heavier 
 
244 PHYSICAL GEOGRAPHY. 
 
 air taking its place. The molecular motion produced in dry 
 air by a heated piece of metal is very different from that 
 due to the passage of heat-waves through moist air. Again, 
 the same layer which, as Tyndall puts it, filters the rays 
 of heat due to solar radiation, so that they do not reach 
 the earth, arrests the passage of heat radiated from the earth. 
 It appears from Tyndall's experiments, that " the aqueous 
 vapour of the air from (several) localities exerted an absorp- 
 tion seventy times that of the air in which the vapour was 
 diffused;" and that at least 10 per cent, of the terrestrial 
 radiation is arrested within 10 feet of the earth's surface. 
 From these facts it is clear that while the earth's loss of 
 heat by radiation cannot be so great as it would be were 
 the air perfectly dry, solar radiation does not contribute so 
 much as it would on the same hypothesis. Further, the 
 greatest possible proximity (Art. 349) of the earth to the 
 sun would not secure the melting of the polar ice, since the 
 evaporation due to the great heat would interpose a sieve 
 which lilfcered the heat rays, and saved the remaining ice from 
 melting. 
 
 226. Absorptive Power of Gases. — But the air contains 
 other matters than oxygen, hydrogen, and aqueous vapour. 
 Carbonic acid and ammonia are also present, and both of 
 them present in greater quantity during warm than cold 
 v/eather. The importance of these gaser' will be apparent 
 from the following table, taken in pavli from that of 
 Tyndall:-''— 
 
 Relafcive fibsorption at 
 
 1 iuoli of pressure. 
 
 Air, ..... 1 
 
 Oxygon, ... - - 1 
 
 Nitrogen, - - - - - 1 
 
 Hydrogen, - ... - 1 
 
 Chlorine, ----- 60 
 
 Bromine, - - - - - IGO 
 
 Carbonic Oxide, - - - . 750 
 
 Carbonic Acid, .... 972 
 
 Nitric Oxide, - - - - 1590 
 
 Nitrous Oxide, - - - - 18G0 
 
 Ammonia, ----- 54(50 
 
 Sulpliurous Acid, - . - - 6180 
 
 * Heat as a Mode, of Motion, p. G20, 
 
TEMPERATURE OF ATMOSPHERE. 245 
 
 Striking as is tlie proportion of heat rays, or calorific rays, 
 arrested by air and ammonia respectively, the power of 
 arresting the passage of radiant heat exercised by perfumes 
 is even more remarkable. In a series of experiments on 
 aromatic herbs, detailed at p. 335 of the above-quoted work, 
 it is shown that the following relations exist : — 
 
 • I ! , ■• A Absorptive Power. 
 
 Air, 1 
 
 Thyme, - - ... 33 
 
 Peppermint, - - - - 34 
 
 Spearmint, ... - 38 
 
 Lavender, ----- 32 
 
 Wormwood, - - - - 41 
 
 Cmnamon, ----- 53 
 
 Allowing for the minute exaggeration in these results 
 due to the presence of aqueous vapour, it still appears that 
 interference with the passage of heat rays is enormously out 
 of proportion to the quantity of solid matter which the per- 
 fumes of these plants contain. The vegetation of tropical 
 regions is less characterised by the brilliancy of its tints than 
 is that of temperate regions. In the latter, evaporation 
 spreads a protective curtain against the solar radiation; in 
 the former, the development of perfume by plants after 
 sundown will tend to check terrestrial radiation. Insiffnifi- 
 cant as these observations may appear to be, they are brought 
 under the notice of the student for the purpose of showing 
 him how many are the points to be taken into consideration 
 before we accept as final any judgment on the rate at which 
 the earth receives or parts with heat. The absorptive power 
 of carbonic acid given above establishes, apart from other 
 reasons, the impossibility of an atmosphere of that gas having 
 existed during carboniferous times, the plants of which we 
 have every reason to believe were physiologically identical 
 with those of the present. 
 
 267. Influence of Ozone. — This gas is an allotropic form 
 of oxygen, is, in fact, condensed oxygen, capable of effecting 
 oxidations which the elementary gas cannot. It is easily 
 decomposable by heat, its atoms resiiming, after expansion, 
 their relations as oxygen, but its absorptive power is 165 
 times greater than that of oxygen. 
 
 ^68, Temperature of Atmosphere.— The sensation of 
 
246 PHYSICAL GSOGRAPHy. 
 
 lieat or cold is relative, like tliat of salt and sweet, or any 
 other pair of contrasting impressions.' The traveller descend- 
 ing from the Alps complains of heat at the same place where 
 the ascending traveller suffers from cold; the impression 
 conveyed by the same degree of actual temperature is deter- 
 mined by the previous conditions. 
 
 Statements founded on mere sensa^tion are open to all 
 kinds of fallacy, only instrumental observations are reliable. 
 The popular saying, that " as the day lengthens the cold 
 strengthens," means, when interpreted by science, that as the 
 drier atmosphere of frosty weather is succeeded by a moister 
 and warmer state of the atmosphere, the great conductivity 
 of the moisture, whereby heat is rapidly abstracted from the 
 body, chills the body, notwithstanding that the thermometer 
 is steadily rising. The contrasting heat and cold on the 
 Alpine glacier have already been mentioned, the air having 
 the same temperature in sunshine and shade. 
 
 The thermometer records the varying amount of heat 
 contained in the complex substance, atmospheric air; and the 
 influence of aqueous vapour is not separated from that of 
 other substances. The temperatm^e of the air of the desert 
 Gometimes rises to 51'6°C,, this being due, however, to the 
 quantity of superheated sand particles suspended in it. This 
 is an extreme case of the phenomenon of heating by convec- 
 tion, or the successive transfer of masses of heated air. Con- 
 tact with warm surfaces raises the temperature of the lowest 
 stratum of air, the amount varying with the humidity. In 
 short, the temperature of the atmosphere is for the most 
 part expressive of the amount of other substances besides 
 oxygen and nitrogen contained in it, and the most potent of 
 these foreign bodies are aqueous vapour and fine sand. 
 
 269. Decrease of Temperature with Height. — The ob- 
 servations of Mr. Glaisher made during balloon ascents 
 {BritisJi Association Re'poTts, 1869), show that the estimate 
 above given — "55° C for every 300 feet of ascent — though 
 fairly representing the facts for air in contact with the 
 ground, is not correct for the air away from that kind of 
 disturbing influence. It appears that there is considerable 
 irregularity in the decrease, consequent on the number of 
 atmospheric currents. At a height of 24^000 feet; the tern- 
 
EFFECTS OF HEAT. 247 
 
 perature in September was -17-7°C., and at 37,000 feet it 
 was - 24r'39°C. At lower elevations, it would appear that 
 the ratio of decrease is slower; and that within 1000 feet it 
 is affected by the state of the sky, being -55^0. for 223 feet 
 when the sky is cloudy, for 162 feet when it is clear. 
 
 270. Effects of Heat. — The expansion of the air under 
 heat amounts to -^r^ of its volume for every . 1 "^ C. ( ^^o" ^^^ 
 every l^F.), so that if a cubic foot of air be heated 273° C, 
 it will double its volume. The co-efficient of expansion, or 
 the fraction by which the volume of a gas is augmented, when 
 its temperature is raised l°C.,is -00366 for hydi'ogen, -00367 
 for air and carbonic oxide. If a cubic foot of air is raised 
 from 0° to 273^0., its volume is doubled; and 1-29 oz. of 
 air is capable of raising through 1 foot the weight of 1 5 lbs. 
 to the square inch, or 2160 lbs. of atmospheric pressure. 
 But if the expansion is prevented, and the volume of the 
 air kept constant, there is a difference in the absolute quantity 
 of heat received by the two cubes of air. A gi-eater araount 
 is required to enable the air to overcome the resistance offered 
 by pressure to its expansion than when the air is heated 
 within unyielding boundaries; in other words, the air which 
 is prevented from expanding has no work to do, the air which 
 expands under pressure does work : it lifts a weight, and heat 
 is spent in doing this over and above that needed for the 
 given quantity of air. The proportion is 1 : 1-421; that is, 
 if the cube with constant volume requires one part of 
 heat, the cube which expands vrhen heated under pressure 
 requires 1-421 parts. The excess is spent in doing work. 
 To take again the cube of air: the lifting of 2160 lbs. 
 through one vertical foot, is the work done by the excess 
 of heat over that needed to raise the temperature of the 
 cube to 490°F. (273°C.). The excess would be sufficient 
 to raise 2-8 lbs. of water 1°F. of temperature, so that divid- 
 ing 2160 by 2-8, it follows that 771-4 lbs. would be lifted 
 1 foot by the amount of force which is made use of in 
 raising 1 lb. of water 1°F. in temperature. Such was Mayer's 
 calculation; Joule has corrected this result, and fixed the 
 amount at 772 lbs. This, the mechanical equivalent of the 
 heat, is known as the unit of heat, as a foot-pound; and 
 the quantity of heat received by solar radiation at different 
 
248 PHYSICAL GEOGRAPHY. 
 
 points of the earth's surface is frequently expressed in these 
 units or foot-pounds. Joule's equivalent is, in terms of the 
 centigrade scale, 1390 foot-pounds for 1° C. 
 
 The increase in volume of heated air corresponds to the 
 diminution during cooling: that which has been heated in a 
 confined space parts Avith its heat when it is allowed to escape 
 and expand. Rarefied air is chilled during expansion, its 
 heat being parted with in the motion of its particles. If the 
 weight which the heated air lifted during its expansion 
 descends during its cooling and contraction, the amount 
 of energy parted Avith is that of temperature, plus that 
 due to the impiilse given to the particles by the weight. 
 Ascent of heated and descent of cooled air, are thus incessant 
 and supplemental of each other. If priority is to be claimed 
 for either phase of the cycle, heating may be regarded as the 
 first step, since if the air were uniformly cool, there would 
 be no cause of disturbance. The replacement of warm by- 
 cold air is not always efiected by two equal parallel move- 
 ments. If a heated chamber is opened, the expanded air 
 rushes out, is checked, and chilled; there is a reflux of cold 
 air, and thus a succession of pulses occur before equilibrium 
 is established between the air within and that outside the 
 room. The movements are more simple when there is free 
 space for the opposite movements. 
 
 271. Analogies of Heat, Light, and Sound. — Heat, light, 
 and sound are vibrations of the particles of bodies; and 
 these are made manifest to our senses by the movements 
 which they impart to the elastic medium between them and 
 our organs of sense. Light and heat are both given off by 
 the sun, and experiments have been already referred to which 
 show that it is possible to stop the one set of vibrations, 
 and allow the other to pass. The independence of light and 
 heat has been asserted by Melloni, but it seems now satis- 
 factorily demonstrated that the solar spectrum represents a 
 series of vibrations of particles or waves, the length of which 
 permits them to be sensible to the human eye as colour; that 
 the length of the waves diminishes from the red to the violet, 
 and that beyond these visible rays other waves exist, those 
 beyond the violet having too short, those beyond the red 
 having too long, a wave length to be visible to the eye. 
 
TRANSMISSION OP LIGHT THROUGH THE ATMOSPHERE. 249 
 
 Finally, the invisible rays beyond the violet are 
 powerful to excite chemical action, those beyond 
 the red produce heat. The vibratory nature of 
 sound is known from the action of the tuning 
 fork, of the violin string, or, to take a more 
 familiar illustration, the result of drawing the 
 teeth of a comb across the edge of a piece 
 of paper. The range of sonorous vibrations 
 capable of being detected by the human ear, 
 varies very widely in different individuals, either 
 naturally or by practice. The limit of apprecia- 
 tion of sound is very wide, ranging, according 
 to Helmholtz, from 16 to 38,000 vibrations in 
 the second, vibrations slower or more rapid thp.n 
 these, respectively, being inaudible. But the 
 series of audible vibrations is very wide, extend- 
 ing over eleven octaves; whereas of the visible 
 rays of light, the shortest vibrations are not quite 
 half the length of the longest. The analogy 
 between these four kinds of vibrations is very in- 
 teresting; in strictness, there are only two sets 
 of phenomena to compare, light and sound being 
 respectively the centres of two series, of the ex- 
 tremes of which we are only made conscious by 
 their effects. 
 
 It appears from what has been stated that our 
 conceptions of external influences on living things 
 would be very imperfect, and necessarily erroneous, 
 if we considered only those phenomena which are 
 capable of direct interpretation by our senses. 
 Kecalling the wide sphere which at the outset we 
 claimed for Physical Geography, it is obvious 
 that the student of that science has only done a 
 portion of his work Avhen he has estimated the 
 effect on living beings of physical features, of 
 movements of the earth's crust, or of the circula- 
 tion maintained in the atmosphere and the ocean. 
 
 272. TranBmission of Light through the 
 Atmosphere. — Light travels at the rate of 
 192,000 miles a second, as inferred from ob- 
 
 
 
250 PHYSICAL GEOGRAPHY. 
 
 servations on tlie occnltation of Jupiter's satellites j or 
 184,000 miles according to the most recent experiments. 
 Their obscuration appears to occur about a quarter of 
 an hour (16m, 26s.) later when the earth is at the oppo- 
 site side of its elliptic orbit from the planet, than when 
 it is at the nearest point of that orbit to the planet; 
 and as the diameter of the orbit is about 190,000,000 of 
 miles, the light of the occulted body takes 15 or 16 minutes 
 to traverse that distance. But this speed is liable to bo 
 modified by the density of the atmosphere. By the undula- 
 tory theory, according to which the vibrations of the lumin- 
 ous particles of a body are communicated to the elastic ether, 
 the transmission would be more rapid the denser the medium ; 
 by the theory of emission, according to which luminous 
 particles infinitely minute pass from the source of light to 
 the eye, increased density would retard the light. Practically, 
 the results are the same whatever theory be adopted, since, 
 as Sir John Herschel points out, the extreme difference be- 
 tween the calculations on one or other theory in an extreme 
 case he puts, amounts to ^oloo ^^ ^ second. 
 
 273. Transparency and Colour of Atmosphere. — Trusting 
 to the evidence of our senses, we should regard the air as 
 normally pure, just as we are accustomed to consider it as un- 
 changing. But the pure binary mixture of oxygen and nitro- 
 gen scarcely exists out of the laboratory. The air may be 
 clear, yet instruments of ordinary delicacy will demonstrate 
 the abundant presence of impurities. Tynclall had great diffi- 
 culty in procuring optically pure air for his experiments, and 
 only obtained it by sifting out the inorganic and burning the 
 organic particles which floated in it. The sunbeam that 
 traces a line of light from sky to earth demonstrates, by 
 becoming visible, the presence of substances which diffuse the 
 light. Aqueous vapour is present even in the clearest aii', 
 so that clearness is not an index of dryness. Absolutely pure 
 air, as obtained in tubes from which all the floating particles 
 have been removed, or as seen in stellar space, is colourless, 
 a dark appearance being due to the absence of anything by 
 which light might be diffused. Great variety of colour is 
 afforded by the play of su»light on dust and vapour : blue is 
 the most frequent; its cause — vapour — being the most abun- 
 
Twilight. 
 
 251 
 
 (lant. Tlie purples, reds, and violets of autumn sunsets are 
 doubtless due to the same source j but this whole question of 
 the colours of the sky is one of great difficulty, and is yet 
 far from being decided. 
 
 274. Refraction; Absorption. — A ray of light entering 
 air does not travel directly through it : its course becomes 
 bent in such a way that the rays from an object below the 
 horizon, which should pass tangentially to the earth's sur- 
 face, become deflected and reach the earth. Now, as the 
 eye necessarily estimates the direction of a whole ray from 
 the direction of that last portion of it which enters the eye, 
 as if it were in a straight line from an object above the hori- 
 zon, the sun, or other luminous body, appears to be above 
 the horizon before or after it is so. The refractive power of 
 the air, water being taken as 1, is "000589 71, both substances 
 being examined at 0°C., and under 29*92 inches of mercury. 
 The amount of refraction depends on the condition of the air 
 as regards density and humidity, being gi^eatest when the air 
 is most humid. 
 
 275. Twilight. — To refraction, coupled with the diffusive 
 or reflecting power of the particles in the atmosphere, we 
 owe the phenomena of twilight. 
 
 In this figure, which represents a section of the earth seen 
 from the North Pole, R being east, S west, H P is the 
 horizon line of an observer at N : when the sun is above 
 that line the atmospheric segment H A D O receives his 
 dii'ect rays; when the sun declines to M, the dii'ect rays illumi- 
 
252 PHYSICAL GEOGRAPHY. 
 
 iiate ADO m, but by refraction a part of tlie atmospliere to 
 the left of the line A C still receives direct ligbt, while the 
 remainder of the space H A C N will receive the light 
 diffused from the aqueous and other particles in the directly 
 illuminated atmosphere. Imagining the sun to be rising in 
 place of setting, the phenomena of dawn are of the same 
 kind. The twilight ends when the sun sinks to 18°, or at 
 the utmost 21° below the horizon ; biit the afterglow of the 
 clear Nubian atmosphere is a secondary reflection from the 
 diffused light of a twilight; thus, if when the sun is at K, 
 the twilight is bounded by the line A C, the afterglow may 
 be seen along the line H IST, or even further to the east. 
 
 276. Absorption: Diminution of Light by Distance. — 
 Light travelling through space^ and diverging as it proceeds, 
 imparts to successive objects of equal superficial area an 
 amount which is inverse to the square of the distance ; thus 
 a surface which receives, at the distance of a yard, a certain 
 amount of light, will receive farther off;, at the distance of tAvo 
 yards, a fourth of that amount; at three yards one-ninth, 
 and so on. But this diminution of intensity, which may be 
 accelerated by the intervention of solid or refractive particles, 
 as when mist or dust blocks the way of the light, is entirely 
 different in its results from that absorption by which certain 
 rays are arrested, and cease to give outward signs of their 
 existence. The solar spectrum, which ranges from violet to 
 red, consists, as has been said, of vibrations of unequal 
 lengths. If the yellow flame of common salt is placed in the 
 track of the sunbeam undergoing analysis, the yellow of the 
 spectrum will show dark lines, the absorption bands, which 
 represent the arrest by the sodium flame of those beams to 
 which it gives rise itself : yellow intercepts yellow. A red 
 object appears black in every beam except red. As white 
 light represents the sum of unequal vibrations of the lumini- 
 ferous ether, absorption means the transfer of the motion of 
 the ether to the particles of a body; but these particles must 
 vibrate in the same time ; absorption means, therefore, coin- 
 cidence of vibrations. It would be beyond the scope of this 
 volume to discuss this subject fully, so that its physiological 
 bearing should be fully explained. But the point which the 
 foregoing remarks may induce the student to investigate for 
 
POLARIZATION. 253 
 
 himself is, tliat as the rays of the solar spectrum have very 
 different influences on vitality, the condition of the atmo- 
 sphere may from time to time change, and thus exert an 
 influence on organic beings, while the sources of that influ- 
 ence can only be revealed by instrumental investigations. 
 
 277. Polarization. — Light which enters any transparent 
 body vertically to its surface, passes directly through; but if 
 it enters obliquely, the parts of the ray are successively 
 retarded, so that the light has an oblique course through 
 the body, emerging at the opposite surface of a body with 
 parallel planes at the same angle which the ray had before 
 entering ; emerging at the opposite surface of a triangular 
 prism at an equal and opposite angle to that of entrance. 
 But a body may have its particles so arranged that light 
 passes more readily in one direction than in another. The 
 whole beam is refracted in water as a single beam ; but in 
 ice there are two planes of refraction, the one pei'pendicular, 
 the other parallel to the freezing plane. In ice, and in Ice- 
 land spar, there is double refraction, and the light of the two 
 rays which result from this property has different properties 
 in each ray. The angle of refraction varies for different 
 substances ; it is also obviously unequal for the two parts of 
 the same beam in the case of double refraction. The light 
 now consists of two parts, each of which can only be trans- 
 mitted through another mass of the same substance in exactly 
 the same direction ; if the planes are altered, that particular 
 ray is arrested. The same property is conferred by reflection 
 from an ordinary reflecting surface, as of a mirror. In these 
 cases the extraordinary ray, as that one is called which 
 rotates round the ordinary ray as round a fixed point when 
 the spar is made to revolve, is composed of vibrations of the 
 ether, which take place in one plane only. But certain 
 crystals, as of quartz, have the power of making the vibra- 
 tions move in a circle, and as the component rays of light 
 have different velocities, the result is that a spectrum is 
 formed by the separation of these rays. The details of the 
 mechanism by which these facts are ascertained belong to 
 the domain of Physics. A beam of light, then, falling on a 
 crystal is partly reflected, partly refracted. As the angle of 
 refraction is fixed for every substance, while that of reflection 
 
254 PHYSICAL GEOGRAPHY. 
 
 depends upon tliat of incidence, tlie beam of light must be 
 shifted from a position perpendicular to the face of the crys- 
 tal towards one of parallelism, till the plane of refiection is 
 at right angles to the plane of refraction ; the angle of the 
 incident beam is then the angle of polarization. .: 
 
 278. Polarization of the Atmosphere. — This angle for 
 air is 45° 0' 32". The neutral points are certain spots at 
 which no polarization takes place. Their positions are, " for 
 Arago's point 18° 30' above the antisolar point when the sun 
 is on the horizon; but if the sun is 11° or 12° above the 
 horizon, and the antisolar consequently as much below it, 
 the neutral point is on the horizon, or 11° or 12° above the 
 antisolar point." After sunset, the maximum distance of 
 the neutral point is 25°. Babinet's point is "as much above 
 the sun as Arago's is above the antisolar point." Brewster's 
 point is between the sun and the horizon. The changes in 
 the polarization of the atmosphere are due to fogs, mist, and 
 ice crystals. The influence of atmospheric moisture is im- 
 perfectly known, and the relations of vapour of water still 
 less so ; but it is probable that in the determination of these, 
 meteorology will find invaluable aid for those prognostica- 
 tions Avhich are its most important practical application. 
 
 279. Transmission of Sound. — By experiment it has been 
 shown tha,t rays of heat, capable of inflaming bodies upon 
 which they are concentrated, may be transmitted through an 
 atmosphere at the freezing point; the ether, v/hose vibrations 
 are translated into heat, is thus distinct from the atmosphere. 
 But the vibrations of sound are de2:)endent on the existence of 
 atmosphere. The enormous range of the appreciable sounds 
 has already been referred to as extending from 16 to 38,000 
 vibrations per second. In a previous paragraph (Art. 80), 
 a wave was described as consisting of a condensation and 
 rarefaction; thus the movements of a tuning fork alternately 
 compress and exj^and the air in contact with it, and these 
 movements are transmitted to other portions of air; but while 
 the pulse is thus propagated, the individual j^ar tides of the 
 air move very slightly to and fro. The transmission of the 
 wave takes place in air at the freezing temperature at the 
 rate of 1090 feet per second. But temperature affects the 
 velocity very importantly, as the following table shows, from 
 
INTENSITY OF SOUND. 255 
 
 which it appears that the velocity increases about two feet 
 per second for every 1° of increased temperature : — 
 
 Temperature of Air, Velocity of Sound. 
 
 0-5°C. 1089 
 
 2-10° 1091 
 
 8-5'' 1109 
 
 120* 1113 
 
 26-6° 1140 
 
 280. Intensity of Sound. — The intensity of a sound 
 depends on the character of the atmosphere in which it 
 originates, not of that through which it travels. A cannon 
 fired on the summit of a high hill may be inaudible in the 
 valley, while, if fired in the valley, it would be distinctly 
 heard on the summit: in the one case the initial intensity is 
 less, the air being rarefied; in the other it is greater, the air 
 being denser. But this does not afiect the law of the dimi- 
 nution of intensity, which is, as in the case of light, propor- 
 tional to the square of the distance. 
 
 The rate of transmission of sounds varies in different sub- 
 stances, the variation being determined by the relation of 
 their elasticity to their density. Mertheim found that sound 
 vv^as transmitted by the water of the Seine — 
 
 Temperature 15° C, at the rate of ^47 14 feet per second. 
 Temperature 30° C, at the rate of 5013 feet per second. 
 Temperature 60° C, at the rate of 5657 feet per second. 
 
 Through gases the rate varies widely, being 858 feet per 
 second for carbonic acid, and 4164 for hydrogen. In solids, 
 the rate varies with the elasticity and the temperature of 
 tlie body. The following are taken from a table quoted by 
 Tyndall:— 
 
 Velocity. 
 at20''O. at 100' C. at 200" C. 
 
 Lead, 4,030 3,951 
 
 Copper, 11,66G 10,802 9,690 
 
 Steel Wh-e (English), . 15,470 17,201 16,394 
 
 Cast Steel, 16,357 16,153 15,709 
 
 Iron Wire, 16,130 16,728 
 
 Iron, 16,822 17,386 15,483 
 
 It appears that while increase of temperature diminishes 
 the transmissive power of some metals, as copper, that of 
 iron is increased up to a certain point; and the above table 
 has been quoted for the purpose of suggesting the difiicultiea 
 
256 PHYSICAL GEOGRAPHY. 
 
 in tlie way of estimating tlie velocity of subterranean sounds. 
 In an interesting series of experiments on the transmission 
 of sounds tlirough different kinds of rock, Mr. Mallett found 
 that much broken rock impaired the velocity very consider- 
 ably, and that heterogeneous structure, as of granite, has 
 the same effect. These results are in accordance with the 
 similar fact, that whereas sound travels along the fibre of 
 the wood of a tree at the rate of 15,314 feet, its speed across 
 the rings is only 4567. It is possible, therefore, that the 
 observed velocity of the earthquake noise may be true only 
 for two directions, and that a different conclusion might be 
 drawn from observations made on the progress of the sound 
 at right angles to them. 
 
 281. Sounds more Intense by Night. — Sounds are more 
 intense, and heard at a greater distance, by night than by 
 day. This is not relative, other sounds being less by night, 
 for, in reality, there is more sound audible than by day, but 
 absolute, the cooler air being more uniform, and permitting 
 more of the sound waves to travel direct. 
 
 282. Refraction of Sound. — Sound, like light, may be 
 concentrated by a lens, the lens being a spherical mass of 
 air enclosed in a highly elastic substance, as collodion. The 
 divergent vibrations thus made to converge have the effect 
 of rendering sounds audible at a distance by increasing the 
 number of vibrations in a given space. 
 
 283. Reflection of Sound. — Again, like light, sound may 
 be concentrated by reflection from a concave surface, and the 
 focal length of the sound mirror accurately determined. The 
 phenomena of echoes are illustrations of reflection, the multi- 
 plication of the surfaces multiplying the repetitions, as the 
 facets of a prism multiply the reflected images. 
 
 284. Resonance. — But the reflection of sound from a solid 
 body is distinct from resonance, which is the intensifying of 
 a sound by the synchronous vibrations of a mass of air, which 
 thus multiplies the vibrations, or rather multiplies the points 
 whence they proceed. Echoes proceed from concave surfaces, 
 which may be resonant only to sounds of a particular pitch. 
 It is to resonance that the thunder-like sound of a gun fired in 
 a narrow valley is due, while the sound may be reflected only 
 from one or two points. 
 
REFLECTION OF HEAT. 257 
 
 285. Transmission of Heat.— Air in contact witli a warm 
 hodj lias its temperature raised by conduction or convection. 
 In the former case, the heat radiated from the body, or, in 
 the language of the mechanical theory, the motion of its 
 particles transmitted to those of the air, is j)assed on through 
 the air by each particle imjDarting movement to the next. 
 Convection is the transfer of this motion, not by single 
 particles, but by groups of particles : masses of air move 
 upwards, other masses take their place, till the loss of motion 
 by the radiating body is equal to the gain by the air. Equi- 
 librium and rest are then arrived at. But this equilibrium 
 would not exist were the air perfectly dry and perfectly pure ; 
 the body would cool by radiation into space; the heat would 
 be thus lost, but the air would receive none of it. The air 
 is therefore in theory neutral to heat rays, it neither absorbs 
 nor radiates. But, as a matter of fact, since it is never pure, 
 it both absorbs and radiates; and, as has been already said, 
 stops 10 jDer cent, of the terrestrial radiation within 10 feet 
 of the earth. Conduction and convection, therefore, are 
 phenomena of air which is not chemically pure. 
 
 286. Reflection of Heat : its Diminution by Distance. — 
 In these two particulars the analogy of heat and light is 
 complete, since the heat or calorific rays may be concentrated 
 by apparatus of the same character as is employed in the 
 case of light; and the ratio of diminution is likewise inverse 
 to the square of the distance. 
 
 SECTION II.— ATMOSPHEHIC CIECULATION. 
 
 Movements of Atmosphere : Influence of Earth's rvotation— Theories 
 as to the Cause of Currents : Hadley and Maiuy — OI)jections — 
 Constant Westerly Current at High Altitudes — "Winds of N. 
 Atlantic Basin : N.E. Trades — Mediterranean — Local Winds — 
 Sirocco : Fohn : Analogous Wind in New Zealand — West Winds 
 of N. Atlantic — Calms of Tropic of Cancer — Hurricane Region — 
 Winds of N. America: S.E. Trades — Pacific Ocean — Indian 
 Ocean: Monsoons — Monsoons in other Regions: S. America; 
 Chinese Seas; Western N. and S. America; S. Africa — Course 
 of Atmospheric Currents — Land and Sea Breezes — Velocity of 
 Wind — Storms — Storms of Acceleration — Tornadoes — Rotatory 
 23 R 
 
258 PHYSICAL GEOGRAPHY. 
 
 Storms : Hurricanes, Cyclones, Typhoons, etc. — Velocity of 
 Botatory Storms : their Area — Storm Waves — Whirlwinds : 
 Waterspouts : Dust Storms — Simoom. 
 
 287. Movements of Atmosphere : Influence of Earth's 
 Rotation. — As the air forms a layer round the globe, and as 
 it is practically a fluid, the movement of the globe within 
 this sphere necessarily involves movement likewise of the 
 investing atmosphere. The globe, rotating on its own axis 
 from west to east, and rubbing against the lowest portion of 
 the atmosphere, sets the whole superjacent mass moving in 
 the same direction. 
 
 But as the sectional area of the globe diminishes towards 
 the poles, it follows that the movement of the air diminishes 
 in rapidity as we approach the poles. That is to say, the space 
 which the air has to travel at 30'' of latitude is less than that 
 which the air at the equator has to travel ; thus the air at 
 the 15 til parallel of latitude has to travel through 869 miles 
 in an hour, while air at 30° N. lat. passes through 450 miles 
 in the same time. As, therefore, the amount of space to be 
 passed through diminishes as the latitude increases, and as 
 eaoli particle of air has the velocity of the point of earth 
 witii which it is in contact, if by any disturbance a particle of 
 air is sent towards the pole, or towards the equator, it lapses, 
 on the one hand into an area of slower, on the other hand 
 into an area of more rapid, movement. Suppose that a par- 
 ticle of air having an eastward movement, at the rate prop)er 
 to the 45th degree of latitude, is directed towards the pole, 
 it will move eastwards more rapidly than the points of the 
 earth over which it passes, and hence it will appear to have 
 a motion very nearly eastward. If^ on the other hand, it 
 moves towards the equator, its rate will be less than that of 
 the regions into which it enters, and it will thus appear to 
 have a westward dii'ection; it lags, so to speak, behind the 
 mass of the earth. 
 
 238. Theories regarding the Causes of Currents: Hadley 
 and Maury. — It has been assumed that the two influences 
 under which the circulation of the currents of the atmosphere 
 take place are heat and the rotation of the earth, heat l)eing 
 the primary oiiginator of its currents ; and Hadley's theory 
 £isserts, (1) that the trade winds which move from the poles 
 
THEORIES REGARDING THE CAUSES OP CURRENTS. 259 
 
 towards the equator, and the counter trades which move from 
 the equator towards the poles, are due to the relatively high 
 temperature of sqnatorial regions as compared with that of 
 polar regions; (2) that the westward tendency of the trades, 
 and the eastward direction of the counter trades, are due to 
 the rotation of the earth. This is the commonly asserted 
 doctrine, and in accordance with it the course of currents in 
 the atmosphere is stated to be after the following order : — 
 1, The over-heated air at the equator rises vertically into the 
 atmosphere, and spreads itself towards either pole, while the 
 colder, and therefore heavier, air from the north travels 
 towards the equator to take its place. 2. The ujoper stratum 
 of air, when it reaches about the 35th parallel of latitude, 
 has parted with its excess of heat, and descends towards the 
 surface of the earth. 3. Part returns again towards the 
 equator, and part passes on towards the pole, forming ■ the 
 south-westerly winds of high latitudes. Maury further 
 imagines that the air, having again reached the equator, 
 becomes heated, but does not part with its onward move- 
 ment, and, m fact, continues its progress across the line to 
 one or other pole, as the case may be, whence it again returns 
 in the same undulating line and completes the surface of the 
 giobe^ the currents thus pictured forming a series of figures 
 of eight, the nodes of w^hich, or the points where the currents 
 meet the surface of the earth, constituting the three bands of 
 calms — the equatorial, and those of Cancer and Capricorn. 
 
 289. Objections to these Views. — Several objections must 
 be taken to this view : in the first place, the regions of 
 calms and variable winds do not form continuous zones all 
 round the earth ; they are like the dead waters of the Sar- 
 gasso Sea in the Atlantic, and of its counterpart in other 
 oceans, limited in dimensions, and do not touch the land on 
 either side. In the second place, if excess of heat were the 
 determining ca,use of the atmosj^heric movements, the north- 
 easterly trades of the Atlantic would be converted into west 
 winds blowing across the continent of Africa, since the in- 
 terior of that continent has a temperature at times 27° higher 
 than that over the Atlantic ; for while air in contact with 
 the sea at the equator seldom exceeds 27-7° C. (82° F.), as 
 much as 544° C. (130° F.) has been recorded in the interior 
 
260 PHYSICAL GEOGRAPHY. 
 
 of Africa. Again, westei-ly winds blow in tlie polar regions 
 both in winter and in summer, although at the one season 
 the ocean area, at the other the land area, is the warmer. 
 Neither can the rotation of the earth be admitted to influence 
 to any great degree the movements of the air currents; be- 
 cause we find that the air actually travels from north-west 
 to south-east, or from south-east to north west in the northern 
 hemisphere; and from south-west, or from north-east, in the 
 southern hemisphere, beyond the 45th parallel, although 
 these directions are contrary to what should take place ac- 
 cording to theory, being polar not equatorial ; and it is certain 
 that the air acquires directly the motion of the earth at the 
 point with which it comes in contact. It seems more probable 
 that the movements of the atmosphere are, as a whole, from 
 west to east, and that the variations from this dominant 
 direction are local in their origin, and are the counterpart of 
 the movements which take place in the ocean. 
 
 290. Constant Westerly Current at High Altitudes. — 
 It has been frequently ob^^erved that ashes from tropical 
 volcanoes travel to windward, contrary to the direction of the 
 steady easterly and north-easterly winds, and that they have 
 been thus transported for hundreds of miles. Frequent 
 observations have been recorded ot the upper clouds travel- 
 ling eastwards, contrary to the movements of the lower 
 strata ; and the observations of many travellers coincide with 
 those of Professor Smyth on Teneriffe, to the efiect that strong- 
 westerly winds prevail at high altitudes, while easterly winds 
 are travelling along the surface of the earth. So far, tliei-e- 
 fore, as these observations go, there seems no reason for 
 doubting that the upper current has a westerly movement in 
 high altitudes, and that that movement is for the most part a 
 very rapid one. In the higher latitudes, both north and 
 south, westerly winds prevail, and these are subject, as they 
 approach lower latitudes, to various deflections, which are the 
 counterpart of those observed in the case of the easterly drift 
 of the Antarctic Ocean. 
 
 291. Winds of the Atlantic Basin : NE. Trades.— For 
 the most convenient description of the various winds which 
 predominate at difierent points of the earth's surface, it will 
 be best to follow the same geographical arrangement as was 
 
MEDITERRANEAN". 261 
 
 adopted in the case of the ocean circulation. In the North 
 Atlantic, between ] 0° and 25° N". lat., the north-east trade pre- 
 vails, the belt thus limited advancing towards or receding 
 from the equator according to the seasons, being nearest to 
 the equator in January, farthest from it in July. Between 
 15° and 20^ the trades blow steadily throughout the year; and 
 this fact is recognised by navigators, Avho know this region as 
 the heart of the trades. Towards the northern limit of this 
 belt the currents have a more northerly direction, and 
 towards the American side the westAvard dii-ection is more 
 conspicuous, this variation being due to the influence of 
 rotation. The winds, however, upon the African shores 
 are deflected at Cape Yerde, and become north-west winds, 
 passing southwards towards the Gulf of Guinea. On the 
 African continent, easterly and north-easterly winds prevail; 
 and under their influence the sand storms of the desert are 
 carried out seawards, the red dust being frequently deposited 
 upon ships at considerable distance from land. Harmattan 
 is the local name of the north-east wind blowing seawards 
 from the Sahara to the south of Cape "Verde in the winter 
 months, especially January and February. The conA^ersion 
 of the trades into easterly winds, towards the West Indian 
 Islands, renders them important auxiliaries of the equatorial 
 current, which sets towards the Gulf of Mexico. On the 
 east side of the Atlantic, beyond the limit of the trades, nor- 
 therly winds, chiefly north-Avest, though sometimes also north- 
 east, preA"ail, and influence navigation importantly, since 
 homeward-bound vessels, returning from Spain, are obliged to 
 go 20° to the we'st before they can turn northwards, though 
 the outward voyage is facilitated by them. 
 
 292. Mediterranean. — The dominant direction of the Medi- 
 terranean winds is from the north. Local names are given 
 to particular winds : thus, the north-west wind of the Gulf, of 
 Lyons and West Italy is knoAvn as the Mistral; the north 
 wind of the Adriatic is the Bora; the Gregale is the north- 
 east wmd which strikes Malta; and the Archipelago is visited 
 by the Etesians, or north-east Avinds of summer; Avhile the 
 Tramontana is the winter Avind of that region. In the Levant 
 I the summer Avdnd is from the north-east, that of spring from 
 the north-Avest; and along the north coast of Africa the 
 
262 PHYSICAL GEOGRAPHY. 
 
 easterly wind known as a Levanter is tolerably steady. The 
 Sirocco is a remarkable exception both to the direction and 
 the character of these Mediterranean winds. It travels from 
 the south-east, carrying with it the red sand of the desert, 
 andj crossing Malta and Sicily, strikes upon Italy, producing 
 there very remarkable physical and physiological effects. It 
 is hot and damp; the temperature rises as high as 35° C, and 
 its severely depressing effects are manifest both on animal and 
 vegetable life. Its character is consequent upon the elevation 
 of the Sahara into dry land, and there is every reason to 
 believe that to its influence may be traced the shrinking of 
 the Italian glaciers upon th6 southern slopes of the Alps. It 
 has been calculated that if the Sahara were again laid under 
 water, the plains of Lombardy would cease to be the richly 
 productive regions we now see them. 
 
 293. Causes of Sirocco: Fohn.— The sirocco crosses the 
 Alps and becomes the fohn, or south-west wind of Switzerland. 
 But it has entirely changed its character; and as the history 
 of this current illustrates several important points, it deserves 
 some attention, more especially as it is even yet a subject 
 of discussion among meteorologists. The JST.E. trades in the 
 Atlantic recede northwards during summer, and the coast 
 winds of Africa are at that season frequently from the west. 
 The north-east winds of the continent are thus impeded in 
 their seaward progress, even blown back, and forced to find 
 an escape to the north as the sirocco over Italy, the solano 
 in Spain. The sirocco blows at all periods of the year, but 
 most frequently in spring and autumn; and at these periods 
 the African interior receives supplies of air, Avarm and moist, 
 from both coasts, from the Gulf of Guinea, and from the Red 
 Sea, as v/ell as from the Mediterranean. It thus starts with 
 a considerable quantity of moisture, which is increased as 
 its over-heated lower strata pass over the sea. Precipitation 
 takes place in the north of Italy: the wind, however, goes 
 on, rises over the Alps, and, cooled during expansion, re- 
 gains a higher temperature on the low gTounds to the north, 
 where it resumes its former density. But the temperature 
 is not quite so high, since the vapour which it still carried 
 over the heights would permit radiation into space, and this 
 heat would not be regained by renewed density. 
 
HURRICANE REGION. 263 
 
 294. Transmontane Wind in New Zealand. — The plieno 
 mena attending its passage over the Alps are exactly parallel 
 to tliose observed in New Zealand, where a north-west Avind 
 throws down rain or snow on the coast and hill face on which it 
 strikes, though this is not always the case, as the precipitation 
 sometimes takes place before the land is reached. On the 
 south- west face of the range, the wind descends the valleys with 
 heat sufficient to cause floods in the glacier streams by the 
 sudden melting of the ice. 
 
 In the northern part of the Atlantic basin, north that is 
 to say of 35°, west winds prevail, and the " Roaring Foi'ties," 
 about 40° N. lat., are tolerably steady in their direction, 
 though in winter they reach the British coasts often with 
 very great violence. The years of greatest loss, by shipwrecks, 
 on the west shores of the British islands are those in which 
 these westerly winds have attained their greatest violence. 
 It may be said that generally west winds prevail over Europe, 
 and reach as far as the Black Sea,, to the south of which the 
 northern tendency reappears. 
 
 295. Calms of the Tropic of Cancer. — Over an area 
 generally similar to that of the Sargasso Sea, to the north, 
 that is to say, of the trades, in 30^ to 35^ W. Ion., there 
 is a variable region known to sailors as the Horse Latitudes, 
 which lies in an area bounded by northerly winds on the 
 east and west; by the west winds to the north; by the 
 trades to the south. The winds in this region may come 
 from any point; they may blow with great violence, or may 
 be succeeded by calms. This change in the character of 
 the winds increases the analogy to the Sargasso Sea, whose 
 shifting position is due to the encroachment of one or 
 another current. 
 
 296. Hurricane Region. — The hurricane district of the 
 "West Indies comprises the islands from Barbadoes north- 
 westv/ard, even to Mexico, and north-eastwards to about TC 
 "VV. Ion,, and there is reason to believe that their influence 
 is felt indirectly upon the shores of Europe. The general 
 track of the hurricanes is from a point to the east of Bar- 
 badoes, in 10*^ to 15® N. lat., thence to the west north-west 
 as far as Florida, and northward to about 40° N. lat., where 
 they die out. They are sometimes continued westwards into 
 
264 PHYSICAL GEOGKAPHY. 
 
 the Gulf of Mexico, but never affect the continent beyond 
 the Alleghanies. 
 
 297. North American Winds. — The continent consists of 
 a central valley lying between the Kocky Mountains and the 
 coast ranges, but open to north and south. The Rocky 
 IMountains cut off well-nigh entirely the westerly winds, and 
 a northerly direction prevails in the central trough as far 
 as about 35° N. lat. ; south of that the direction is unceas- 
 ingly from the south-west, though south-east i^ frequent at 
 the foot of the Rocky Mountains, the wind following the 
 curve of the high ground. The winter winds are chiefly from 
 the north-west, a direction observed even in the Caribbean 
 Sea. These cold winds, which blow down the whole American 
 valley, become more and more westerly among the islands, 
 whence they pass to rejoin the westerly Avinds of the North 
 Atlantic. They are the northers of the Mexican Gulf, the 
 counterpart of which exists on the homomorphic western 
 coasts of the Pacific. 
 
 298. South-east Trades. — The south-east trades are very 
 steady, and in general stronger than the north-east. They 
 occupy a belt, likewise variable in position, between 0° and 
 25*^ S. lat. ; in winter they reach 5'^ N. lat. On the African 
 coast they are more southerly in direction, and in the Gulf 
 of Gumea westerly Avinds prevail, the direction shifting from 
 west north-west to west south-west, and calms are of frequent 
 occurrence. This region of variable Avind extends as far as 
 S"^ W. Ion., and the names by which it is knoAvn are: Region 
 of Equatorial Calms, Region of Variable Calms, Region of 
 "Variable Winds and Calms, Region of Constant Precipitation, 
 Doldrums, or the Rains of earlier navigators. This region, 
 in fact, corresponds generally with the triangle left between 
 the two roots of the equatorial ocean current, and is thus 
 an atmospheric dead-water, in which west winds struggle 
 for predominance. To the north and south of it the trades 
 sometimes coalesce into an east wind, blowing towards the 
 Caribbean Sea. On the opposite sides of the S. Atlantic basin, 
 opposite tendencies exist, the northerly prevailing on the 
 American shore, the southerly prevailing on the African side, 
 thouoh these are converted into west winds durino: winter. On 
 the southern continent of Africa^ as on the northern, the trades 
 
PACIFIC OCEAK. 265 
 
 are, as a rule, distinctly felt. On the American continent con- 
 siderable variations, however, exist; thus, the trade passes up 
 the Amazon valley as an east wind, which from July to Janu- 
 ary — the dry season — blows with increasing strength ; but in 
 the wet season dies down almost to the coast line. In Para- 
 guay north-east winds pi'e vail, north-west being rare. But the 
 l^amperos, farther to the south, are north-west, west, or south- 
 west winds, which blow across the pampas, often with great 
 and sudden violence. From 40* S. lat. the westei'ly circum- 
 polar winds prevail as stead^r and strong currents which^ 
 near Cape Horn, form westerly or south-westerly gales. 
 Another region of calms exists between 0° and 1 5"^ W. Ion. ,27*^ 
 and 37*** S. lat., and represents an ellipse whose position is 
 variable, and whose long axis is from north-west to south-east. 
 299. Pacific Ocean. — The Pacific Ocean manifests the 
 same westerly currents in high latitudes as does the Atlantic. 
 The north-east trades form a belt between 0' and 20'' N. lat. 
 The south-east trades range from 0* to 25° S. lat. The 
 north trades are not so steady as those of the Atlantic, and 
 both they and the south trades have a more easterly direction. 
 The north-east trade, striking upon the Philippine Islands, 
 acquires a southerly direction, thus following to some extent 
 the coiu'so of the Japan current. In the north China seas 
 the winds undergo seasonal changes, being south-west in 
 summer, and north-east from November to April; and, to 
 complete the resemblance which this region presents to the 
 corresponding portion of the American continent, we have 
 hurricane regions in which typhoons occur between August 
 and October, their position being from 10° to 23° N. lat. 
 To the seasonal variations the term monsoon is applied, 
 although the phrase is more strictly applicable to the variable 
 winds of the Indian Ocean. As in the latter region, the 
 south-west monsoon is the wet summer wind, the north-east 
 is the dry wind of winter. On the American side, the trade 
 winds go farther to the north, and have, as on the corre- 
 sponding shores of Africa, a more northerly direction, even 
 acquiring a north-westei'ly inclination. On the Mexican 
 coast, the winter winds are north or north-west, the summer 
 south-west or south-east, but across the open ocean westerly 
 winds prevail above 40° of latitude; and between this region 
 
M6 PHYSICAL GEOGRAPHY. 
 
 and the sliifting limits of the trades, a vai'iahle region exists, 
 which does not, however, reach the land upon either side. 
 In the southern Pacific Ocean, the south trade loses strength 
 westward, and among the islands is replaced by irregular 
 westerly winds. On the Chilian coast, monsoons vary from 
 south and south-west in summer to north, commencing in 
 March; and here, as in many other cases, the polar wind is 
 dry, the equatorial moist. But, on the Peruvian coast, the 
 summer wind is south-west and moist, though heavy fogs 
 accompany the rare north winds. In winter, the north wind 
 blows from the Caribbean Sea as a papagayo or tehuanfcepecer. 
 
 300. Indian Ocean. — The winds to the north of 12° S. 
 lat. in the Indian Ocean, are alternately from north-east and 
 south-west, and are hence known as monsoons, though that 
 term is originally applied only to the alternating winds on 
 the Arabian coast. The south-east trade blows in winter to 
 the equator; but, in summer, its northern limit retreats to 
 12^^ S. lat. In winter, it is more and more easterly towards 
 the equator; but, in summer, the area of 12° which it has 
 deserted, is traversed by winds from the west, north-west, 
 or south-west. The trade shifts its southern limit also, 
 and thus the variable region, the calms of Capricorn, which 
 separates it from the constant Antarctic west winds, is 
 carried up and down. This region is only a patch separated 
 from Australia by the winds which blow northwards to feed 
 the trade, while the African winds again limit to the west. 
 We have here another example of an atmospheric backwater, 
 in which the westerly winds struggle to prevail, and the 
 weather is thus very irregular and broken. 
 
 Summer and winter are those of the north and south hemi- 
 spheres respectively. 
 
 301. Monsoons. — It is impossible, within the compass of 
 this chapter, to make clear the great irregularity of the 
 movements Avhich are comjDrised under the one phrase, 
 monsoons. It is usual to state their character summarily, 
 as that of a v>rind which, from May to September, blows 
 from the south-west; and from October to April, from tlie 
 north-east. But this is only true for a few localities, as the 
 coast of Arabia. It may be said, in general terms, that the 
 winds north of the equator, from October to April, are 
 
MONSOONS. 2G7 
 
 northerly, chiefly from the east of north, but even from 
 the west of north; and that the winds from May to September 
 are more markedly westerly in the open sea. The deflec- 
 tions may be grouped according to the coast lines. Thus the 
 parallel lines, from Calcutta south-eastwards and from Cutch 
 to Cape Comorin, from Calcutta to Ceylon and the west side 
 of the Indian Ocean, show parallel movements of the winds. 
 On the Malabar and Burmah coasts, the N.E. monsoon, which, 
 at sea is steady, blows from north-west; the S.W. monsoon 
 is from the west, or north of west, on the Malabar coast; it 
 blows north-west past Point de Galle, and returns as a south- 
 easter to the Coromandel coast, though it is more strictly 
 south in the open sea; and again south-west when it blows 
 inland at the north end of the bay. The Red Sea, during 
 the N.E. monsoon, i.e., from October to April, is traversed 
 by a south-east wind; and from May to September, by a north- 
 west. Some of the summer variations are tabulated below. 
 
 N.W., . . . Gulf of Oman. 
 
 S.W.; S.S.W., . S.E. Coast of Arabia. 
 
 N.; N.E., . . EedSea. 
 
 W.; S.W.; S.E., . Bay of Bengal (Avest and north). 
 
 W. ; N., . . . Bay of Bengal (east coast). 
 
 W.; W.N.W., . Malabar Coast. 
 
 N.W.; W.N.W., . Cape Comorin. 
 
 N.W., . . . Kilglierries. 
 
 S.W.; W.; KW., . Off Hooghly. 
 
 The similarity of the movements on parallel coast lines, 
 and the remarl^able variations in the Ked Sea, indicate the 
 power of local geographical conditions as superior to any 
 general influence, such as rotation of the earth. 
 
 To the south of the equator, the south-east trade is only 
 variable in position, is not reversed in direction by the 
 seasons; this is true at least for the area from 12° to 25° S. 
 lat. The seasonal reversal between 0° and 12°, gives during 
 winter the eastward movement, which may be regarded as 
 the extreme deflection of the south-east trade; while in 
 summer a westerly wind traverses the ocean, and passes into 
 the Pacific through the island channels. This westei-ly wind, 
 the N.W. monsoon of the South Indian Ocean, is more 
 northerly near Madagascai', and even may blow from the 
 east of north. The influence of local features is again illus- 
 
268 PHYSICAL GEOGRAPHY. 
 
 trated in tlie winds of the Mozambique Channel, which blow 
 during winter from the south-west, during summer from the 
 north-east. Now, as the wind blows along the coast of 
 Africa from the mouth of the Arabian Gulf as a north- 
 east wind during the months October to April, and as in 
 summer the S.E. trade retreats to 12° S. lat., the Mozam- 
 bique wind of that season is a jorolongation of the N.E. 
 monsoon of the North Indian Ocean, while the winter wind 
 is the deflection of the S.E. trade, which at that season 
 blows to the equator. 
 
 302. Monsoons in other Regions. — The word monsoon 
 is now used for alternating winds in other regions; and as 
 this custom is becoming popular, it would, perhaps, be well 
 to use the term, in books on Physical Geography, for any 
 winds at any locality whose direction shifts with the seasons, 
 and which divide the year, however unequally, between 
 them. 
 
 Using the term, then, in this wider sense, we have monsoons 
 in the China seas, in the Mexican Gulf, on the coasts of 
 Africa, and South America. 
 
 303. S. American Monsoons. — The N.E. trade reaches 
 the north-east of South America, north of the equator, from 
 December to April; but, for the rest of the year, the venda- 
 bales, wet westerly or south-westerly winds, alternate with 
 south-easterly winds along the coast. There is here repeated, 
 and for exactly the same reason, what haj)pens on the east 
 coast of Africa in the Mozambique Channel. The northern 
 summer carries the trades to the north, and the S.E. trade 
 blows into the Caribbean Sea, while the southern summer 
 brings the N.E. trade nearer the equator. In the Gulf of 
 Mexico itself an alternation of the same kind is seen, but 
 its periods are not sharply defined; and the north Avinds 
 come down the Mississippi valley, their movements having 
 reference to other influences. 
 
 304. Chinese Monsoons. — From October to April, the 
 Formosan winds are steady at north-east; and, like the polar 
 winds on the American side of the Pacific, they are dry and 
 clear. For the rest of the year, with the exception of the 
 typhoon months, the wind is from the south-west, and is 
 even more southerly at the Phili23pine Islands. These regular 
 
COURSE OF ATMOSPHERIC CURRENTS. 269 
 
 seaso]ial alternations extend, there is every reason to believe, 
 into the interior, the floods of the Yang-tse-Kiang not being 
 explicable by the melting of snow, which does not lie on the 
 Thibetan high gronnds; so that a wet season, that of the 
 southerly or south-westerly monsoon, is the only other source 
 of the excessive moisture. 
 
 305. Monsoons of Western N. America. — From June 
 till October, or rarely November, the winds are, on the west 
 coast of Mexico, from south, south-west, or south-east, and 
 are wet, as equatorial winds are for the most part. In winter, 
 over the same area, north-westerly winds prevail, and may 
 reach even to the equator, if they pass their customary limit 
 about 10° N. lat. 
 
 306. Monsoon of Western S. America and S. Africa. — 
 The north-west winds, with much moisture, increase in their 
 duration as we advance southwards on the American coast, 
 till in Patagonia westerly winds occupy most part of the 
 year ; but these are derived from the westerly ^^^Jlds of high 
 latitude, and the alternate preponderance of the northern 
 or southern element is recognisable. But the winds on the 
 American continent liaA'e at once greater obstacles to encounter 
 in the mountain chains, and greater freedom of escape over 
 the open Pacific. Hence their movements have much less of 
 periodicity than where, as in the Indian Ocean, the space is 
 more restricted. In the South Atlantic, the African coast is 
 lower than that of America, while the great projection of 
 the northern portion acts more effectually than the similar 
 prominence of western North America, in checking the 
 movements of the air and forcing the currents more into the 
 interior. Hence the African coast is characterised chiefly 
 by the frequency of its calms. 
 
 307. Course of Atmospheric Currents. — There are two 
 startmg points from which to trace the winds. According 
 to the common notion, the starting point is the trades, which 
 are the indraught of cooler air to replace the warmer air 
 which ascends over the thermal equator, or line of greatest 
 heat. This in July reaches as far as 10*^ north of the 
 equator, and in January to 2° or 3° south of the equator. 
 The ascending current over this region, then, is the starting 
 point, whether or not the existence of those undulations 
 
270 PHYSICAL GEOGRAPHY. 
 
 already refeiTecl to (Art. 288) is believed in. But, on the 
 other hand, the region of a steady westerly wind seems a 
 much more reliable area to start from, than one in which, 
 so far from a vertical cii'cnlation being proved, easterly and 
 westerly winds struggle, and calms frequently indicate the 
 absence of all motion, while excessive rainfalls likewise tell 
 of arrested motion in the air. Starting, then, with the 
 westerly current as consequent on the rotation of the 
 ea.rth, we find that it strikes the level of the sea at 40*^ lat. 
 N. and S. Its progress eastward is arrested by the shores 
 of the Atlantic and Pacific, But these present very unequal 
 obstacles. The mountain backbone of America is low only, 
 in the centre. To the north and south it stops almost 
 entirely every direct influence from the west. The eastern 
 shores of the Atlantic, on the other hand, consist chiefly 
 of low grounds, or of chains which run from east to west, 
 and thus present less powerful obstacles to the progress of 
 the current. The wind passing over Europe is deflected by 
 the central mountain chains through gaps in Avhich the 
 mistral, the bora, and other north-westerly currents pass, 
 while the Ourals and the colder, denser air of north and central 
 Asia check its eastward progress, the eastern and western 
 direction of the great ranges permitting its free passage as far 
 as the western rampart of the Thibetan plateau. From Hay- 
 ward's explorations, it appears that the winds of eastern 
 Turkestan are dominantly from the west, and that there are 
 passes through which these winds enter the country to the 
 north of the Himalayas. But to the south of the sources of 
 the Kashgar river, a mountain range, with peaks of upwards of 
 21,000 feet in height, forms a wall which reaches the Hindu 
 Kush, and thus gives that southerly direction which Burnes 
 records as prevalent in Bokhara. Thence it may be followed 
 to the south-east, till it enters the low ground south of the 
 Himalayas, and descends the main valleys of the Ganges and 
 Indus. Its eastward direction is resumed on the east sides 
 of the Indian Ocean, but it divides on the north-east corner 
 of Africa, part turns south-westwards with the African coast, 
 while the other part follows the hollow of the Bed tSea. 
 The S.W. monsoon corresponds in time with the easterly 
 monsoon which blows south of the equator. But when 
 
COURSE OF ATMOSPHERIC CURRENTS. 271 
 
 the nortli-westerly monsoon, or more strictly tlie westerly 
 monsoon, blows in the summer of the south hemisphere, the 
 N.E. monsoon is blo\ving in the north; the two winds, there- 
 fore, intervening between the N.E. monsoon and the S.E. 
 trade, are in reality different phases of the same westerly 
 wind, Avhich in the narrower Atlantic fails to attain the 
 same fetch, and thus blows with less strength in the open 
 ocean; in the Indian Ocean, moreover, it has an escape 
 throu£>"h the insular channels, The N.E. monsoon is in 
 position the equivalent of the N.E. trade, and thus compar- 
 ing this with the two great oceans, the equatorial calm 
 and variable region is here greatly enlarged, sv,dnging alter- 
 nately to the north and south of the equator through a 
 greater range 
 
 The deflections of the trades have still to be accounted for. 
 If there is no continuous range of mountains across Africa, 
 there seems some obstacle sufficient to prevent the European 
 winds passing south, and to foi'ce them -s^estwirds to join the 
 trades, which, however, start farther north than this source 
 of supply. The air over the land is appealed tj as, in defect 
 of other more poweiful barricades, sufficient to deflect a cur- 
 rent. It would appear as if the equatorial belt of westerly 
 wnids and calms presented such a case; the trades acquire 
 increased easting as they approach the equator, and the belt 
 shifts with them to north and south. As that belt is one of 
 low barometeric pressure, it must be only difference in quality 
 of atmosphere which prevents its being torn away and 
 absorbed into the trades on either side; the contrary takes 
 place in the west of the Atlantic and Pacific oceans, where 
 the trades contribute to the westerly winds which sometimes 
 blow in the equatorial calm region. 
 
 The order here indicated is founded on theoretical con- 
 siderations, which have not yet met with entire acceptance. 
 Into the arguments in support of the arrangement it is 
 impossible here to enter. The student will find them dis- 
 cussed in Laughton's work already referred to, and in various 
 communications in journals. It is only necessary to say 
 further, that, while the primary motion is ascribed to a 
 westerly current, and the great majority of the phenomena 
 are easily strung on this theoretical thread, the auxiliary 
 
272 PHYSICAL GEOGRAPHY. 
 
 influence, under certain circumstances, of pressure, and differ- 
 ence of temperature, is by no means excluded. 
 
 308. Land and Sea Breezes. — Thus the alternate move- 
 ment of air to and from the land, by day and night, seems 
 due to the combined influence of increased pressure and 
 difference of temperature. The air over the sea receives, as 
 the heat increases, a large amount of vapour by evaporation, 
 and thus acquires greater elasticity than that over the land, 
 upon which it thrusts itself. During the night, when evapo- 
 ration is checked over the sea, the air, cooled down by radia- 
 tion and by diffusion, loses more elasticity than that of the 
 land, which thrusts itself out seaward. A certain approach 
 to equilibrium between the sea and land air, however, is 
 needed; thus these breezes are not known where the shores 
 are bare of vegetation, and the temperature of day and night 
 are thus very different. The east wind of night on the 
 Tapajos as observed by Bates to replace the west wind, 
 doubtless comes from the interior, and thus gives a transition 
 from the diurnal variation to a seasonal variation, of which 
 many examples occur in the monsoon area on a small scale. 
 
 It is again, as in the case of ocean currents, defective 
 knowledge or extreme statements which give rise to contro- 
 versy; and as there is still uncertainty as to the machinery 
 of atmospheric circulation, that view has been here adopted 
 which gives greatest coherence and simplicity to the subject. 
 The student must bear in mind that it is a question of how 
 much influence each agent may fairly be credited with, and 
 upon these amounts the most conflicting statements are made. 
 
 309. Velocity of Wind. — The same necessity for reference 
 to a fixed standard exists in the case of wind, as of heat and 
 light; perhaps even greater, since the circumstances under 
 which excitement is likely to make the unaided senses worth- 
 less giiides, are those against which it is most important to 
 guard, by learning most accurately every detail of their 
 occurrence. Wind gauges of various kinds are used, and 
 these of course give sure results. But in defect of these, it 
 is necessary to have a scale commonly intelligible, by reference 
 to which the risks of error are diminished. The Beaufort 
 scale recognises thirteen grades of movement, which are 
 determined by reference to the speed of a ship or the sails 
 
STORMS. 
 
 273 
 
 slie can safely cany. But it is more common to use a scale 
 of seven grades, — 6, between which the intermediate steps 
 are marked '5. The formulae for calculating the velocity V, 
 and pressure P, are P = Y^ x -005; and V = \/200 x P. The 
 pressures and velocity are given in the following table, which 
 is copied from Buchan, p. 211. 
 
 4) 
 
 Pressui-e. 
 
 Velocity. 
 
 
 t) 
 02 
 
 lbs. per 
 
 miles per 
 
 
 0—6 
 
 sq. foot. 
 
 hour. 
 
 
 0-0 
 
 0-00 
 
 
 
 
 
 0-5 
 
 0-25 
 
 7-1 
 
 1 
 
 1-0 
 
 1-00 
 
 14-1 
 
 2 
 
 1-5 
 
 2-25 
 
 21-2 
 
 3 
 
 2-0 
 
 4-00 
 
 2S-8 
 
 4 
 
 2-5 
 
 6-25 
 
 35-4 
 
 5 
 
 3-0 
 
 9-00 
 
 42-4 
 
 6 
 
 3-5 
 
 12-25 
 
 49-5 
 
 7 
 
 40 
 
 16-00 
 
 56-6 
 
 8 
 
 4-5 
 
 20-25 
 
 63-6 
 
 9 
 
 5-0 
 
 25-00 
 
 70-7 
 
 10 
 
 5-5 
 
 30-25 
 
 77-8 
 
 11 
 
 6-0 
 
 36-00 
 
 84-8 
 
 12 
 
 Beaufort's Scale. 
 
 Calm, . . . . 
 
 Light air, . . . 
 Light breeze, 
 
 Gentle breeze, . 
 Moderate breeze, 
 
 Fresh breeze, 
 
 Strong breeze, . 
 
 Moderate gale, . 
 
 Fresh gale. . . 
 
 Strong gale, . . 
 
 Whole gale, . . 
 
 Storm, .... 
 
 Hurricane, . . 
 
 Just sufficient to make steerage 
 way. 
 "I With M'hich a ship with ) 1-2 kts. 
 |- all sail set would go in > 3-4 ,, 
 ) smooth water. ) 5-6 „ 
 
 Royals, etc. 
 Single reefs and 
 T.G. sails. 
 Dble. reefs, jib, &c. 
 Triple reefs, etc. 
 Close rfs. & courses 
 ( In which she could jiist bear 
 < close-reefed maintopsail and 
 ( reefed foresail. 
 Under storm staj'sails or trysails. 
 Bare poles. i 
 
 In which 
 siie coiild 
 just carry 
 
 But while, as has been said, the estimated force may be 
 exaggerated, the instrumental observation is liable to error, 
 both of excess and defect. In every gale gusts may, for a few 
 minutes, give a pressure of 80, though the storm is far short 
 of a hurricane; and, on the other hand, the instruments do 
 not fully record the suddenness which constitutes the chief 
 dangers of the revolving storm to sailing ships. 
 
 310. Storms. — Though it may not be easy to say at what 
 point a storm begins in temperate regions, where the accele- 
 ration of the wind is often gradual, the ti-opical storms are 
 abrupt enough to mark them sharj^ly ofi" fi-om the ordinary 
 states of the atmosphere. Their suddenness gives them a 
 distinct character apart from their rotation, but does not 
 afford a basis of classification. We may, perha2">s, regard 
 storms as belonging to two groups. 
 
 1. Those which are accelerations of the prevailing winds, 
 whether caused by increase of pressure from behind, or by 
 diminished i^ressure in front. 
 
 2. Those in which the prevailing direction is altered. 
 
 23 S 
 
274 PHYSICAL GEOGRAPHY. 
 
 311. Storms of Acceleration. — The hot v/inds, the sii'occo 
 and simoom, both southerly, both coming from heated sandy 
 deserts, are perhaps the best examples of winds driven by 
 pressure from behind. The gales of the N. Atlantic, of the 
 Patagonian coast, and of other localities where the prevailing 
 v/ind comes in contact with a current of different temperature 
 and different elastic tension, are due to the sudden diminu- 
 tion of pressure by condensation, which is propagated back- 
 wards in one or other current by the steady advance of that 
 which has after the contact the greater amount of tension. 
 
 312. Tornadoes. — The harmattan has abeady been de- 
 scribed (Art. 291); but it remains to add that the southward 
 shifting of the thermal equator has probably to do with the 
 occasional acceleration of this wind, by the amoimt of sudden 
 precipitation which takes place. This certainly seems to be 
 the case v/ith the tornadoes of West Africa, in which the 
 barometer and the dii-ection of the wind are unchanged 
 during the height of the storm. 
 
 313. Rotatory Storms. — But difficult as it is to trace the 
 origin of storms whose direction is rectilinear, perplexing as 
 is the effort to fix the relative importance of barometric 
 variations, it is still more difficult to accept, as conclusive, 
 any of the explanations hitherto given of the rotatory storms. 
 The meeting of antagonistic air currents, differences of elec- 
 tric tension, rotation of the earth on its axis, as well as 
 variations of atmospheric pressure, have been appealed to. 
 The facts seem to be: — 1. That the hurricanes of the West 
 Indies, the cyclones of the Indian Ocean, the Chinese 
 tyi^hoon, start from the areas in which currents mingle from 
 different directions. 2. The rotation is, in the West Indies, 
 N.W.S.E.,vdth the sun; in the South Indian Ocean N.E.S.W., 
 against the sun, but with the hands of a watch. 3. That 
 the centre of the revolving mass is an area of low pressure, 
 two inches lower than outside. 4. That the track of the 
 spiral follows the course of the prevailing wind. 5. That 
 the barometer falls before the storm. 6. That heavy rainfall, 
 and frequently electric displays, accompany the storm. 
 7. That when the storm approaches and touches the shore, 
 a storm-wave is hurled on the land with terrible effects. 
 The unsettled questions in meterology are : Is the barometric 
 
ROTATORY STORMS. 275 
 
 depression "before the arrival of the storm the cause, or is it 
 an effect like the Avave which precedes a swift steamer? is 
 the rainfall a cause here of movements which in other tropi- 
 cal localities it does not cause? 
 
 The months in which these storms occur are given below, 
 the West Indian figures being a mean of several years ; 
 those for the cyclones the statistics of one year ; while for 
 the typhoons the months of greatest frequency are merely 
 marked with a star. 
 
 Jan. Feb. Mar. Apr. Mav. Jan. Jly. Aug. Sep. Oct. J^ov. Dec. 
 West Inches, 1-5 2 3 2 1-5 37 28-5 24 20*5 5 2 
 S. Indian Ocean, 10 16 17 10 13 
 
 Typhoon, '" ^- '■'' 
 
 The track of the hurricane has already been described 
 (Art. 296). It practically starts along the oblique line formed 
 by the northward passage of the S.E. trade. Its parallelism 
 to the course of the Gulf Stream seems to indicate that both 
 are directed in their movements by the trend of the coast- 
 line as soon as they are clear of the Mexican Gulf. But 
 when a portion of the storm enters the gulf, it there also 
 follows the coast line and sweeps round the shores. 
 
 The origin, course, and period of the typhoons are singu- 
 larly repetitive of those of the hurricane. Their range in 
 latitude is from 10<^ to 24°. 
 
 The cyclones are on the northern limit of the S.E. trade, 
 as it recedes southwards, followed by the N.W. monsoon. 
 They start far to the east, near to Java, and travel along the 
 margin of the trade, bending towards the south-west near 
 Mauritius, the N.AV. monsoon there blowing farther south 
 than elsewhere, just as the S.E. trade in the Atlantic reaches 
 far towards Barbadoes. 
 
 In the North Indian Ocean they start near to the Nicobar 
 Islands, and reach to Calcutta or to Madras, for their course 
 varies in different years. They are less frequent to the west 
 of the Indian peninsula, along whose shores they commonly 
 travel. 
 
 These are the principal areas in which revolving storms 
 occur as regular periodic events. But in temperate regions 
 cyclonic storms sometimes are observed, and of these baro- 
 metric variations are probably the efficient causes. 
 
276 PHYSICAL GEOGRAPHY. 
 
 The Storm Chart of Europe for ^^ovember 2, 1863, given 
 by Buchan, page 242, illustrates the relation of the strongest 
 winds to the area of minimum barometric pressure, 28*9 
 (that of Euro2:)e being on the average 29*9). The track of 
 the two storms was from the west of Ireland to the head of 
 the Gulf of Finland on the one hand, through Denmark 
 towards Riga on the other. As the former of these is the more 
 common for the European storms, and as it is from S.W. 
 towards N.E., it is again an example of the prevailing winds 
 fixing the course, whatever may have been the origin of the 
 movement. 
 
 314. Velocity of Rotatory Storms. — Two distinct velocities 
 must be kej)t apart ; that of the wind, which may attain to 
 100 miles an hour, shifting round the compass, and that of 
 the storm which travels at from 10 to 15 miles an hour — 
 though the hurricane of 1866 is said to have approached 
 Bermuda at 30 miles an hour. The storms of Europe travel 
 at rates varying from 15 to 45 miles an hour, 18 miles being 
 the most common rate; and it is worthy of note that those 
 which occur in the area of westerly wind?, are more rapid 
 than those of the trade return currents. The determination 
 of velocity is of importance with reference to the transmission 
 of storm warnino's. 
 
 o 
 
 315. Area of a Rotatory Storm. — The diameter of the 
 revolution is variously stated. Many are known to have 
 had a diameter of 50 miles : 100 miles is proved to have been 
 the diameter of one. But the extreme breadth assigned, 
 1500 miles in some cases, requires strong proof. Buchan's 
 chart, already referred to, shows that the proximity of two 
 distinct storms might have led to their reference to one had 
 the intervening points not been observed ; and apart from 
 this, the observations on which such wide limits are asserted 
 would requii-e to have been made simultaneously. 
 
 316. Storm Waves. — The centre of a storm is an area of 
 low barometer ; but this is not due to centrifugal force, for 
 it is certain that the movement of the wind is in reality 
 vorticose. Precipitation in the centre of contiicting winds 
 gives rise to vortex movement, just as water escaping through 
 a hole entails secondary whirls in the air above it. This 
 sometimes takes place in the tornadoes, and in the gales 
 
DUST STORMS: SIMOOM. 277 
 
 formed, as at tlie mouth of tlie La Plata, "between the sea 
 and the river winds. The consequence of precipitation is 
 diminution of elasticity; thereafter the vorticose movement 
 increases in strength, and rises as air is drawn in along the 
 ground, for in all these cases the wind is not parallel to the 
 ground, but strikes down and is reflected from it. The 
 centre is thus a cup with the mouth downwards, into which 
 the water is raised; and when an obstacle occurs, when the 
 cup is broken, the water, no longer supported, " holds its way," 
 and is thrown down in mass over the land. 
 
 317. Whirlwinds: Waterspouts. — Every gusty day one 
 may see eddies round corners, which illustrate the j)rinciples 
 laid do-svn as to the movement of wind and water w^hen they 
 come against an obstacle. But the phenomena which have 
 now to be considered are of another kind. They are in tem- 
 perate regions usually connected with electric disturbance. 
 These local rotatory storms are sometimes called tornadoes; 
 and etymologically the name is better applied than to the 
 usually rectilinear gales of West Africa. The account of 
 that of Chatenay, near Paris, in 1839, quoted by Noad, de- 
 scribes what is sometimes seen on a small scale on the high 
 grounds of north Scotland; the lower part of a thunder- 
 cloud swelled downwards and became a conductor between 
 the upper clouds and the earth. The inverted cone became 
 surrounded by dust and light objects, drawn up to and wheel- 
 ing round it; the now continuous pillar travelled forward 
 for some time, iiprooting and twisting everything in its way, 
 and finally, the upper half was withdrawn into the clouds, 
 while the base of the pillar sank, a mass of rubbish, to the 
 earth. A similar description is given of the waterspouts 
 whose formation has been observed in the Pacific; and it is 
 interesting to note that the descent of the cloud-funnel to 
 meet the ascending cone of water was preceded by veering 
 winds, in the axis of which the spout was formed. Palgrave 
 describes deep circular hollows in the Arabian desert which 
 were probably formed by such Avhirlwinds, and in the Aus- 
 tralian deserts similar appearances have been detected. 
 
 318. Dust Storms : Simoom. — The dust storms of India 
 consist of a number of whirlwind columns moving together; 
 those of Nubia, described by Baker, move independently, as 
 
278 PHYSICAL GEOGRAPHY. 
 
 do tliose of tlie Saliara. The phenomenon may be imitated 
 by the ascent of a number of adjacent columns of smoke from 
 smouldering cotton, the spirals of each pillar and of the 
 whole mass being distinct if the air is stilL 
 
 The simooms of the Arabian desert, and the samiel farther 
 to the east, seem to belong to this group; they certainly 
 differ from the sirocco, inasmuch as their origin is within 
 the limits of the desert, and local heating of the surface 
 geems the only adequate explanation of theii* origin. 
 
 SECTION III.— ELECTPvICITY AND MAGNETISM. 
 
 Terrestrial Magnetism — Magnetic Equator — Line of no Variation-^ 
 Annual and Daily Variations — Intensity — Magnetic Stoi'ms — 
 Aurora Borealis : Cause of its Light — Atmospheric Electricity 
 —Diurnal and Annual Changes — Relation of Electricity to Heat : 
 Conductors — Conditions affecting the Amount of Electricity in 
 the Air — Thunderstorjns. 
 
 319. Terrestrial Magnetism. — The magnetic needle, when 
 suspended horizontally, does not point to the north pole of 
 the earth, but at Greenwich points between 20° and 21^^ 
 west of north. This, the north magnetic pole, has its counter- 
 part in an antarctic pole, which is correspondingly east of 
 the earth's pole. This variation of the needle is not constant. 
 The pole has shifted since a.d. 1576 from 11° 15' E. to due 
 north in 1657, thence to its westward maximum, 24° 27' 18'' 
 in 1815, from which it is now moving back. 
 
 But the change has not been in one plane. If a needle is 
 suspended so that it can swing verticoily, it will not rem^ain 
 horizontal, but its north pole will be deflected downwards at 
 an angle of about 68^ to the horizon at Greenvfich. The 
 nearer to the magnetic pole the greater Avill be the deflection, 
 so that at the pole it should be vertical. The deflection 
 diminishes towards the magnetic equator, and again increases 
 in the opposite dii^ection in the southern hemisphere. Now 
 this dip of the needle has diminished, at London, from 
 74° 42' in 1720, to 68° 2', and this indicates a greater distance 
 from the pole now than formerly. Taken in connection with 
 
INTENSITY OP MAGNETIC FORCE. ^7^ 
 
 the variation from east to west, tlie magnetic pole seems to 
 have described a circle round the terrestrial pole from east 
 to west, and Mr. E.. A. Proctor assigns about G50 years as 
 the period of the complete revolution. 
 
 The earth is thus a magnetic mass which afiects the needlo 
 as any other magnetic mass would ; and the student caii 
 repeat the main features of the phenomena with a magnetic 
 bar suspended in a paper globe, v.diile a needle is carried to 
 and fro outside. 
 
 320. Magnetic Equator, — Latitudinal bands have been 
 described round the globe, passing through the points of 
 equal dip. The equator, or line of no dip, is north of the 
 equator between the meridian of Greenmch and ISO*^ AV. 
 Ion., south of the equator for the rest of the globe's circum- 
 ference. 
 
 321. Line of no Variation. — This line, though not regarded 
 as of great importance, is of some interest. From the mag- 
 netic pole it passes to the svest of Hudson's Bay, thence 
 south-east outside of the Antilles, crosses the eastern promi- 
 nence of South America to the south magnetic pole ; thence 
 crossing the western portion of Australia, it reaches the 
 teiTestrial equator in about 75*^ E* Ion., and passes north- 
 ward to the magnetic pole. 
 
 322. Annual Variations. — Between the vernal equinox 
 and the summer solstice, that is, from April to July, the 
 westward variation diminishes ; for the rest of the year it 
 returns westward. 
 
 323. Daily Variations. — The needle moves alternately 
 eastward and westward towards the sun, whether he is above 
 or below the horizon : the maximum of the easterly move- 
 ment is reached at 7 a.m., the maximum westerly movement 
 at Ih. 10m. P.M., whence it recedes eastward till 10 p.m.; the 
 mean deviation for the day varying through 9' 8". These 
 are the mean movements, but they are modified seasonally. 
 In summer the extreme range is 13' 27''; in winter the 
 minimum daily range is 7' 2". In winter the westerly move- 
 ment is continuous throughout afternoon and night ; in 
 summer the eastward movement is continuous from 7 p.m. 
 to 7 A.M. 
 
 321 Intensity.— The intensity of the magnetic force 
 
280 PHYSICAL GEOGRAPHY. 
 
 varies, being at a maximum at the magnetic pole, and in 
 Siberia, for the northern hemisj^here; near the magnetic pole 
 for the southern hemisphere. The points of least intensity 
 are near the equator in the middle of the Pacific, and near 
 St. Helena in the South Atlantic. 
 
 325. Magnetic Storms. — The needles under observation 
 are disturbed from time to time, and as these disturbances 
 extend simultaneously over the globe, the influence to which 
 they are due is presumably a very widely acting one. The 
 relation of the magnetic storms to solar disturbances, indi- 
 cated by changes in the form and position of the solar spots, 
 has long been matter of observation; but their coincidence 
 with auroral displays is better known. 
 
 326. Aurora Borealis. — At either pole a dark arch rises, 
 Tvdiose plane is at right angles to that of the magnetic 
 meridian. Above the dark arch the auroral light is de- 
 veloped as a band of light from which long peaks are pro- 
 jected to the zenith ; or else a succession of concentric bands 
 of light, separated by dark arches, gives the appearance of 
 vertical curtains, seen in perspective, even the changing folds 
 of the drapery being to appearance recognisable. The horizon 
 of the auroras, or the line along which they are most fre- 
 quently seen, extends from New York through St Peters- 
 burg, cities whose parallels of latitude are 15^ apart, though 
 they are approximately in the same magnetic parallel. 
 
 327. Cause of Auroral Light. — Spectroscopic investiga- 
 tions, conducted during the Scandinavian Polar Expedition, 
 seem to indicate the presence in the atmosphere of iron and 
 carbon in a fine state of division, and of snow as ocmtribut- 
 in 2: to the character of the auroral liijlit. 
 
 328. Atmospheric Electricity. — Pure air offers little or 
 no resistance to the ^^^ssage of electricity through it: in 
 rarefied air the resistance is so diminished that it may be 
 looked on as a conductor rather than as an insulator. Hence 
 the intensity of electricity increases with height. The elec- 
 tricity of the atmosjDhere is positive, 10,000 observations, 
 extending over three years, showing 3*17 j^er cent, of negative 
 indications. In fair weather, Thomson found negative indi- 
 cations to precede a change from N.E. to a westerly wind, 
 and explains it by the accumulation in the air, over any 
 
RELATION OF ELECTRICITY TO HEAT. 281 
 
 locality towards whicli winds blow from different points, of 
 the earth's electricity, conducted off by trees, etc. The con- 
 dition of air and earth are then temporarily reversed, and 
 the same result might be expected to follow a whirlwind. 
 
 329. Diurnal and Annual Changes. — The intensity of 
 the atmospheric electricity varies periodically, increasing from 
 June to January and decreasing again to June. The daily 
 chanfijes are more marked in winter than in summer. The 
 following summary of his observations by Quetelet represents 
 also the conclusions drawn from the extended observations 
 at Kew: — 
 
 1. The electricity of the air, estimated always at the same height, 
 undergoes a diurnal variation, which generally presents two maxima 
 and two minima. 
 
 2. The maxima and minima vary according to the seasons of the 
 year. 
 
 3. The first maximum occurs, in summer, before 8 a.m., in winter 
 towards 10 a.m. ; the second maximum occurs, in summer, after 9 p.m., 
 in winter towards 6 p.m. The inters^al of time which separates the 
 two minima is therefore more than thirteen hours at the epoch of the 
 summer solstice, and eight hoiirs only at the winter solstice. 
 
 4. The minimum of the day is towards 3 a.m. in summer, 1 a.m. 
 in •\\dnter. 
 
 5. The mean electric state of the day is best represented about 
 11 a.m. 
 
 330. Relation of Electricity to Heat: Conductors. — 
 Electricity is commonly spoken of as a polar force, as if it 
 differed therein from other forces. But the conversion of 
 heat into electricity, or, to speak more correctly, the change 
 of that motion which appears as heat into that motion which 
 appears as electricity, though unknown to us save by its 
 results, is manifestly a change either in the transmissive 
 power of the particles in a particular line, or an alteration 
 in the line of transmission. The similar relation between 
 electricity and magnetism, makes an illustration available 
 from diamagiietism. Some bodies, held between the poles 
 of a horse-shoe magnet, swing axially, others swing with their 
 poles pointing towards those of the magnet. Among the 
 diamagnetic substances, or those which placed themselves in 
 the line connecting the poles of the magnet, is bismuth. 
 Tyndall prepared a rod of bismuth powder, made firm by gum 
 water, and found it possessed the property of the metal j but 
 
282 PHYSICAL GEOGRAPHY. 
 
 ■wlieii tlie rod was laterally compressed it became magnetic, 
 the molecular aggregation apparently controlling the dii^ection. 
 Between heat and electricity there is a remarkable coincidence 
 in properties, by which both are conducted equally by the 
 same bodies : a good conductor of the one is a good conductor 
 of the other. 
 
 331. Conditions Affecting the Amount of Electricity 
 in the Air. — The passage of a current is retarded by heat, 
 accelerated by cold: hence the seasonal differences already 
 stated, as well as those consequent on rarefaction. The amount 
 of electricity may be said to vary with the amount of mois- 
 ture in the air ; but the movement of moist air develops no 
 electricity by friction, unless the vapour has already assumed 
 the vesicular form; in other words, has undergone some 
 condensation. Probably the largest amount of electricity is 
 due to chemical action ; evaporation of perfectly pure v/ater 
 in still air develops none; but when compound solutions 
 are evaporated, when, therefore, chemical combinations are 
 formed or altered, electricity is generated. Hence combus- 
 tion, as a particular case of chemical combination, is a soui^ce 
 of electricity. Yf hen the evaporating compound is acid, and 
 when combustion takes place, positive electricity is given 
 off; when the solution is alkaline, negative electricity is 
 given. And it is worthy of comparison with the effects of 
 perfumes on radiation, that the relation is reversed by the 
 presence of vapour of turpentine in the discharge pipe through 
 which steam issues. 
 
 As pure air is as little retentive of electricity as of heat, 
 the accumulation of electricity in it can only be in proporticn 
 to that of substances capable of retaining it, chiefly watei*. 
 But mere moisture is not sufficient; the same quantity of 
 water at different degrees of- temperature occupies different 
 cubic space in the air, and the electric intensity is in jDropor- 
 tion to the density; hence the contrast in winter and summer, 
 in the heat of day and the cold of dewfall. 
 
 332. Thunderstorms. — The formation of every cloud is 
 therefore the accumulation of electricity; but if the formation 
 is slow, equilibrium is maintained by the escape of a portion. 
 But if large masses of vapour are suddenly accumulated, and 
 if these are in opposite electric states, a spark passes between 
 
THUNDERSTORMS. 283 
 
 them, and tliis is continued till equilibrium is established. 
 The reaccumulatlon may be very rapid, as in summer even- 
 ings, v.^hen the flashes continue for more than an hour from, 
 at least apparently, the same cloud. Forked lightning is due 
 to the breaking up of the flash by unequal conductivity of 
 different atmospheric layers ; this very frequently occurs 
 when the spark passes between cloud and earth, though it 
 also occurs v/hen the spark passes from cloud to cloud. 
 Sheet-lightning and silent lightning are probably the reflec- 
 tion on the clouds of thunderstorms at variable distances, 
 beyond or just within the limits at which the thunder or its 
 echo may be heard. Thunderbolts or fuloriirites. the track of 
 the flash through the soil, have been artificially produced 
 with a lar2;e friction machine discharginGj into salt and 
 sand, the vitreous tube thus formed exactly resembling those 
 found in the desert. Whirlwinds and . waterspouts have 
 already been mentioned, but they must be referred to hero 
 as the gentler restoration of electrical equilibrium between 
 the earth and the atmosphere, than that which takes pla,ce 
 during a thunderstorm. The report heard after the flash 
 is due to the sudden displacement of the air by expansion, 
 and the consequent inrush to fill the space; its propagation 
 is impeded by the resistance of the air, so that it does not 
 travel so far as might be expected from its initial intensity. 
 The number of seconds between the flash and the first sound 
 of the thunder multiplied by 1090, the average velocity of 
 sound through air, gives approximately the distance of the 
 thunderstorm, in feet, from the observer. 
 
CHAPTER YIL 
 
 CLIMATE AND WEATHER. 
 
 Climate and Weather — Relation of Temperacure to Latitude — 
 Equivalent Periods in both Hemispheres — Isothermal, Isotheral, 
 Isocheimal Lines — Continental and Insular Climates — Influence 
 of Currents ; East and West Shores — Climate of British Isles — 
 Climate of Lake Regions — Influence of Marsh Land — Form of 
 Ground — Decrease of Temperature in Altitude — Influence of 
 Barometric Pressure — Surface Temperature of Land and Sea — 
 Influence of Vegetation — Changes of Climate — Cycles of Climate 
 — Influence of Eccentricity of Earth's Orbit — Influence of Obli- 
 quity of Ecliptic — Coincidence of Extreme Eccentricity and 
 Obliqiiity — Geological Evidence of Climatic Cycles — Influence of 
 Geographical Changes — Weather — Deviations from Normal Tem- 
 perature — Lunar Influence — Weather Prognostics. 
 
 333. Climate and Weather. — Climate was at first intended 
 to express the annual temperature of a place, when tempera- 
 ture and latitude were believed to correspond. Additions 
 have been gradually made to this limited meaning, and now it 
 has even been used''' to include food as one of the external 
 conditions to which animals are subjected. This is, however, 
 too extended a sense to give to a term for which at present 
 no exact definition can be framed. Climate may be regarded 
 as the general tendency of a district towards mild or severe, 
 average or extreme, temperature, moisture, atmospheric pres- 
 sure. Weather is the variation from time to time in resi:)ect 
 of all or any of these conditions. A man's constitution is 
 a popular phrase for his tendencies towards any particular 
 kind of disease, and his capacity to endure changes : health 
 means his daily departures from, or return to, this condition 
 of equilibrium; for it is only in sanitary science that the 
 health of a district means disease. 
 
 334. Relation of Temperature to Latitude. — The alter- 
 * Juke's 3Ianual of Geology. Third Edition, p. 486. 
 
KELATIOy OF TEMPERATURE TO LATITUDE. 
 
 285 
 
 nate exposure of tlie nortlierii and soutliern liemispliere to 
 the sun (Art, 2) destroys the j)arallelism to the equator of 
 the zones of equal solar heat. As a consequence of the obli- 
 quity of its axis, the earth at B, which represents the northern 
 
 summer solstice, has the sun above the horizon of any place 
 in the northern hemisphere for more than twelve hours ; and 
 within tlie Arctic circle, that is, within 23° 27' 30", the sun 
 never sets. As the temperature of a region is in proportion 
 (omitting minor modifications) to its exposure to the sun's 
 rays, the northern hemisphere receives, in this phase, as much 
 more than the average as it receives less than the average in 
 the opposite phase D, which shows the southern summer 
 solstice. In the intervening jDositions A and C, the day 
 and night are equal all over the earth; and at these times, 
 spring and autumn respectively, the vernal and autumnal 
 equinoxes, the solar radiation on both liemispheres is equal. 
 Climate, therefore, presents for every locality a seasonal 
 maximum and minimum. If the hemispheres were identical 
 as regards the distribution of land and water, solar and 
 terrestrial radiation would balance each other ; but the 
 northern, the land hemisphere, is more rapidly heated and 
 cooled than the southern or water hemisphere. The summer 
 and winter temperatures of the former are, therefore, extreme ; 
 of the latter, nearer the mean. The following, which shows 
 also that the whole earth receives more heat in one half of 
 the year than in the other, is Dove's approximate estimate : — ■ 
 
 N. Hemisphere. S. Hemisphere. WlioleEartb 
 
 Temperature for July, 21 -0° C. (summer) 12° C. (winter) 16-7°C. 
 „ Jan., 9-3°C. (winter) 15°2C. (summer) I2'3°C. 
 
■2SQ IPHYSICAL GEOGRAPHY. 
 
 335. Equivalent Periods in Both Hemispheres. — The 
 hottest month in the N. temperate region is July ; in the sub- 
 tropical regions, May and July; near the equator, April and 
 August; at the equator, March and September. While, 
 therefore, the greatest warmth of the temperate regions, 
 nortli and south, is separated by six months from the greatest 
 cold, there are, nearer the equator, two warm seasons. As 
 these two hot months correspond to the passage of the sun 
 across the equator, it follows that the interval which separates 
 them decreases towards the poles, till at last the summer of 
 June indicates the time when the sun approaches to verti- 
 cality in the northern temperate zone. From this diagram 
 the student will recognise the seasonal correspondences, the 
 italicised names in the inner circle giving the cold seasons at 
 each quadrant. 
 
 Summer Solstice. 
 
 
 
 
 
 June. 
 
 
 
 
 
 
 
 December. 
 
 
 
 
 
 
 May. 
 
 
 July, 
 
 
 
 
 April, 
 
 
 
 August. 
 
 
 
 (Vernal 
 
 
 
 
 (Autumnal 
 
 
 March 
 
 Eq.uinox 
 
 June. . 
 •) 
 
 
 . Eq^uator. . . 
 
 .December 
 
 Efiuinox.) 
 
 September 
 
 
 
 February. 
 
 
 October. 
 
 
 
 
 
 Janua 
 
 June. 
 December. 
 
 November. 
 
 
 
 
 
 W 
 
 INTER Solstice. 
 
 
 336. Isothermal, Isotheral, Isocheimal Lines.^ — Maps 
 have been constructed on which the mean annual temperature 
 of the earth is shown by a series of lines passing through the 
 sj^ots which have the same temperatures. But these isother- 
 mal lines only give a very general idea of climate. Thus, 
 the mean of St. Louis is 12-2°C., of Algiers, 13-2°; but St. 
 Louis has a summer temperature of 23 "S", a winter of 0*5°; 
 v/hereas summer and winter in Algiers show 23-3° and 12*2°; 
 it is this range of 23° in the one case, of 11° in the other, 
 which characterises these two places. But St. Louis is 7'' 
 farther north tban Algiers; Nev/ Orleans is 2° farther south; 
 the tcmj^eratures at the latter place are: mean, 20*5*^0.; 
 summer, 27*7*^; winter, 13*3^. The mean does not tell of 
 
INFLUENCE OF CURRENTS. 287 
 
 tlie range tlirougli 14°, nor does the clifierence of latitude 
 account for the higher mean. 
 
 Fully to realise all the conditions on which the climate of 
 a region depends, would require the comparison of daily 
 thermometric readings at many points. But the broad facts 
 of the distribution of temperature may be gathered from 
 isotheral and isocheimal maps on which the lines of equal 
 temperature for summer and winter, respectively, are recorded, 
 or are still more obvious from the quarterly charts,'* 
 
 337. Continental and Insular Climates. — The influence 
 of sea is well seen by the comparison of localities iieai'ly on 
 the same latitude. Thus : — 
 
 Miiiimuiii. Maximum. Diff. jroan. 
 
 (EioJaneho, 20°C. 26-l°C. CrC. 23'C. 
 
 \ St. Helena, 14-4'' 17 7^ SS 16° 
 
 (Mauritius, 23-8^ 27-7" S'O 25° 
 
 i Honolulu, 22-7= 26 1° 4-1 244° 
 
 {Mexico, 12-2° 18'3° 61 15" 
 
 JBatavia, 25-5° 25-5° 25 5° 
 
 J Lima, 21° 255° 45 235° 
 
 ( St. Petersburg, .... -7-2° 16-1° 233 4-6° 
 
 I Reykjavik, -1'8° 11-6° 13 5° 
 
 i St. Louis, ■ 0-5° 23-8° 18 12° 
 
 j Kizeljelgah, E. Turkestan, -17° +3-8° • 20-8 -7° 
 
 Nain, ....... 88° -15° 238 -36°' 
 
 Fort York, Hudson's Bay, 15-5° -20o° 36 -3° 
 
 Sitka, 12-2° 12-2 61° 
 
 Edinburgh, 14-4° 3-3° 11 '1 8-8° 
 
 These examples will serve to illustrate the extreme range 
 of places, even on the shores of continents as compared with 
 islands, on the same latitudes. The insular as compared with 
 tlie continental climate is characterised by moderation, in 
 consequence of the different specific heat of land and water 
 (Art. 75). The isothermal lines show the greatest variations 
 of curvature over land surfaces, least over ocean areas. 
 
 338. Influence of Currents: East and West Shores.— 
 But the Gulf Stream oives a remarkable northward convexitv 
 of the curves, so that the summer temperatures at sea are 
 considerably to the north of their proper latitudinal position, 
 
 * Studenfs Adas of Physical Geography ^ Maps X. and XI. 
 
288 PHYSICAL GEOGRAPHY. 
 
 fclie winter isotlierm of 5°C. reaching to 58^ N". lat. in the 
 Atlantic, to 50° IST. hit. in the Pacific; on the other hand, it 
 descends below 40° N. lat. on the American coast. The 
 Arctic current thus indicates its course on the surface. 
 Similarly, the summer line of 25° C. in the southern hemi- 
 sphere is wavy, bending northwards with the prolongations 
 of the Antarctic drift. The efiect on the land climate is 
 not by radiation, but by the passage of heated air from sea 
 to land. Hence the western shores of the N. Atlantic are 
 colder than the eastern. The mean annual temj)eratures of 
 the following places show this : — 
 
 Halifax, - - 44° 39' N., G-2''C. Faroe Islands, 62° 2' N., 7-l°C. 
 Boston, - - 42° 21' K, 9-6°C. Dublin, - 53° 21' N., 9-6°C. 
 Falkland Islands, 52° S., 8-2°C. Port Famine, 53° 21' S., 5-3°C. 
 
 The third line shows the temperature of corresponding 
 southern latitudes. In the S. Atlantic the west shores are 
 the warmer, Rio Janeiro having a mean annual of 23° C, 
 Cape Town of 17-7°. 
 
 But the warm air is accompanied to the land by moisture, 
 and this, checking radiation (Art. 265), helps to maintain 
 the temperature. 
 
 339. Climate of British Isles. — The summer and winter 
 temperatures of the British Isles are instructive. Through- 
 out the year the source of its higher temperature, the Gulf 
 Stream, is indicated by the direction of the isothermal lines, 
 which in winter are on the whole parallel to the trend of the 
 coasts, in summer run from south of west to north of east. 
 In winter, temperatures above 4°C. are found to south and 
 west of a line from the Straits of Dover through the Isle of 
 "Wight, by Bristol, to the Irish Sea. The centre of Ireland 
 averages 3*8°C., but the line of 4° embraces a narrow margin 
 of the island on the east side, and a very broad tract of the 
 south and west. The west coast of Scotland has an average 
 of 3-8°C., the rest of Great Britain is 2*7°. In summer the 
 relations are reversed; the highest temperatures are still to 
 the south-w^est, but they now occupy the interior of Britain, 
 and the temperature of places to the east is higher than that 
 of places to the west on the same parallel of latitude. Hence, 
 while the difierence of summer and winter temperatures at 
 Galway is 9°, in the valley of the Thames it is 15°. On a 
 
INFLUENCE OF BAEOMETRIC PRESSURE. 289 
 
 minor scale, therefore, the east part of the area approaches 
 tlie continental character of climate. The difference between 
 east and west would be greater, but that a portion of the 
 cold water steals down as far as the west of Ireland. 
 
 340. Climate of Lake Regions. — In K America, the 
 freezing of the lakes seems to exercise the same influence as 
 if they were solid land. The isotherm of 0°C. (January), 
 curves to the south of the lake area, as far as it does over 
 the arid continent of Asia, whereas in summer it retreats 
 within the Arctic circle. The specific heat of water tells 
 therefore in summer; but in winter, ice gives the same 
 results as an equal mass of land. A similar deflection of the 
 0"^ line is due to the Baltic Sea. 
 
 341. Influence of Marsh Land. — The depth of water in 
 the lakes gives them a beneficial effect in summer, whereas 
 the evaporation of the thin layer of water over the swampy 
 grounds of Arctic America and Asia has the oj^posite effect, 
 of keei^ing down the summer temperature, as did the ill- 
 drained lands of Britain in former times. 
 
 342. Form of Ground. — A mountain range acts in two 
 ways; it forces the air into greater altitudes, refrigeration 
 being the consequence; and it condenses the moisture, thus 
 drying the wind. If the rarefied air descends on the other 
 side it reacquires a part, but only a part, of the heat it had 
 on striking the hill range, having lost a portion by radiation. 
 But it has also parted with moisture; it offers less impedi- 
 ment to radiation, and hence the diurnal range of temperature 
 is greater. Thus the Scandinavian chain separates two areas, 
 of which the western shows a difference of 18° between the 
 summer and winter temperatures, while on the east side the 
 difference is 23''. 
 
 343. Decrease of Temperature in Altitude. — It has been 
 already stated (Art. 269) that temperature diminishes with 
 height, and that the diminution is less rapid, and less regular, 
 away from mountains than in air which is in contact with 
 earth. Solar radiation is more powerful at great heights, 
 the air being drier, but terrestrial radiation makes up for this, 
 and the nocturnal loss of heat powerfully aids the influence 
 of rarefaction on the air in depressing the mean temperature. 
 
 344. Influence of Barometric Pressure. — It appears 
 23 T 
 
290 PHYSICAL GEOGRAPHY. 
 
 tliat tlie position of high barometric pressure strongly 
 affects the temperature of particular regions. Buchan 
 tabulates the barometric pressure over Europe''" for several 
 seasons of unusual warmth and cold, showing that when 
 the pressure in January 1867 gradually diminished from 
 30*262 inches in Iceland, to 29 '604 inches in Jersey, 
 the cold was 3*5° below the average in Orkney, and 1-6 in 
 Jersey, the mean temperature of Scotland being as much as 
 6-1° below the average of the month. Again, in November 
 1867, the pressure in Jersey was 30*278 inches, in Iceland 
 29*957 inches, a slight difference, yet the average tempera- 
 ture was above that of the month by 5*5° at Paris, 3° at 
 Orkney. Lastly, in December 1860, the barometric slope 
 was from 30*7 inches in Siberia, to 29*7 over Britain; and 
 the mean temperature of eastern Scotland was 15° C. below 
 the average on Christmas day. The movement of the air was 
 along the slopes indicated, and these examples abundantly 
 strengthen Buchan's appeal for the regulation of weather 
 telegrams, as a certain means of enabling the physician to 
 take precautions against dangers to health, as great as* the 
 dangers to shipping, to avert which storm warnings are 
 issued. 
 
 345. Surface Temperatures of Land and Sea. — The 
 depth to which daily variations of temperature are felt 
 extends, for the sea, to 100 fathoms in the Indian Ocean, 50 
 fathoms in the N. Atlantic, the limit of constant temperature 
 being 1700 and 1000 fathoms respectively (Art. 75). The 
 curve of the snow line has already been stated (Art. 207). 
 The limit of constant temperature of the soil is speedily 
 reached. The soil can heat only by conduction downwards, 
 and the impediments are such that the diurnal changes are 
 probably nowhere perceptible beyond four feet. The evapora- 
 tion from different soils has already been mentioned (Art. 
 174). The heat of the surface also depends on the character 
 of the soil, sand attaining the highest temperature, 70° in S. 
 Africa; but the Arabian and N". African mean is 33° to 35°0. 
 The more compact the material, the greater may its con- 
 ductivity be in general expected to be; hence the lower 
 temperature of solid rock than sand. Clays, on ihe other 
 * Handhooh, pp. 129 et seqq. 
 
CYCLES OF CLIMATE. 291 
 
 Iiancl^ are shielded by evaporation from the sun's direct 
 rays. 
 
 346. Influence of Vegetation. — Plants protect the soil 
 from being highly heated, radiation preventing the accumnla- 
 tion of heat. On the other hand, the moisture which accom- 
 panies vegetation helps to maintain a more equable, and 
 therefore a higher mean temperature. 
 
 The importance of forests has already been more than 
 once alluded to (Art. 60), and it is only necessary here to 
 point out that, as their evaporation increases the rainfall, 
 and as the condensation of moisture liberates a certain 
 qiiantity of heat which became latent on vaporization, the 
 direct as well as the indirect influence of vegetation on tem- 
 perature is considerable. 
 
 347. Changes of Climate. — The investigation of the 
 changes on the earth's surface has prepared the student to 
 understand the influences by which climate may be altered : 
 
 a. Elevation or depression of a coast line may alter the 
 du'ection of currents. 
 
 h. The greater or less height of a mountain chain may 
 stop or permit the passage of the v/inds from one basin into 
 another; even the lowering of a pass is of importance, since 
 migTatory birds, though high in air, are found to follow the 
 Alpine passes in their southward flight. 
 
 c. The deflection of a warm or cold current, as of the 
 Gulf Stream or the Labrador current, would be productive 
 of considerable change. 
 
 d. The removal or the increase of vegetation. 
 
 e. The draining of land. 
 
 All these are sources of slow change, and their occurrence 
 is irregular, as the movements on which chiefly they depend 
 are not subject to any law, so far as is yet known. 
 
 348. Cycles of Climate. — But the facts mentioned in the 
 first chapter regarding the earth's varying distance from the 
 sun, and the phenomena of precesssion and nutation, corre- 
 spond to variations in the amount of heat received from the 
 sun by different parts of the earth's surface. And as these 
 are periodically recurrent, though the intervals may be 
 affected by the attraction of other planetary bodies, astrono- 
 mical cycles correspond by their indirect influence to climatal 
 
292 PHYSICAL GEOGRAPHY. 
 
 cycles. Changes of physical geography, modifications of the 
 distribution of land and water, may increase or diminish the 
 temperature of particular regions, but they are subordinate 
 and uncertain influences compared with the astronomical. 
 
 349. Influence of Eccentricity of Earth's Orbit. — When 
 the eccentricity of the earth's orbit is at the maximum, the 
 earth in aphelion would be 8,641,876 miles more distant 
 from the sun than now. The two hemispheres would have 
 very diflerent temperatures if the winter solstice of one 
 ha2:>pened in aphelion. The northern hemisphere would be 
 reduced by 25*^ C, while the southern hemisphere, the winter 
 of which occurred in perihelion, would enjoy a more equable 
 climate. In consequence of the great reduction of temperature 
 in the north, the thermal equator would be much to the south 
 of its present position, the north-east trades, representing the 
 return of the westerly currents arrested far to the south of 
 their present limit, would have greatly increased force, and, 
 crossing the earth's equator, would drive the warmer tropical 
 waters to the south, so that the equatorial drift would not 
 enter the Caribbean Sea. Now, as the temperature of Scot- 
 land is 15-5* C. in excess of that proper to its latitude, the 
 withdrawal of the Gulf Stream would lower the temperature 
 by that amount; and as that current of warm water carries 
 to the north 3234 times the heat which would be conveyed 
 by a current of air of the same volume, the cessation of that 
 influence, and its transfer, even in a modified form, to the 
 Antarctic regions, would be of great impoi-tance. But the 
 northern summer occurring in perihelion would be, at first 
 sight, very warm. It must be remembered that the winter 
 cold would cause precipitation to take the form of snow, 
 and the summer heat would be largely spent in melting the 
 winter's accumulation; but melting leads to evaporation, 
 and the fogs thus resulting stop the heat rays from the sun, 
 while the snow and ice reflect the heat rays, and, at the same 
 time, cool the air by contact. A Avarm summer, therefore, 
 ministers to the snow and ice of the pole. The transfer, 
 under the influence of the strong N.E. trades, of warm water 
 to the south pole, would tend to diminish the ice there, and 
 ultimately to remove it. The loss of heat by radiation into 
 space, whieh warm air at the equator sustains by ascending 
 
COINCIDENCE OF ECCENTRICITY AND OBLIQUITY. 293 
 
 into higher altitudes, must also be taken into account, since 
 it diminishes the influence of a warm summer. The reversal 
 of the winters, the occurrence of the northern winter in 
 perihelion, would lead to a contrary condition of things; 
 and the Gulf Stream would then have as much more heating 
 power than it at present possesses in northern latitudes, as 
 it had less in the case just stated. It follows from the 
 relative power of warm water currents, as compared with 
 warm air currents, that an equatorial ocean would have 
 much greater power in moderating the severity of the polar 
 climate. 
 
 350. Influence of Obliquity of Ecliptic. — The maximum 
 obliquity of the ecliptic (Art. 1) would, according to Meech's 
 calculations, increase the amount of heat received at the poles 
 by Yg, that is, if the thermal days at the equator at present 
 are 365*24, and at the pole 151 '59, these numbers would be, 
 at the maximum obliquity, 363*51, and 160 '04 resjDectively, 
 a diminution of 1*73 in the one case, an increase of 8*45 in 
 the other; and this increase of ■—. would represent a rise in 
 the mean annual temperature of the poles to the extent of 
 from 7*5"^ to 8*5° C, if the polar region were free of ice and 
 snow; but the increase of temperature would, in reality, be 
 spent in melting part of that ice and snow, the air not rising 
 above 0°C. The conjunction of extreme eccentricity and 
 obliquity with Avinter in aphelion would be to moderate the 
 severity of the climate of that hemisphere whose winter 
 occurred in aphelion, and to diminish the ice at the opposite 
 pole. But the conjunction of maximum eccentricity and 
 minimum obliquity would tend to increase the cold of the 
 aphelial winter, and to diminish the warmth of that in 
 perihelion. 
 
 351. Coincidence of Extreme Eccentricity and Obliquity. 
 — The coincidence at remote periods, say 11,700 years ago, 
 is not determinable with certainty, nor is the rate of preces- 
 sion uniform vso far as is known. Mr. Croll gives 11,700, 
 33,300, and 61,300 years as periods when the winter solstice 
 of the northern hemisphere was in aphelion, the intervals 
 being 21,600 and 28,000 years respectively, and to this 
 extent, therefore, the statement in Art. 6 must be modified. 
 The further back calculations are carried, the less certain 
 
29^ ' PHYSICAL GEOGRAPHY. 
 
 become tlie periods at wliich phenomena are believed to have 
 occurred, the greater the chance of perturbations having 
 interfered with the reguhtrity of the movements. All that 
 it is proposed in these paragra2)hs to indicate is that, in 
 obedience to laws, the details of whose operations are not 
 known with certainty, the position of the earth relatively to 
 the sun has changed, and the temperature of the north and 
 south hemispheres has varied in correspondence with this 
 change, 
 
 362. Geological Evidence of Climatal Cycles. — Within 
 the Arctic circle the remains of plants and animals proper to 
 regions now greatly warmer have been found, and, beyond 
 the limits of existing species, carboniferous fossils prove 
 resemblance, even identity, of forms in polar and temperate 
 regions. The foregoing paragraphs suggest that these facts 
 are ex^Dlicable by reference to astronomical movements, whose 
 date we cannot, however, even approximately determine. 
 The presence of the elk, rhinoceros, and hippopotamus, in 
 Europe, is among the most recent palseontological evidence 
 of change of climatal conditions ; while the boulder clay and 
 striated rock surfaces bear testimony to a recent period of 
 great cold in regions where formerly permian, old red, and 
 Cambrian glaciers probably existed. The student will find 
 the full discussion of this most interesting problem in Mr. 
 Croll's papers in tlie Reader and Fliilosophical Magazine 
 since 1864, and in Sir Charles Ly ell's Frinciiiles, vol. i. 
 
 353. Influence of Geographical Changes. — The position 
 of the great masses of land at the poles, or at the equator, is 
 an important element in the change of climate. Equatorial 
 land would part more rapidly with heat into the air, and 
 radiation keeps the upper strata of the atmosphere cool, in 
 conjunction with rarefaction. It would also diminish the 
 area of warm water at the equator, and stop its movement 
 toward the poles. An equatorial ocean and polar land is a 
 hjrpothetical case, the conditions of which it is not easy to 
 determine; but, supposing that the preponderance of land 
 lay at the poles, and that in equatorial and temperate regions 
 there were still land masses (and there is no reason for 
 believing that land has ever ceased to exist in those regions), 
 the arrangements would be presented by which atmospheric 
 
Weather prognostics. 295 
 
 ^nd oceanic circulation would take place as at present, though 
 the details might not be identical. 
 
 354. Weather. — The student will have seen that much 
 of what has been said regarding temperature might be con- 
 sidered as affecting weather, not climate, and will therefore 
 be prepared to recognise climate as the mean of the weather, 
 thus making both terms the symbols of different quantities 
 of the same thing. 
 
 355. Deviations from Normal Temperature. — There is 
 seldom a regular gradation of temperature from the hottest 
 to the coldest months, or the converse. The departures from 
 a regular movement are either storms, or periods of heat and 
 cold in excess of that proper to the season. Some of these 
 deviations recur with great regiilarity, and are due to varia- 
 tions in barometric pressure, such as have been already men- 
 tioned, and which are probably determined by equatorial 
 disturbances of greater or less area. The deviations observed 
 in Scotland, and some of them are also European, are as 
 follows :* — 
 
 Cold, 7-10 Feb. , 11-14 Apr. , 9-14 May, 29tb Jn.-Uh .Ty. , C-11 Aug. , C-12 Nov. 
 Waru), 12-15 July, 12-15 Aug., 3-9 Dec. 
 
 356. Lunar Influence. — Popular tradition assigns great 
 power to the moon, and the lower temperatui-e^ after full 
 moon, has been ascribed to lunar heat dispersing clouds and 
 increasing terrestrial radiation. What the lunar heat may 
 be which is arrested in the atmosphere we cannot tell, but 
 Zengerf has shown, from a large number of observations, 
 that changes in the moon's distance are really followed by 
 differences of temperature. The preponderance of S. and W. 
 winds in the first, and of N. and E. winds in the last half of 
 the moon's revolution, is an isolated observation as yet; but 
 it is worthy of inquiry whether there are not in truth atmo- 
 spheric tides as there are of the ocean, movements not 
 identical in kind nor coincident in time with those of the 
 sea, but to which some peculiarities in the distribution of 
 barometric pressure may be traced. 
 
 357. Weather Prognostics. — These belong to the province 
 of practical meteorology, at present an empirical branch of 
 
 * Buchan. t Philosojphkal MafjazinCf 1868. 
 
296 PHYSICAL GEOGRAPHY. 
 
 science, and likely to remain so till long time or increased 
 points of observation shall yield the data on which more 
 certain principles shall be established. At present the care- 
 ful application of physical laws to data, comparatively scanty, 
 has led to conclusions, the general accuracy of which is 
 popularly discredited by the dishonest use which ignorance 
 and prejudice make of the failures. 
 
CHAPTEE VIII. 
 
 Hypogene, or Subterranean Changes — Intensity of Forces — Volcanoes 
 — Structure of the Cone — Ashes in Sedimentary Deposits — Trans- 
 port of Ashes : their Size and Composition — Texture and Com- 
 position of Lava — Quantity of Lava poured out — Alteration of 
 Cone and Crater : Course of Lava below Ground— Dormant and 
 Extinct Volcanoes — Distribution of Volcanoes — Latitudinal Vol- 
 canic Chains — Earthquakes — Eartlnvave Twofold — Form of 
 Earthwave — Modification of the Wave Shells — Earthquake 
 Wave at Sea — Tests of Direction of Movement — Change of Surface 
 — Phenomena accompanying Earthquakes — Area of Disturbance 
 — Distribution of Earthquakes — Causes of Volcanoes and Earth- 
 quakes — Hypothesis of connection between Sea and Volcanic 
 Centres — Vapour and Thermal Springs — Intermission of Geysers 
 — Periodicity of Earthquakes and Volcanoes — Secular Move- 
 ments of the Earth's Crust. 
 
 358. Hypogene, or Subterranean Changes. — Yolcanic 
 eruptions, earthquakes, and upward and downward move- 
 ments of the earth's crust, may occur apart or in conjunction, 
 or may succeed each other so as to suggest that their develop- 
 ment is alternative. Their association thus renders them 
 a natural group for systematic description, if it does not 
 necessarily indicate their common origin. But the operations 
 to which the metamorphism of rock masses is due, whether 
 they have 'been chemically or mechanically altered, or have 
 undergone both changes simultaneously or in succession, are 
 for the most part wdthout equivalent in degree at the surface 
 of the earth, though they may have representatives in kind. 
 In the present imperfection of our knowledge regarding the 
 chemistry of metamorphism, it is safest to keep that sulycct 
 apart from those previously mentioned, and to disregard to 
 some extent the probability of the common origin of all these 
 phenomena. 
 
 359. Intensity of Forces. — From the necessarily slow 
 changes in a rock undergoing alteration — through the oscilla- 
 
298 PHYSICAL GEOGRAPHY. 
 
 tions of level which are so gradual that even a minute change 
 is only detected by comparing the records of generations — to 
 the earth(^uake which ruins a region, and may have its horrors 
 intensified by simultaneous volcanic eruption, there is a scale 
 of intensity v/hich, if estimated by work done in equal times, 
 is in the order of enumeration; but, if tested by the total of 
 work done, the volcano is dwarfed by the side of the slow 
 elevation. 
 
 The volcanic outburst, it must be remembered, is only the 
 last in a long series of events : it represents the slight excess 
 of force in some direction v/hich overthrows, the balance and 
 sets in motion operations which tend to restore equilibrium. 
 Violent as the eruption may be, it is only a symptom; it 
 cannot be regarded as more than a very subordinate event, 
 and in speculating on the progressive diminution of volcanic 
 energy, it is not the outburst but the force of which it is the 
 expression, that must engage our attention. Now, the occur- 
 rence from time to time of violent events is a part of the 
 doctrine of uniformity, which means identity in kind, but 
 not necessarily in degree, of the processes which have gone 
 on at all times of the earth's history. The degree may have 
 varied, so that the intensity is greater now, or was greater 
 in the ]3ast than now. The amoimt of energy in the earth 
 is undergoing diminution, but at what rate we cannot tell : 
 if we assume that volcanic activity depends on a store of 
 materials, or of force which has not been renewed, the dimi- 
 nution in amount of volcanic activity is a necessary conse- 
 quence of the tendency to equilibrium manifested by all 
 chemical change. But geology gives no reason for believing 
 that there has been less activity in recent times, and furnishes 
 evidence that defective observation has exaggerated the 
 intensity of the past, by massing together events which were 
 really far apart. On the other hand the conflict of opinion 
 among competent chemists and physicists, as to the conditions 
 of volcanic activity, proves that speculations are far from 
 resting on a sufficiently wide induction of facts, and that even 
 the chemical elements of the jDroblem are undetermined. 
 
 360. Volcanoes. — In such long mountain chains as those 
 of America and Central Asia, numerous volcanic peaks occur 
 which were not the cause of the mount?vin elevation, and 
 
STRUCTURE OP THE CONE. 299 
 
 may have been developed at any period before or since the 
 elevation. The most obvious phenomena connected Avith 
 volcanic outbursts are the events which take place at and 
 near the seat of eruption. The majority of the orifices by 
 which material escapes from the interior of the earth's crust 
 open on the summit of elevations of greater or less height. 
 Usually the crater of the volcano, as this aperture is de- 
 nominated, opens on the top of a conical hill, more or less 
 abruptly truncated, and for the most j)art having one side 
 higher than the other, the smooth outlines of the cone con- 
 trasting with those generally presented by ordinary denuded 
 hills. The crater gives exit to lava, ashes, steam, sulphurous 
 vapours, nitrogen, hydrogen, and hydrochloric acid. Several 
 of these are present in all eruptions, but in very various 
 quantities. The lava may be the principal material; in others 
 no lava flows out, only ashes are driven forth, or hot water, 
 sulphurous vapours, or other gaseous emanations may escape 
 alone. Lava pours out of the orifices of the crater, or, as 
 very frequently hapj)ens, through apertures upon the side of 
 the cone ; it flows out in a stream or coulee, which may not 
 descend below the cone, or may travel do■s^^l to the plain, and 
 even — if the quantity of lava is large — pass for miles over 
 the adjacent country. The slope, at first high, as much as 
 30*^, gradually becomes less as the stream approaches the low 
 grounds, until it finally terminates, usually with a more or 
 less vertical face. This description aj^plies to the typical 
 volcano, consisting of a cone, through which passes a single 
 supply pipe. But such a simple case is of comparatively 
 rare occurrence, the volcano, for the most part, presenting a 
 complicated structure, due to the presence of several orifices 
 more or less distinct from each other. 
 
 361. Structure of the Cone. — The cone, as has been 
 said, terminates the truncated extremity, and its orifice 
 leads into a funnel-shaped cavity, the materials on the sides 
 of which slope downwards towards the orifice of the supply 
 pipe. The gi-eater height of one side is due to the manner in 
 which a cone is formed; for the volcano is not a mountain 
 of elevation, it is in reality a mass which grows at the sum- 
 mit. Volcanic ashes, that is to say, the molten matter 
 which is blown into a coarser or finer powder by the force of 
 
300 PHYSICAL GEOGRAPHY. 
 
 the explosions, are thrown npAvards, and descending for the 
 most part round the orifices, gradually pile up a hill which 
 has the same typical form as the heap of sand in the hour- 
 glass. If the prevalent winds have any strength, the high 
 side of the volcano will be to the leeward, and by repeated 
 eruptions the height of the mass may go on steadily increas- 
 ing. Its incoherent materials are liable to be suddenly re- 
 moved, as has happened in Java, where the cone has suffered 
 a diminution of 4000 feet after the close of a series of erup- 
 tions. If a long period of quiescence follows, the cone 
 becomes affected by atmospheric waste ', its height is slowly 
 reduced, and ravines are furrowed out of its sides, so that, as 
 Junghuhn said of the Javan cones, they look like umbrellas, 
 the ridges representing ribs. 
 
 362. Ashes in Sedimentary Deposits. — The ashes, how- 
 ever, do not always fall in the immediate vicinity of the 
 crater; they may be transported to a considerable distance. 
 If they fall into the sea, they become incorporated with the 
 sedimentary deposits there going on, which thus present 
 transitions from the purely volcanic to the purely sedimen- 
 tary formations (diagram, p. 25). Again, steam is the almost 
 constant accompaniment of eruptions, and, becoming con- 
 densed immediately after its ejection, is precipitated upon the 
 surface of the volcano, carrying with it the finer ashes, and, 
 flowing down towards the lower grounds in a stream of mud, 
 is perhaps as destructive as a stream of molten lava. These 
 floods, known as moya in South America, form deposits which, 
 if preserved by subsequent lava flows, ofier a close resem- 
 blance to subaqueous accumulations. 
 
 363. Transport of Ashes. — But the force of the eruption 
 may carry the ashes into the air for a considerable distance, 
 so that they may actually pass into the upper stratum, and 
 be carried by the steady westerly winds. Thus, in 1815, the 
 ashes of Sumbawa were carried to Amboyna, a distance of 
 800 miles to the north-east; the ashes of Coseguina were 
 carried to Kingston in Jamaica, a distance of 700 miles, in 
 four days; while the ashes of Hecla reached the Shetland 
 Islands, transported, however, in this case, by a lower cun-ent, 
 to the S.E. The quantity of this kind of material is various. 
 What it may be, however, is suggested by the fact that eight 
 
COMPOSITION OF LAVAS. 301 
 
 leagues to tlie south of Coseguina tlie aslies formed a layer of 
 three feet in thickness; and ashes formed the chief part of 
 the material which buried Pompeii and Herculaneum. 
 
 364. Size and Composition of Ashes. — The size of the 
 ejected material varies very much : blocks as large as an ox 
 have been thrown out, and, falling upon sedimentary strata 
 in course of formation, have sunk into them. The soft layers 
 are carried downwards by the weight; new layers, as they are 
 laid down, arch over the block, and thus a record is preserved 
 of the periods at which eruptions may have occurred. The 
 quality of the aslies varies with that of the lava in the same 
 eruption ; being, therefore, siliceous or basic, as will be imme- 
 diately explained 
 
 365. Texture of Lava. — The lava poured out varies in 
 character, being in some cases more tenacious than in others: 
 thus. Von Buch describes the lava of 1805 as shooting down 
 the cone of Vesuvius, the velocity being probably several 
 hundred feet in a few seconds; but, for the most part, it is 
 somewhat more viscid. Whether it overflows the lip of the 
 crater, or passes out by lateral orifices on the cone, it parts 
 with its heat rapidly from the suil'ace, and thus becomes 
 coated with a dense layer, which graduallv retards its speed. 
 The vertical section of a lava flow shows that the central 
 portion of the mass is more compact, while the upper and 
 lower surfaces form a layer of scorise of greater or less thick- 
 ness, the included gases expanding as they approach the sur- 
 face, and escaping with more or less violence at the upper 
 surface, so as to give the coulee that ragged aspect Avliich is 
 preserved in some of the Auvergne outflows as freshly as if 
 they had been of yesterday. Rock being a bad conductor, 
 the formation of this hardened outer layer diminishes the 
 speed of radiation, and thus the heat of the central mass may 
 be retained for a considerable time; thus the lava of Jorullo 
 retained sufiicient heat after eight years to light a cigar a few 
 inches below the surface. It is this incrusting of the mass 
 with a solid covering w^iich gives to the termination of the 
 coulee its usually abrupt form. 
 
 366. Composition of Lavas. — Lavas are trachytic, or 
 doleritic; contain, that is to say, a larger amount of silica 
 on the one hand, and of tho hcayier basic ingi'edients on 
 
302 PHYSICAL GEOGRAPHY. 
 
 tlie other. Tlie mean composition of the two types is given 
 below : — 
 
 
 
 Trachj'tes. 
 
 Dolerites 
 
 Silica, ■^ . , 
 
 
 66-5 
 
 51-0 
 
 Alumina, 
 
 
 17 
 
 140 
 
 Potash, 
 
 
 50 
 
 0-2 
 
 Soda, 
 
 
 4-0 
 
 3-4 
 
 Lime, 
 
 
 1-4 
 
 100 
 
 Magnesia, 
 
 
 ri 
 
 5-5 
 
 Oxides of Iron and Manganese, 
 
 8 
 
 14-7 
 
 Loss by ignition, 
 
 . ... 
 
 10 
 
 11 
 
 But though these varieties frequently occur separately, they 
 are also met v/ith in combination as a product of the same 
 volcano, being either discharged by distinct orifices, or at the 
 same orifices at different periods, giving rise to alternate layers 
 of the two species. As their specific gravity is different, that 
 of trachyte being, on the average, 2-6, while the dolerites 
 are nearly 3, it has been suggested that they represent 
 different layers of molten material in the interior. But the 
 fact of their indiscriminate occurrence makes it more probable 
 that the difference is due rather to the mode of segregation 
 from a common mass than to any original differences in the 
 source of supply. Similarly, in one and the same coulee 
 varieties in the proportions of the ingredients may be found, 
 due to the greater or less distinctness of the crystallization, 
 or, in other words, to the greater or less pressure to which 
 the mass has been subjected; and the differences must bo 
 borne in mind, as they help to explain the unequal modifica- 
 tion which lava flows have undergone. The scoriae resemble 
 the slag of a glass furnace, and they may present the appear- 
 ance kno\vn as ropy, if, by the onward movement of the lava, 
 they become twisted and carried forward in curved lines, 
 which recall the curves across the surface of a river or of a 
 glacier. By the onward movement the gas bubbles, striving 
 to reach the surface, may be protracted so as to present, not 
 spherical, but lenticular, cavities, or even elongated lines, a 
 character of use in the more ancient lavas or trap rocks, as 
 enabling us to recognise the direction in which the stream 
 has moved. 
 
 367. Quantity of Lava at one Eruption. — The total 
 quantity of lava emitted at one eruption is very various ; 
 
COURSE OF LAVA BELOW GROUND. 303 
 
 the coulee may never pass beyond the cone, or it may, like 
 that of SkajDtar Jokul, Iceland, in 1783, extend over forty- 
 five miles, with a varying breadth of seven to fifteen miles, 
 and a depth of one to six hundred feet, a mass which, as 
 Lyell calculates, would stretch from Hampstead to Gloucester. 
 The quantities poured out in the older formation often cover 
 an enormous area, and to a great thickness. But it is im- 
 possible to recognise the portions belonging to indivdual 
 outflows, and the position of the crater is equally beyond our 
 knowledge. 
 
 368. Alteration of Cone and Crater. — It has been' men- 
 tioned that a crater is an incoherent mass consisting w^holly 
 of ashes, or of lava and ashes interstratified. Such a struc- 
 ture is not likely to be permanent, and it is only very rarely 
 that we find any trace of the actual crater of volcanoes which 
 have, since their formation, been exposed to marine denuda- 
 tion. Etna ofiers a grand example of this : the Val del Bove 
 is excavated by atmospheric waste to a depth of 3000 feet, 
 entirely out of the softer materials of many outpourings. 
 The summit of Arthur Seat, at Edinburgh, is a fragment of 
 a tertiary cone, part of the ashes being still preserved in 
 place as they lay round the orifice, which is now filled by a 
 plug of basalt. 
 
 369. Course of Lava below Ground. — The ancient trap 
 rocks give us information as to what may be called the ana- 
 tomy of the volcano. The diagram (Art. 27) sums up that 
 anatomy, and shows that between the contemporaneous ?-nd 
 the so-called intrusive masses there is no distinction save 
 that of position, a,nd of such textural characters as result 
 from greater or less pressure, or speed of cooling. Every 
 volcano breaks through sedimentary strata; and whether 
 the direction of outburst is determined by the eftbrt of the 
 lava to reach the surface, or the lava takes advantage of 
 fissures already created (and both things may have concuried), 
 it is certain that in its upward progress, more especially if 
 movement be retarded at the surface by any cause, the 
 force from behind will compel the molten matter to escape 
 in any direction, wherever a line of Aveakness exists. It 
 would, therefore, tend first to pass between the planes of 
 stratification. The direction of the lava would be altered, 
 
304 PHYSICAL GEOGRAPHY. 
 
 further, by joints or fissures across tlie strata, and, if it actu- 
 ally reached the sui'face, its place of escape would be regarded 
 as another lateral orifice. It is this mechanical tendency to 
 take advantage of weak points which originates the many 
 dykes that traverse the cones of most volcanoes. Denudation 
 makes the position of dykes obvious by wearing away the 
 sedimentary strata; but the trap itself is sometimes wasted 
 so as to leave a parallel-sided gap: the projection in the 
 former case, the gap in the latter, correspond to the Scottish 
 and Cumbrian sense of the word dyke, respectively. 
 
 370. Dormant and Extinct Volcanoes. — We do not know 
 if any volcano, now quiescent, is extinct in the sense of final 
 cessation of possible activity. Probably there is no such 
 extinction any more than there is reason to believe that a 
 volcano cannot break out at a spot where such an event 
 never happened before. A volcano may be quiescent for 
 centuries, and its dormant state may end in a very violent and 
 long-continned period of activity. We cannot even regard 
 as proof of extinction an interval of, it may be, many thousand 
 years, since, for aught we know, the j)i'Ocesses which termi- 
 nate in an outburst may be going on beneath, but do not 
 make themselves manifest, either because the operations are 
 slow, or because their tension is relieved in other directions. 
 Thus the interval between the carboniferous and the tertiary 
 eruptions in Scotland was probably greater than that between 
 the tertiary and the present time. 
 
 371. Distribution of Volcanoes. — The distribution of 
 volcanoes over the surface of the earth at the present time 
 presents a certain kind of system, and lines may be traced 
 as mountain chains have been traced, though in the one and 
 in the other case we cannot regard these lines as evidences 
 of simultaneous action at all points. We ought properly to 
 look upon them as a series of successive actions, which, occur- 
 ring at different periods, have maintained the same general 
 direction. The American continent furnishes the most con- 
 tinuous line. The continuity, taken in conjunction with the 
 alternating activity of different points, suggests that they are 
 situated over a single longitudinal fissure. Commencing at 
 the south we have those of Fuego, the highest of which is 
 7000 feet; and the Patagonian volcanoes; situated about 
 
DISTRIBUTION OF VOLCANOES. 305 
 
 5i° south latitude. In the Andes the chain extends from 
 43*^ to 30° lat. S., and numbers more than thirty peaks, 
 Aconcagua being the highest; those of Bolivia, seven or 
 eight in number, from 21^ to 15*^ ; the Quito district, from 
 2^ lat. S. to 3^ lat. IST., includes about twenty peaks, nearly 
 all above 14,000 feet of elevation. About 9^ lat. N". the 
 volcanoes of Central America, Mexico, and Western America 
 commence ; nearly sixty are known, and several are recorded 
 whose history is, however, im^^erfectly known. The series 
 terminates in Mount St. Elias, whose height is about 18,000 
 feet. The great Mexican volcano, JoruUo, is 123 miles from 
 the nearest ocean, but the rest present the usual relation of 
 most active volcanoes, namely, close proximity to the coast 
 line. The West Indian Islands are connected with this 
 great north and south line by a line passing from Quito 
 through Granada. The West Indian Islands form two 
 parallel chains; in the western there are ten volcanoes, 
 while the eastern consists chiefly of calcareous rock. The 
 American line is connected with the Asiatic by a chain 
 stretching through the Aleutian Islands, of which more than 
 twenty are volcanic, to Kamtschatka, which contains about 
 twenty volcanoes. Southwards, the Kurile Islands, twelve 
 of which are volcanoes, and the Japanese group, with twent}'- 
 five, form a fringe on the west shores of Asia. In the 
 Philippine Islands, with about twenty volcanoes, the chain 
 is traversed by a very numerous series, passing from the 
 Indian Ocean eastwards to New Guinea, through Sumatra 
 and Java, each of which contains about fifty active or dor- 
 mant volcanoes. The line curves away to the south-east of 
 New Zealand, and nearly comi)letes a ring round the Pacific. 
 Though the general direction of the line thus traced is tole- 
 rably continuous, it is probable that detailed enquiry will 
 prove the existence of several distinct axes belonging to diffe- 
 rent periods of activity ; but it is interesting to note that 
 the two opposite shores of the Pacific manifest exactly the 
 same kind of distribution of the volcanic chain. Another 
 line has been traced from the borders of China westwards 
 into Asia Minor, and thence by the Mediterranean to the 
 Cape Yerde Islands ; but the interruptions of this line pre- 
 vent us from regarding it as a common axis. It is certain 
 23 u 
 
306 PHYSICAL GEOGRAPHY. 
 
 tliat Spain and the north-west of Africa form a single vol- 
 canic region, while the north-eastern corner of Africa, with 
 Arabia, is singularly free from, all record even of former 
 activity. Isolated volcp.noes occur which cannot be asso- 
 ciated with any of the recognised lines ; thus, in the North 
 Atlantic, Jan Mayen and Iceland are separate centres, though 
 they may possibly be connected. Passing southwards, the 
 Azores, the Cape Yerde Islands, the submarine eruptions at 
 the equator. Ascension Island and Tristan d'Acunha, form for 
 the most part perfectly isolated volcanoes, though two of 
 them, the Azores and the Cape Yerde group, have been held 
 to belong to the Mediterranean axis. St Paul's Island and 
 the Mauritius group are likewise isolated, and other volcanoes 
 have been recorded still farther to the south, whose relations, 
 however, are absolutely unknown. The diagram on p. 66 is 
 intended to show the deviation of the axial lines among the 
 islands of the Pacific. The remarkable way in which they 
 contrast with the lines of existing continents is of great in- 
 terest in connection with the hypothesis that the Pacific and 
 part of the Indian Ocean were, since the appearance of man 
 on the earth, the seat of a continent, that it had been so for 
 long ages previously, and that Australia, JSTew. Zealand, Tas- 
 mania, are the last fragments of this changing land surface 
 (Art. 38). 
 
 372. Latitudinal Volcanic Chains. — The volcanoes 
 hitherto sj)oken of are either on the margins of continents, 
 in chains of islands, or isola^ted in ocean. But the mountain 
 lines across Asia are to some extent associated with volcanic 
 phenomena, mud and vapour vents occurring at intervals 
 across the continent. But these are most numerous in the 
 western area, near the Ponto-Caspian area, and no active 
 volcano exists in the interior, far a,v/ay from a water area. 
 The subjoined diagram shows the relations of the leading 
 mountain chains to the lines of volcanic activity, enmnerated 
 in last paragraph. 
 
 373. Earthquakes. — An earthquake is a vibration of the 
 earth's crust, a disturbance whose effects are immediately 
 appreciable, which is therefore connected vv^ith violence. The 
 movement of the surface of Yesuvius before an eruption is 
 only detected by Palmieri's delicate instruments, but it fore- 
 
MOUNTAIN AXES, 
 
 507 
 
 MouirrAiN Axes of Europeo-Asiatic Continent. 
 
 coo 
 
 c p a 
 
 <<?' 
 
 .^J-'^ 
 
 5>^ 
 
 
 vv' 
 
 o>*- 
 
 .s^^^ 
 
 W. 
 
 Tj-rcnees. CaiitaLrian Mts. Sien'a Nevada. 
 
 Central A]^.^. Balkan. Germanic Mts. 
 
 %, 
 
 % 
 
 E, 
 
 ^/ 
 
 \ 
 
 v> 
 
 
 ■^4: 
 
 % 
 
 "% 
 
 % 
 
 
 
 Caucasus. Aimeniau Mts, Taurus. 
 
 Antjtaurug, 
 
 Ilindo Kush. Altai. Kuon Lun. 
 
 Thian Shan. 
 
 Aldau Mts. Neilgherries. 
 
 
 •'■/■a < 
 
 
 '\ 
 
 <5> 
 
 V V" 
 
 5s^' 
 
 
 
 Mountain Axes of America. 
 
 
 / 
 
 Guiani^. 
 
 Trinidatl 
 
308 PHYSICAIi CEOClRArilY. 
 
 tells and passes into tLat violent action amid wliich even 
 tliunder is unheard. -■r- 
 
 " 374. Earthquake Wave Two-fold. — A shock of some 
 kind is communicated to a mass of the earth's crust; it gives 
 rise to two distinct movements: 1. The shock displaces the 
 particles relatively to each other, at the point of impact, and 
 this displacement tends to travel in a straight line. 2. From 
 the point of shock a wave of elastic compression travels for- 
 wards at right angles to the line of shock. The latter travels 
 more rapidly than the former. In the Calabrian earthquake, 
 Mr. Mallet found that the rates were 789 : 13 in feet per 
 second. Keverting to the definition of a wave given in Art. 
 38, the student will see that the transmission of the wave- 
 form,, each particle returning to its place, must be more 
 rapid than the propagation of motion in which each particle 
 displaces that which precedes it. Standing on the shore on 
 a calm day when steamers are passing at various distances, it 
 "will be seen that each steamer causes two sets of waves to 
 break, the one later than the other, by an interval which is 
 directly as the distance of the steamer. The first waves are 
 those of elastic compression, which are converted into move- 
 ments of translation against the shore; the second, more 
 powerful, are the waves of displacement. 
 
 375. Form of Earth Wave. — An upward blow on the 
 earth's crust tends to travel vertically uj^wards, and this, the 
 seismic vertical, produces at the surface a vertical displace- 
 ment upwards and downwards; but from the point of shock 
 other waves reach the surface at angles which gradually 
 increase with distance, the tendency to horizontality of the 
 consequent movement likewise increasing. Thus, if C is 
 at the surface of the ground, the line A B C is the seismic 
 
 4 3 210 1234 
 B 
 A 
 
 vertical. If B is the point of impulse, lines Bl, B2, B3, 
 represent the increasing angles of emergence; the intensity 
 of the shock is inverse to the angle, and the tendency to 
 horizontality increases till the line of shock is at right angles 
 to the vertical. If the disturbed medium were homogeneous, 
 the points 11, 2 2, 3 3, which are the coseismic points^ 
 
EARTHQUAKE WAVE AT SEA. SOD 
 
 would be in the circumference of circles; and tlie angle of 
 emergence would be equal all round the seismic verti- 
 cal as a centre. The waves would thus form conical shells. 
 The deeper the point of shock, as at A, the acuter the 
 angle of emergence, Al, A2, etc., to start with, and the 
 wider the area over which the concentric shells would extend. 
 It is obvious that, though the sensations of an observer 
 suggest that the ground undulates from the point C out- 
 wards, the undulation is in reality made up of a series of 
 movements at many consecutive points of the surface, and 
 each of these starts independently from A or B. If, there- 
 fore, the direction of movement at 1, 2, 3, could be ascer- 
 tained, the depth of the point of shock might be calculated. 
 This is the substance of Mr. Mallet's teaching, and the perusal 
 of his reports to the British Association will repay the 
 student who desires to follow the application of mechanical 
 principles to natural phenomena. 
 
 376. Modifications of the Wave Shells. — But the medium 
 is not homogenous : it consists of layers of different texture, 
 density, and thickness; and as the earth wave, like other 
 waves, is due to reflection and refraction, the figure which 
 the coseismal points would describe on the surface must 
 vary. In the foregoing diagram, the lines lAl, 2A2, repre- 
 sent inverted cones, with a circular base on the surface of 
 the ground, but by such inequalities of movement as have 
 been suggested, the circle may become an ellipse, or some 
 still more irregular figure. A fault line in sedimentary 
 strata has its dii'ection altered in passing through layers of 
 difierent density, as the ray of light is deflected in passing 
 from air to water; and the earth wave undergoes the same 
 change of direction, so that the area at the surface, bounded 
 by any set of coseismal points, may be smaller or larger 
 than it should be were the angle of emergence, proper to 
 the distance of these points from the vertical, at its normal 
 value. 
 
 377. Earthquake Wave at Sea. — Though these move- 
 ments at sea are not registrable like those on land, it is 
 obvious that, after the earth wave is commmiicated to the 
 water, its behaviour mil be more regular in the homogeneous 
 fluid, and the movements already produced will travel with 
 
310 PHYSICAL GEOGRAPHY. 
 
 less unequal speed. The wave of translation thrown on the 
 land, always a powerful agent of destruction, depends for 
 its power on one or two conditions. Thus the horizontal 
 tendency increases from the seismic vertical to a certain 
 point, whose distance depends on the depth below ground of 
 the shock, diminishing thereafter. If it strikes on a shelv- 
 ing shore, it gathers strength as it advances, in the same 
 way as the tidal wave under similar circumstances. 
 
 378. Tests of the Direction of Movement. — As the plane 
 of the wave tends to coincide with that of the earth's surface, 
 its effect on buildings varies. The less the angle of emergence, 
 the greater is the intensity, and the more of the vertical 
 height of a pillar or building which shares in the horizontal 
 movement. Hence, near the vertical, a pillar of sik feet high 
 will be thrown forwards j at a greater distance the base will 
 be shifted forwards, but the inertia of the upper part will 
 make it fall behind, as a man falls when a carriage suddenly 
 moves. A wall running in tlie dii-ection of the wave is 
 fissured at right angles to its plane, so that if the wall and 
 the wave are, say, in a plane from E. to W., the fissure will 
 be oblique from above downwards, from W. to E. 
 
 The twisting of spires seems due to the reflection of the 
 wave, and to be the joint product of the first and second 
 movements already mentioned. 
 
 Fissures of the ground take place in very irregular fashion, 
 either at right angles to the line of shock, or radial, as if 
 vorticose movements had occurred. 
 
 379. Changes of Surface. — These fissures may be tempo- 
 raiy or permanent; in the latter case they sometimes become 
 the seat of mud or other springs. The permanent elevation 
 of coast lines, as Chili and north Australia, and the conver- 
 sion of valleys into closed basins, as has probably occurred in 
 the Andes, are interestincj as connecting these sudden violent 
 disturbances with the more gentle movements of elevation 
 and subsidence. Equally interesting is the formation of 
 lakes, as in the Sunken Country of the Missouri in 1812. 
 
 380. Phenomena Accompanying Earthquakes. — The 
 phenomena accompanying earthquakes are, as Mr. Mallet 
 sums them up — 1, The gi-eat earthquake wave; 2, the wave 
 which is formed by the vertical displacement of the ocean 
 
CAUSES OF VOLCANOES AND EARTHQUAKES. 311 
 
 floor, and the consequent overflow, in all dii'ection«, of the 
 water thus elevated; 3, the wave of sound through the 
 earth, which may or may not precede that of the shock; 4^ 
 the wave of sound transmitted through the air or sea; 5^ 
 the grea.t wave of translation which represents, in the sea, 
 the displacement eflected by the emergent movements of land. 
 
 381. Area of Disturbance. — The area over which earth- 
 quake movements are felt is often considerable; but an uncer- 
 tainty prevails when very gi-eat distances are alleged to be 
 included imder one movement, more especially when sea 
 intervenes between the different points affected ; for we 
 cannot be sure that there have not been intermediate points 
 of disturbance which have escaped observation, while their 
 results seem to carry forward the one recorded event. 
 
 382. Distribution of Earthquakes. — The earthquake areas 
 on the surface of the globe correspond generally to the vol- 
 canic districts; and it is noteworthy that, for example in 
 the case of the Mediterranean, disturbances extend upon 
 either side of the long axis which passes through that 
 region, the phenomena showing, at least northwards, where 
 they have been best observed, gradual diminution of intensity. 
 Beyond the Alps tremors are experienced even as far as the 
 British Islands; and the considerable disturbance which co- 
 incided with the great earthquake of Lisbon in 1755, was 
 looked upon as proof of the extent to which that shock 
 reached. It is possible, however, that the tremor in Scot- 
 land was a simultaneous — one might call it a sympathetic — 
 disturbance, due to alterations in the subterranean cavities 
 consequent upon the great change to the south. The process, 
 Avhatever it was, which overthrew Lisbon, probably disturbed 
 the relations of the fluid cavities which, as has been already 
 stated, are believed by Sir William Thomson to exist. 
 
 383. Causes of Volcanoes and Earthquakes. — The belief 
 in the common origin of earthquakes and volcanoes rests 
 upon the very frequent coincidence of earthquake movements 
 with volcanic outbursts ; and probably no great developments 
 of the latter ever take place without very important develop- 
 ments of the former. In the Mediterranean area a remarkable 
 alternation has been observed between the volcanic eruptions 
 of the Archipelago and the earthquakes of Syria; and the 
 
312 PHYSICAL GEOGRAPHY. 
 
 same relation is believed to exist between Ischia and Vesu- 
 vius. These alternations are similar to those whereby the 
 volcanoes of the great American chain have their maximum 
 intensities at different periods, and it is possible to explain 
 the phenomena by reference to the existence of subterranean 
 lava lakes. 
 
 384. Hypothesis of Connection hetween Sea and Vol- 
 canic Centres.— It has been already more than once remarked 
 that the majority of volcanoes are in the vicinity of the 
 ocean, Jorullo, in Mexico, being but an apparent exception, 
 since it seems to be a member of a chain, the great part of 
 which is certainly close to the shore. The volcanoes in 
 the Caucasus are distant indeed from the ocean, but close to 
 the borders of the Caspian Sea. The constancy of this rela- 
 tion has suggested the probability of water being the principal 
 agent by which volcanic activity is called into operation. 
 The passage of water downwards into the volcanic foci, seems 
 to have the effect of calling into fresh activity those chemical 
 changes by which heat is evolved, and, as a consequence, the 
 expansive power of vapours is increased, and even the solids 
 themselves come to occupy a larger cubic space. Common 
 salt has been obtained from the fumes of Vesuvius — is 
 even thickly de^Dosited with other chlorides after eruptions. 
 Hydrogen is known to escape from volcanoes, although its 
 flame is not readily detected amidst the more powerful light 
 of red-hot cinders. In its discharge and conversion into 
 water Avhen burnt in the open air, we find one explanation 
 at least of the steam which constantly occurs in eruptions. 
 The absence of magnesia, which seems a difficulty in the way 
 of this theory, is exj)licable by the circumstance that the 
 chloride of magnesium is decomposed into hydrochloric acid 
 and magnesia, the latter coming to form a very important 
 constituent of lava (Art. 366). That materials from the 
 the surface have been introduced is well known from the 
 fact that Vesuvius has ejected, from time to time, infusorial 
 cases amongst the ash, these animal remains having obviously 
 reached the interior by fissures of some sort. It is not 
 quite so easy to imderstand the source of the nitrogen which 
 is obtained from the craters of active volcanoes, and is 
 observed in the waters of thermal springs. Perhaps tho 
 
VAPOUH AND THERMAL SPRINGS. Si 3 
 
 only plausible explanation of its occurrence is that, as the 
 columns of molten matter and of heated water surge upwards 
 and do^vnwards in the supply pipe, the vacuum created by 
 their sudden retirement becomes filled with air, which ulti- 
 mately reaches the interior. It is not altogether out of 
 place to recall the suggestion put forward by Sir Charles 
 Lyell, that the loss of heat which the earth is kno^vn to 
 sustain may be replaced in some measure by electro-magnetic 
 force from the sun. It is defended by him on the ground 
 that, although it may appear like an attempt to establish 
 perpetual motion, our knowledge does not yet permit us to 
 be content with an epigrammatic condemnation of the sugges- 
 tion, since farther knowledge might prove the possibility of 
 the hypothesis, just as observation has established the unex- 
 pected fact that the radiation of heat is retarded by atmo- 
 spheric moisture. It must of course be remembered that 
 the tendency of all chemical change is towards equilibrium, 
 and that equilibrium must ultimately be arrived at, further 
 change being thereafter impossible without the introduction 
 of fresh material, or the disruptive action of some force 
 different in its manifestations from that of ordinary chemical 
 combination. The only question of any irnportance is as to 
 the rate at which we are tending towards equilibrium, as 
 to the rate therefore at which it is probable physical and 
 organic changes Avent on in the past; and on this problem 
 we are scarcely yet in a condition to speculate profitably. 
 
 385. Vapour and Thermal Springs. — It has been men- 
 tioned that some volcanoes, as those of the Andes, emit lava 
 comparatively rarely, and chiefly give escape to ashes and 
 vapours. One step more brings us to those orifices from 
 which steam alone, or hot water alone, or gaseous vapours 
 alone are discharged, a solid material, even in the form of 
 fine ashes, never accom])anying the emission. To this group 
 belong the geysers, or hot springs of Iceland, of the extinct 
 volcano of Ischia, of the island of St. Paul's, and many other 
 localities which are obviously directly associated with volcanic 
 activity, either in the past or the present; second, the solfa- 
 taras, from which sulphurous vapours alone are emitted, these 
 being either on the volcanic cone or near it, or, as sometimes 
 happens, at considerable distances from craters active or 
 
314 PHYSICAL GEOGRAPHY. 
 
 extinct; tliircl, fumarolos, frora whicli boraclc acid is dis- 
 cliarged; fourth, naphtha; or, fifth, carbonic acid vapour. To 
 the same category belong the sulphurous, siliceous, and gyp- 
 seous springs in various regions, as in Yellowstone Park 
 (Art. 160), Savoy, and Germany, where, at least within 
 recent geological times, no volcanic activity has been mani- 
 fested. Chemically, the substances found in these springs 
 and lakes are identical with those obtained from volcanoes; 
 and although it might be difficult to establish by direct proof 
 the actuaJ connection between all these kinds of phenomena, 
 still the probabilities are in favour of, at least, the community 
 of their origin. 
 
 386. Intermission of Geysers. — A very interesting physi- 
 cal problem is associated with the intermittence of the Ice- 
 landic geysei's, and Tyndall has given a satisfactory explanation 
 of it, and an illustration by a simple experiment. He carried 
 down a metal tube from the centre of a basin full of water, 
 and surrounded the bottom of the tube, as well as a part of 
 its length, with a ring of fire. The water being thus sub- 
 jected to a considerable heat at two points, he procured 
 eruptions of hot water and steam at irregular intervals of five 
 minutes; for the water at the bottom, becoming heated, ex- 
 panded and lifted the water above it for a certain distance. 
 Relieved to some extent of pressure, and its boiling point 
 thus lowered (as in passing from thirty-eight to thirty-two 
 feet), it expanded into steam, and the heat evolved in the 
 process generating steam in the mass beneath, the whole 
 suddenly burst into ebullition and propelled the superincum- 
 bent mass out of the tube. It fell back chilled into the 
 basin, descended again into the tube, and the process went 
 on again until the temperature of the whole mass was suffici- 
 ently raised to permit of another explosion. The application 
 of this experiment to the geysers relieves us of the necessity of 
 imagining underground caverns containing water and steam, 
 and restricts the mechanical production of the phenomena to 
 the heating of the fissure through which the spring rises. In 
 the subjoined table the boiling temperatures are those at 
 which water should boil at that depth and pressure : the ob- 
 served temperatures are the actual ones ascertained by Bun- 
 sen in the tube of the great geyser, and these are below the 
 
Secular movements of the earth's crust. 315 
 
 boiling point ; it is obvious tbat the water must be raised in 
 tlie tube before it can pass into steam. 
 
 Observed Temperature, Feet. Boiling Temperature, 
 
 85-5° C. 6 107° a 
 
 110° 22 116° 
 
 32 120-8° 
 
 121-8' 38 123-8° 
 
 124° 50 130° 
 
 126° 64 136 
 
 Of the effects produced by these and other sj)rings, which 
 contribute solid matter to the sedimentary strata, enough has 
 been said in a previous chapter Avhen treating of springs. 
 
 387. Periodicity of Earthquakes and Volcanoes. — It 
 only remains to speak of the periodicity which, it is alleged, 
 may be observed in earthquake and volcanic phenomena. That 
 tolerably equal intervals have been noted, as the thirteen 
 year periods of Icelandic disturbances, is true ; but the rai-ity 
 of such observations, when taken in conjunction "svith the much 
 better observed irregularity of the events in other regions, 
 makes it probable that the supposed periodicity rests upon 
 coincidences. The history of Vesuvius, as given by Professor 
 Phillips, and of Etna, does not bear out any periodicity of 
 either, or even any regularity of their alternations. Only 
 one observation seems to suggest the possibility of external 
 influences securing regularity in the phenomena, namely, that 
 the great majority of volcanic eruptions have taken place in 
 winter. M. Perrey believes that there is a greater amount of 
 activity when the moon is nearest the earth, and when the 
 earth is in perihelion. If a sufficient number of observations, 
 sufficiently authenticated, should confirm this suggestion, a 
 certain amount of periodicity might be traceable to thcso 
 astronomical influences; but in the meantime the data are 
 insufficient to warrant the general conclusion, which is 
 scarcely reconcilable with the teaching of physicists regard- 
 ing the effects of tides in the earth's interior. 
 
 388. Secular Movements of the Earth's Crust. — Inti- 
 mately associated with this subject is that of slow elevations 
 and subsidences. It is known that rocks imdergo a certain 
 change of dimension in passing from the fluid to the solid 
 state. The estimates of the amount of this change vary 
 considerably^ thus Bischofs calculations make it appear 
 
316 PHYSICAL GEOGRAPHY. 
 
 tliat granite suffers a diminution in volume of twenty-five 
 per cent, in passing from the fluid to the crystalline con- 
 dition, while Delesse calculates that the contraction is only 
 nine to ten per cent. These materials, however, occupy a 
 comparatively small part of the earth's crust, while the sand- 
 stone, upon which extensive observations have been made, are 
 more important as regards their mass, and, therefore, as re- 
 gards the influence they may perchance have. Lyell calcu- 
 lates that, according to the data given by the experiments of 
 Totten on building stones, a mass of sandstone a mile in 
 thickness would, if raised to a temperature of 93*3° C, lift 
 the rock above to a height of ten feet, while the heating of 
 a mass fifty miles in thickness to a temperature of 316° or 
 426° might yield an elevation of 1000 or 1500 feet, the sub- 
 sequent cooling producing a corresponding amount of depres- 
 sion. The contraction, again, of clay rocks under the influence 
 of high temperature might yield subsidence. Taking this in 
 conjunction with the supposed falling in of the roofs of sub- 
 terranean chambers, we have, in the volcanic foci, an adequate 
 cause at once for the violent and the slower variations of level 
 at the surface; while, as has already been suggested, the per- 
 colations of springs below ground may likewise, oj the removal 
 of soluble rocks, lead to depressions on the smaller scale. 
 
 It is alleged that the globe is still undergoing contraction, 
 and that the elevation of mountain chains is attributable to 
 this cause, which also takes share in the production of vol- 
 canic outbursts. But it is difficult to adopt this view, for 
 we have abundant evidence of rej^eated elevations and sub- 
 sidences in the same area at very different periods, and to 
 very unequal amounts. Africa, though a continent for an 
 immense period, has no gi'eat mountain clmins such as this 
 theory would require, and some of the oldest mountain chains 
 have not that enormous height which their antiquity might 
 be expected to involve, while the highest chains are those of 
 most recent date. It must on the other hand be remembered 
 that the great ocean basins are of great antiquity, and that, 
 while subordinate movements have occurred in abundance, 
 the geological record only tells of one grand alternation, that 
 whereby the Atlantic and Pacific, once land, have become sea, 
 while land has taken the place of the former great oceans. 
 
CHAPTER IX. 
 
 DISTRIBUTION OF PLANTS AND ANIMALS. 
 
 Fauna and Flora: Aspects of Life — Relation of Existing to Former 
 Faunas and Floras — Aspect orFacies of a Region: How Determined 
 — Laws of Distribution — Non-coincidence of Botanical and Zoo- 
 logical Provinces — Aquatic and Subaerial Animals : Not Essen- 
 tially Different — Influence of Climate — Parallel Regions in 
 Latitude and Altitude — Marine Batlij-metrical Zones — Pro\*inces 
 Determined by Physical Conditions — Specific Centres — Biological 
 Provinces Laiequal — Sclater's Provinces: Neotropical; Ethiopian; 
 Indian; Australian; Palfearctic — Common Character of Neo- 
 tropical, Ethiopian, Indian, and Australian Provinces — Ana- 
 logous or Representative Forms — Migration of Species and Ex- 
 tension of Area— Results of ^Migration — Natural and Artificial 
 Selection: Survival of Fittest — Variations: How Beneficial — 
 Mimicry : Protective and Independent Resemblances — Repre- 
 sentative Species — Dangers Incident to Migi'ation — Tabular View 
 of Organic "World — Homotaxis — Insular Faunas and Floras — 
 Hj'pothesis of Lost Continents — ^Marine Provinces : N. Atlantic ; 
 Caribbean; Indo-Pacific ; Australian; Western S. America — 
 Pelagic Forms — Deep-sea Faunas — Continuity of the Cretaceous 
 Epoch — Extension and Replacement of Species — Persistent Typos 
 — Progressive Development. 
 
 389. Fauna and Flora: Aspect of Life. — The animals 
 "wliicli inhabit any area constitute its fauna; the plants con- 
 stitute its flora : and these terms are equally applicable to 
 the inhabitants of contiguous or of distant areas. They are, 
 in fact, quantitative terms, while the qualitative comparison 
 of these faunas and floras shows that there are diflerenccs 
 which give to each fauna and flora its aspect or fades. The 
 aspects may be identical in closely-contiguous areas, utterly 
 uiilike when distant areas under dissimilar conditions are 
 compared, or representative of each other in the case of dis- 
 tant localities under similar conditions. It has already been 
 gtated that the provinces of the earth at the present day are 
 
318 PHYSICAL GEOGRAPHY. 
 
 tlie same in kind as tlie provinces of former periods, Tlie 
 comparison of the faunas and floras of tlie past with those of 
 the present shows that the plants and animals of the present 
 are very closely allied to, if not identical with, those which 
 flonrished in the most recent times over the same area; that 
 the difference between the existing and former faunas and 
 floras increases the further back we 'go, and that the organic 
 forms in one locality at one period frequently strongly re- 
 semble those of another locality at another period. 
 
 390. Relation of Existing to Former Faunas and Floras. 
 ■ — The fact that the most recent fossils are of the same genera, 
 or even species, as the living beings is clearest where, as in 
 South America and Australia, the characteristic forms are 
 restricted v/ithin narrower limits. The gravel and cave bones 
 are of marsupials in the one case, of sloths or their allies in 
 the other; but equally good, though less-striking, examples 
 are to be found in every country. The wealden fauna and flora 
 are very unlike those of Europe now; but there are points 
 of resemblance to those of Australia; and it may be said, in 
 general terms, that the mesozoic animals have an aspect to 
 which at the present time the Australian area offers the 
 only resemblance. Principal Dawson has shown that the 
 pfe-carboniferous flora of N. America has greater affinity to 
 the secondary flora of Europe than to the carboniferous or 
 subsequent floras of America. But it must be remembered 
 that in plants the non-preservation of the parts essential for 
 safe classification, these being usually of very soft tissues, 
 places great difiiculty in the way of this systematist, and 
 renders his conclusions imperfect and insecure. 
 
 391. Aspect or Facies of a^Region : How Determined. — 
 From what has been said in earlier chapters it is plain that 
 the modifications in texture or composition of rocks must be 
 great before they can give rise to marked differences between 
 two countries which have been subjected to precisely similar 
 influences. The physical aspect, therefore, results from a 
 smaller number of factors than the organic. But while the 
 practised eye may gather the main points in the geological 
 history of a country during a rapid visit, general impressions 
 do not count for much as regards the life of a country. 
 Brilliant descriptions of tropical vegetation, after all, teach 
 
ASPECT OR FACIES OF A REGION. 319 
 
 little except tlie well-known fact that great lieat favours 
 anotlier kind of vegetation than that found in temperate or, 
 still more, in cold regions; or the generalization may be 
 extended to this, that endogenous plants are more prominent 
 than exogenous in the tropics. But if by facies or aspect 
 is meant, not the obvious featiu^es merely, but those which 
 give individuality to particular regions, we find that here, as 
 in the domain of physics, some standard more reliable than 
 that of the senses must be appealed to. The following table, 
 given by Pokorny, shows how changing is the statistical guide, 
 even in the case of plants, which are, in one way, more easily 
 obtained and enumerated than animals : — 
 
 Linuceus enumerated in 1754, about 7728 species 
 Persoon ,, 1801, ,, 21,000 „ 
 
 Sprengel ,, 1828, ,, 30,000 ,, 
 
 Stendel „ i840, ,, 87,000 „ 
 
 linger „ 1852 ,, 92,6G2 ,, 
 
 In 1859, the number of species was variously estimated, 
 according to Hooker, at 80,000, and 150,000 according to 
 the opinion of the systematist as to what constitutes a species. 
 And this source of difficulty increases with the number of 
 new forms observed in new localities explored. 
 
 But these numbers do not, after all, represent a fair sum- 
 mary of the whole earth. Some regions are as little known 
 as others have been exhaustively investigated; and this is 
 still more true for the animal kingdom. The lists of fossils, 
 necessarily incomplete for any region, since we have not 
 exhausted, and never shall exhaust, the contents of its rocks, 
 are for most countries fragmentary, while in many the explo- 
 ration can scarcely be said to have commenced. If, therefore, 
 the aspect of the life of any region at any period helps us to 
 trace out its geographical history, we must qualify our con- 
 clusions by the recollection of our imperfect knowledge, and, 
 setting aside our impressions, trust only to careful enumera- 
 tions of all the species that have been recorded. It is a 
 common impression that a group of animals tells by simple 
 inspection its native place: but, in reality, the judgment is 
 based, not on the impression conveyed by the whole, but on 
 the rapid recognition of the individual species which make 
 up the group. 
 
320 PHYSICAL GEOGKAPHY. 
 
 392. Laws of Distribution. — The facts as to tlie presence 
 or absence, tlie luxuriance or insignificance, of species, genera, 
 or families, in particular regions, and the conditions of tem- 
 perature and the like which these regions present, have been 
 embodied in propositions which are spoken of as the laws of 
 distribution. The phrase is an unfortunate one, since it 
 suggests too great a value for the results of experience in 
 this branch of science. All laws are, in one sense, the fruit 
 of experience, and their experimental verification is the test 
 of their scientific value, for they can be received as laws only 
 if they enable us to announce what is past, and to anticijDate 
 the future. The law of the attraction of bodies is so exact 
 as to lead to the recognition of the exsistence of planets, 
 before they have been seen, the influence of gravitation being 
 determinable at any point in the orbit of a planet. The 
 mechanical theory of heat enabled James Thomson and 
 Magnus to arrive simultaneously at the conclusion that the 
 freezing point might be lowered by pressure, and the experi- 
 mental verification of this opinion proved it to be an accurate 
 deduction from an exact generalization, a law in the scientific 
 sense of the word. The constant sequence of phenomena in 
 these two cases is, so far as our knowledge and experiments 
 go, uninterrupted; but the necessary connection between 
 the antecedents and consequents in any phenomenon, that 
 peculiarity of the things which makes their relation to others 
 not merely invariable but inevitable, it is impossible for us 
 to ascertain. But while the phenomena of the inorganic 
 world manifest the nearest approach to absolute certainty 
 with which experience makes us acquainted, those of the 
 organic world display an uncertainty which increases with 
 their complexity, or, to use more precise language, their 
 sequence is obscured by the number of contending influences. 
 There is no law operative in the inorganic world which is 
 not also operative in the organic; but as the tissues which 
 make up an organised body are multiplied, their proportions 
 vary, and the j^i'eponderance of particular parts undergoes 
 change. Hence the sequence of phenomena, which by 
 analogy we are assured must be normal, is concealed, because 
 we cannot isolate events. Our estimates, therefore, of the 
 power of any one kind of influence on an organised body, 
 
LAWS OP DISTRIBUTION. 321 
 
 are only reliable in proportion to the number of instances on 
 which they are based; for it is only by multiplying the cases 
 that we can diminish the chances of error. The theoiy of 
 probabilities teaches that a certain event must recur in a 
 certain number of trials, when we know the number of con- 
 tingencies upon which it depends; but in the organic world 
 the number of contingencies is not and cannot be known, 
 so that a generalization based on 1000 instances may be at 
 variance with the next instance, the details of which are 
 unkno^vn and cannot be foreseen. And if this is true for any 
 one physiological process, the results of which are compared 
 in 1000 individuals, it must be true also for the sum of the 
 processes which take place in 1000 individuals. The external 
 influences to which organic beings are subjected are com- 
 paratively few in number and kind; but the internal modi- 
 fying influences are indefinite in number and variable in kind. 
 In previous chapters it has been shown that these external 
 influences are perpetually changing ; the multiplication of 
 instances, therefore, is of itself apt to multij)ly the chances of 
 error, since, if they are taken over a wide range in sj^ace or 
 in time, the introduction of new influences is more probable. 
 Now, as plants and animals are in themselves thus variable, 
 and as the external influences are likewise liable to change, 
 any general proposition regarding the distribution of organic 
 beings is true only for the cases that have been observed. 
 It is not, and cannot be, a law, as that word has already been 
 defined, for it cannot make us sure that the like has 
 happened in the past or will in the future. The distribu- 
 tion of the elephant at the present time would not lead us 
 to expect that the remains of any member of that genus 
 should be found in Britain or Siberia, still less would the 
 characters of existing genera have led us to believe that any 
 members of the family had been protected by a hairy cover- 
 ing. If it is impossible to assign a higher value than that of 
 a careful summary of facts to the results of our comparison 
 of many individuals, or groups of individuals, i.e., classifica- 
 tions, it is still less possible to attach greater importance to 
 the results of comparison of the faunas and floras of different 
 regions, I.e., botanical and zoological geography. The ante- 
 cedents in many cases are connected Avith the consequents in 
 23 -I 
 
622 PHYSICAL GEOGRAPHY. 
 
 a way wliicli it is impossible for us to understand; tlius we 
 say that increase or diminution of temperature has driven out 
 the animals of a region, but we cannot frame an explanation 
 of the way in which heat or cold affects an animal or vege- 
 table, so that the difterence of a few degrees in the annual 
 average may render a country favourable to it, or the reverse. 
 
 393. Non-coincidence of Botanical and Zoological Pro- 
 vinces. — Botanical and zoological geography do not coincide ; 
 the natural provinces of the one or other kind of life have 
 different limits and centres, as might be expected, when we 
 consider their different physiological endowments and sus- 
 ceptibility to external influence. 
 
 394. Aquatic and Subaerial Provinces. — The most 
 obvious distinction of distributional areas is that of aquatic 
 and subaerial, and the aquatic forms are restiicted to fresh 
 or salt water, to lakes or rivers. But the difference between 
 gills and lungs is more of form than essence : in the one case 
 the oxygen is held in suspension in the water ; in the other 
 it is in mass in the atmosphere. The distinction between 
 aquatic and subaerial animals, though greater than that 
 between aquatic and subaerial plants, is after all less than is 
 commonly supposed; for in both cases the gas passes through 
 membrane to reach the circulating blood, and the temperature 
 of the animal is in proportion to the amount of air absorbed, 
 which is least where the water flows over the membranous 
 projections covering blood-vessels, and increasingly greater 
 vv^here the air is received into membranous sacs, on the out- 
 side of which the vessels are distributed. The possession of 
 gills is permanent or temporary in the life of the individual ; 
 thus the fish is aquatic throughout life, but in the groups of 
 amphibians some retain gills permanently, others speedly 
 lose the gills and acquii*e lung sacs; while in a third group 
 both organs are present, and are alternately wsed as the 
 necessity arises. 
 
 395. Climate. — If vital phenomena were as limited in 
 their range as are inorganic, or, in common phrase, physical 
 processes, the isothermal lines might be expected to corre- 
 spond more or less closely to the limits of larger or smaller 
 groups. The correspondence is, however, not very exact, 
 since physical features, affecting atmospheric and oceanig 
 
PARALLEL REGIONS IN LATITUDE AND ALTITUDE. 323 
 
 currents, cause local variations from the normal temperature, 
 or that proper to tlie latitude; but more striking is the want 
 of correspondence when the seasonal temperatures are con- 
 sidered in place of the mean annual temperatures. Thus the 
 annual migrations of birds, as of the swallows from the south, 
 the fieldfares from the north, into our own area, are examples 
 of the uncertain limits which average temperature imposes 
 on distribution. But, besides these extreme cases, insjoec- 
 tion of the maps of distribution shows that the same species 
 may have very wide limits, passing through several zones of 
 temperature; thus the bison and Virginian opossum range 
 from Canada to the Mexican Gidf. The tailless hare and 
 the tiger extend from about 55*^ N. lat. as far south as Java, 
 but their range to east and west is restricted. This absence 
 of animals from adjacent regions under the same conditions 
 is one of the reasons for assigning to climate a subordinate 
 influence in distribution. 
 
 396. Parallel Regions in Latitude and Altitude. — 
 Ascending a mountain under the equator brings the traveller 
 into successive belts of temperature, which repeat the expe- 
 rience of one passing towards the poles, and as each zone in 
 latitude has a characteristic aspect of life, so have the corre- 
 sponding belts in altitude ; the snow line is the limit of 
 abundant life in both cases. The following are the zones of 
 distribution of plants useful to man:"' — • 
 
 Zone. 
 Tropical, . . . . > 
 Subtropical or warm 
 
 temperate, 
 Temperate, 
 
 Subarctic, 
 Arctic, 
 
 Approximate Latitudes. Charactei-istics. 
 
 0° to 23° 30' ^ ^ic6» iiaaize, palms, spices, 
 
 sugar. 
 
 23" 30' to 45° 
 45no55° 
 
 \ Wheat and tropical grains, 
 \ Olive, fig, grape, citron. 
 ( Wheat and northern grains. 
 < Orchard fruits, deciduous 
 ( leaved trees. 
 I 55° to 66° 30' \ Northern grains. Berries, 
 
 66° 30' to 90° Saxifrage, mosses, hchens. 
 
 The eight zones tabulated by Meyen give undue imjoort- 
 ance to mere temperature, and are therefore artificial in their 
 construction, as well as arbitrary in their grouping, of the 
 plants in each. The zones in altitude corresponding to these 
 
 '■^ Yeats, 
 
324 PHYSICAL G:]OGrvAPHY. 
 
 horizontal areas are shown on the figure, Student^ s Physical 
 Atlas, Map XVIII., and their temperate limits ai-e better 
 founded. They are, in ascending order, the regions of — 
 (1), bananas and palms on the low grounds; (2), tree-ferns 
 and figs, up to 2020 ft. ; (3), of myrtles and laurels, 4050 ft. ; 
 (4), of evergreen dicotyledonous trees, 6120 ft.; (5), of 
 European dicotyledonous trees, 8100 ft. ; (6), of pines, 10,140 
 ft.; (7), of rhododendrons, 12,150 ft.; (8), of Alpine plants, 
 14,170 ft.; (9), the plantless region, commencing with the 
 snow line, 16,200 ft. The correspondence of these two kinds 
 of zones is not, of course, exact ; but the agreement is in 
 keeping with what has been already said (Art. 207) of the 
 shells of temperature with which the earth may be regarded 
 as surrounded. 
 
 397. Marine Provinces: Bathymetrical Zones. — The 
 areas of distribution in the sea are less distinct in climatal 
 demarcation than those of land, the temperature being more 
 uniform. The belts distinuished by Professor E. Eoi'bes are, 
 therefore, purely topographical, and approximately correct 
 for every region, representative forms occupying correspond- 
 ing positions. The five bathymetrical areas are : 1. The 
 littoral zone, between tide marks, or at the water's edge in 
 tideless seas. 2. The circum-littoral zone, from low tide to 
 15 fathoms. 3. The median zone, from 15 to 50 fathoms. 
 4. The infra -median, from 50 to 100 fathoms. 5. The 
 abyssal zone, from 100 fathoms downwards. Since it is now 
 known that life descends to the bottom of the ocean, the 
 fifth zone may yet be subdivided, but the data are in the 
 meanwhile imperfect. 
 
 398. Provinces Determined by Physical Conditions. — 
 The influence of temperature is controlled by the physical 
 features of the land above and below water. A mountain 
 chain, such as the Andes, is an insuperable barrier to the 
 passage of plants or animals. The deserts of Africa separate 
 regions as sharply as does the sea. In the sea, where a 
 superficial view might lead us to expect great uniformity, 
 there are provinces recognisable, though their demarcation 
 is not always clear. But a current may be a bar as absolute 
 as a mountain range, and a deep central trough, as that of 
 the Atlantic, seems as eflLCctual a means of separation as a 
 
SPECIFIC CENTRES. 325 
 
 desert. In the N. Atlantic basin warm and cold areas wero 
 described as connected with the features of the sea bottom 
 (Arts. 75-85), and by their interlacement the faunas of 
 northern and southern regions are found under the same 
 parallels. Among the anomalies of distribution must be 
 mentioned the absence of animals from an adjacent area from 
 which they are separated by obstacles seemingly out of pro- 
 portion to the result. Thus the distinctness of Madagascar 
 from Africa, and of the two parts of the Malayan Archi- 
 pelago, even of Ireland from Great Britain, leads us to what 
 is the primary cause of the features of the botanical and 
 zoological geography at the present time. 
 
 399. Specific Centres. — Plants and animals are grouped 
 under species, and these under genera; these again imder 
 families, orders, classes. The number of species included 
 under a genus varies, and the number of varieties under a 
 species is likewise unequal. But the proportions in both 
 cases have reference to the extent of country they cover, and 
 to the variety of its surface. There is little variety in the 
 animals of a country which, like Africa, is remarkable for 
 uniformity; on the other hand, the changes of conditions 
 jDresented by such a region as the Amazon valley are 
 associated with the presence of many local varieties among 
 the butterflies, as Mr. Bates has described. Though they do 
 not prove, these two converse facts support the view, that 
 species are modified as they spread from the area or centre 
 in which they first appeared. It is now accepted as indis- 
 putable, that species which have become extinct, do not 
 recur in the geological series, and the forms which have 
 descended almost unchanged from remote times to the present 
 confirm us in the belief that continuity of descent is insepar- 
 able from continuity of character. The isolation of specific 
 or generic types at the present time is, therefore, explicable 
 by reference to the geographical changes which have been 
 described in previous chapters of this volume. It is beyond 
 the scope of such a book as this to discuss the question of 
 the origin of species; but the view here adopted is, that 
 species are developed out of other species by modification. 
 Considerable difference of opinion exists as to the nature of 
 the processes by which this modification is accomplished; 
 
32^ PHYSICAL GEOGRAPHY. 
 
 but the theory affords the only means of co-ordinating a mas^ 
 of facts which no other doctrine yet proposed has brought 
 into harmony. To physical changes, therefore, must be 
 assigned the primary influence in bringing about the distri- 
 bution of organic forms such as we now see. 
 
 400. Biological Provinces Unequal. — Reference has been 
 made in previous paragraphs to the unequal plasticity of 
 organised forms, to the unequal power they possess of endur- 
 ing vicissitudes in the conditions to which they are subjected. 
 It may be said in general terms that this power of endurance 
 is in proportion to the simplicity of the organism; that the 
 simplest forms are those which have the widest range iu 
 space, and have had the longest existence in time. As the 
 complexity of an organism increases, its power of endurance 
 diminishes, as a rule, in the same proportion; so that the 
 continued existence of its offspring depends upon their 
 capacity to change at the same rate as the conditions alter. 
 If the capacity is limited, extinction of the type must ensue. 
 If the capacity is great, or if the external changes are slow, 
 the modification gives rise to new species. In some groups, 
 both of plants and animals, the changes in a given time must 
 be greater than in others, and after long intervals the specific 
 centres, say of the grasses or of the molluscs, must have 
 undergone considerable change of position. A series of maps, 
 each devoted to the distribution of a distinct species or genus, 
 would present very great differences, the amount of which 
 may be inferred from the comparison of those given in Mr. 
 A. MMYrsifB Distribution of Mammals, 1866. But while the 
 centres of maximum development of animal and vegetable 
 species do not coincide, there is sufficient agreement to render 
 possible the division of the earth into regions, each of which 
 is characterised by the presence of an assemblage of plants 
 and animals whose association has reference to the previous 
 geographical conditions of the areas. 
 
 401. Sclater's Provinces. — The provinces, which thus 
 harmonize biological and geographical changes, are those first 
 set forth by Dr. Sclater for the birds, and since approved by 
 Mr. Wallace for the animal kingdom generally, while they 
 include tolerably well the main facts as to the distribution of 
 plants. 
 
ETHIOPIAN REGI02T. 827 
 
 Regions. 
 
 Areas. 
 
 1. Keotropical, . 
 
 2. Nearctic, . . 
 
 3. Palasarctic, 
 
 4. Ethiopian, . . 
 
 5. Indian, . . . 
 
 6. Australian, 
 
 South America, Mexico, West Indies. 
 
 The rest of America. 
 
 Europe; Asia, north of the Himalayas, as far 
 
 as Japan; Africa, north of the Sahara. 
 The rest of Africa; Madagascar. 
 Southern Asia; west half of the Malayan 
 
 Archipelago. 
 East half of Malayan Archipelago; Australia; 
 
 and most of the Pacific Islands. 
 
 402. Neotropical Region. — South America contains, as 
 its characteristic animal forms, the platyrhine monkej^s, 
 which are distinguished from the Old World forms by the 
 thicker septum between the nostrils, and by the possession 
 of 36 not 32 teeth. The antiquity of the province is shown 
 by the exclusive presence in the tertiary dejDOsits of monkeys 
 belonging to this type, no member of the Old World group 
 having been discovered. The sloths and armadilloes are 
 similarly related to their predecessors, the megatherium, 
 glyptodon, and the like, and Mr. Bates regards the arboreal 
 habit of the sloth as acquired during long residence in a 
 ■wooded country. The llama, alpaca, and guanaco, form a 
 characteristic group confined to the higher plains of the 
 eastern slopes of the continent, while the vampire bats are 
 also confined entirely to this region. 
 
 403. Ethiopian Region. — The submersion of the Sahara 
 down to tertiary times, formed a barrier which helps to 
 explain the separation of the north coast of Africa from its 
 tropical and southern districts, and the specific distinctness 
 of its elephant, of its three species of rhinoceros, and the 
 restriction to it of the hippopotamus. The chimpanzee and 
 gorilla are the anthropoid apes of this region, while the 
 lower groups are numerous, but all belong to the catandiine 
 division. Madagascar, separated from Africa by the Mozam- 
 bique Channel, contains very few mammals common to it 
 and the continent (Art. 419). The characteristic lemurs 
 have been found fossil in Europe, and the difference between 
 two countries so near, and placed under similar climatal 
 conditions, indicates the antiquity of their separation. 
 
328 PHYSICAL GEOGRAPHY. 
 
 404. Indian Region. — The eastern boundary of tMs area, 
 the strait between Borneo and Bali on the one hand, Celebes 
 and Lonibok on the other, is only fifteen miles broad, but its 
 depth is greater than 100 fathoms, and this corresponds to a 
 geographical separation of great antiquity. The western 
 boundary is less definite geographically, the Ethiopian and 
 Indian types being mingled on the continental land surface 
 of j^rabia. The scitaminese, zingiberaceae, and bananas are 
 the most characteristic forms of plants, and the fitness of 
 local climates for certain species of food plants from other 
 regions, has been experimentally proved. The elephant, 
 tapir, and rhinoceros are wide spread over the area, and the 
 species of ox {Bos gaurus and B. gavialis), though kindred 
 to the S. African species, have remarkably local areas. The 
 reptiles of the Indian area are well defined, the exceptional 
 occurrence of tropical snakes in Japapi, and of a few tropical 
 butterflies, telling of a former southern land connection of 
 Japan, whereby these Indian forms gained a footing among 
 a mass of palsearctic species. They represent groups which 
 otherwise would have been unrepresented, and their continu- 
 ance looks like the fulfilment of a function in nature. 
 
 405. Australian Region. — But though mammals, birds, 
 and reptiles agree in defining the Indian region by the Straits 
 of Lombok, the insects wander over the line on both sides, 
 the mixture of species being here due to power of flight, as 
 in Arabia it was due to absence of obstacles. The other 
 groujDS of animals leave no doubt as to the definiteness of the 
 Australian region. The marsupials, and the absence of 
 placental mammals, are the leading features of the greater 
 portion of the land. Among birds, the emu, mound-building 
 megapodius, the honey-suckers, and loris, form a characteristic 
 assemblage, while the kivi or ajDteryx still represents the 
 struthious birds, of which the moa was the last survivor, 
 having become extinct probably within the last century. 
 The gum trees, eucalypti, heaths, epacrides, the proteacese, 
 casuarinas, and gummiferous acacias are among the most 
 characteristic for the continent, but the variety of climate 
 which difierent portions of its borders ofler, is accompanied 
 by local peculiarities in the grouping of the plants. The 
 grass plains of the interior are like those of the north tem- 
 
PARALLEL REGIONS IN LATITUDE AND ALTITUDE. 329 
 
 perate zone; the scrub is peculiar, since no turf accompanies 
 it, and the heath -like, vertically placed leaves offer little 
 protection to the soil beneath. The combination of American, 
 Australian, and Antarctic plants in New Zealand, the latter 
 being confined to the high gi'ound, renders it probable that 
 the present aspect of the vegetation was acquired after many 
 great changes of the geography of the South Pacific, and 
 after the dispersive influence of a southern glacial epoch had 
 been felt. 
 
 406. Palaearctic Region. — The Europeo-Asiatic continent 
 from Japan to Spain, and to the north of a line from the 
 Atlas through the north of Arabia and the Himalayas, pre- 
 sents a singular uniformity in the distribution of mammals, 
 76 per cent, of European species being common to Amoor- 
 land, while Algeria contains forty-seven sjDecies and about 
 twenty-eight genera common to the lands north and east of 
 the Mediterranean. There is a similar agreement as regards 
 the other groups of vertebrates, with the exception of the 
 Japanese serpents, which are tropical. The land molluscs 
 ngree, on the whole, with the vertebrates; and the plants 
 likewise render this a natural province, save in the eastern 
 Asiatic region, in which American affinities indicate relation- 
 ship to the miocene flora. 
 
 407. Common Character of Neotropical, Ethiopian, 
 Indian, and Australian Regions. — The bird fauna of the 
 great area thus included is of great interest. To these regions 
 are confined the struthious birds, including tlie ostrich of 
 Africa, emu and apteryx of Australia, cassowary of the Malayan 
 groujD, rhea of America. The edentates, including the sloths 
 and armadilloes of America, the pangolin of Africa and 
 Malaya, and the orycteropus of Africa are confined to this 
 area; Avhile the anthropoid apes range from Eastern India 
 to the west of the Ethiopian region. The tapir is American 
 and Indian, the elephant Ethiopian and Indian, the marsu- 
 pials American and Australian. To this must be added the 
 common characters of the plants of S.America, S. Africa, and 
 Australia, which, though not comparable in importance with 
 the zoological resemblances just stated, contribute to the 
 evidence in favour of a former connection between these 
 different land masses in the southern hemisphere, though the 
 
830 PHYSICAL GEOGRAPHY. 
 
 connection was not simultaneous. Oscillations of land, such 
 as the coral islands show to be still going on in that region, 
 formed at various periods bridges between the different areas, 
 and it is probable that some of these connections date from 
 the triassic times, when S. Africa assumed the continental 
 form which it has since retained. 
 
 408. Analogous or Representative Forms. — A distinction 
 must be observed between these common types and what are 
 known as representative forms. As a general rule, the habits 
 of species under the same genus are closely similar, if not 
 identical, the affinity of structure which constitutes them 
 members of the same group offering a ]3resumption in favour 
 of such resemblance. Thus, the elephants of Asia and Africa 
 are equivalent, so far as their share in the economy of nature 
 is concerned. The llama and its kindred, again, represent the 
 camel, but are not equivalent to it. These relations are con- 
 sequent on the diffusion of the descendants from a common 
 stock, the forms, divergent in their migi'ations as well as in 
 their structure, being dissimilar also in habits according to the 
 character of their new surroundings; but the departure stops 
 short of that utter difference which would be associated with 
 changes so extreme as to efface even the family resemblance. 
 But another kind of resemblance, for v/liich analogy is the 
 best term, exists between members of distinct orders. Thus 
 the marsupials present, within the limits of that group, most 
 of the habits and corresponding modifications of form which 
 occur among the placental mammals. The carnivore has its 
 analogue in the thylacinus, the ruminant in the kangaroo, the " 
 insectivore in the myrmecobius. This remarkable parallelism 
 is due to the long occupation by the marsuj)ials of their iso- 
 lated territory, and to the consequent competition among 
 them, resulting in the adaptive change which has specialized 
 each group by its mode of life. 
 
 409. Migration of Species, and Extension of Area.— 
 Every species tends to spread in all directions as the number 
 of individuals increases ; and if the conditions are similar on 
 all sides the spread will be equal, an event vv^iich can rarely 
 happen. But this extension of area must be distinguished 
 fi'om transfer of the point of maximum development. This 
 takes place — ^.-^ 
 
MIGRATION OF SPECIES. 331 
 
 a. \Vlien climatal changes occur. All species are not 
 equally capable of migration, lience extinction in the primary 
 locality may go along with increase of numbers on the margins 
 of the area, while expulsion in mass will shift the area of the 
 more locomotive species. The glacial period was the climax 
 of a series of migrations whose progress is traced in the 
 changes of the pliocene faunas. 
 
 h. "When geogi'aphical changes occur. But as there is 
 every reason to believe that these are, on the whole, gradual, 
 the changes are more likely to have been by migration, and 
 expansion on the margins of the specific area, than by ex- 
 tinction in one locality. 
 
 c. When the food supply oi a species axters. Whatever 
 affects the vegetation of a district must affect, hoAvever slightly, 
 all the inhabitants of that district. Any disturbance of the 
 balance, say by the failure of one kind of plant, will to that ex- 
 tent alter the diet of the graminivorous mammals, while space 
 will be left for the free development of other kinds, or their 
 importation from elsewhere; and importation of new plants 
 usually brings new animals, whether then: influence be im- 
 portant or not. The annual migrations of animals, commonly 
 referred to instinct, are, in many cases, rather due to appetite. 
 Their occurrence at the time of seasonal change has led them 
 to be referred to climatal causes. But the migTation of the 
 swallow southward at the close of the summer is towards a 
 region where he will find during winter abundant insect 
 food, wanting in that he has quitted. The bison travels after 
 the food which he exhausts in each successive locality by 
 destruction as well as consumption ; and the lijsrds have their 
 return timed by the renewal of the herbage. The quest of 
 water leads to other gi-eat migrations, and these movements, 
 of irregular periods, are identical in kind with those which, 
 from analogy, may be assumed to have some share in this 
 shifting of sj^ecies. 
 
 d. When enemies or competitors appear or disappear. The 
 vegetation of New Zealand seemed that best adapted for the 
 countrj'-, but its development was due to long isolation ; for, 
 where European plants were imported, the white clover is 
 displacing the ferns and native grasses; the cow-gi'ass {Poly' 
 goniim aviculare) follows all the road lines; the dock {liumex 
 
332 PHYSICAL GEOGRAPHY. 
 
 ohtusifoUus and crispus), sow-thistle, and water-cress spread 
 with great rapidity, and attain wonderful luxuriance. Dr. 
 Hooker's correspondent, who gives the above facts, says 
 that the Maori will disappear before the white man, as the 
 Maori's grass, and rat, and fly have been driven away before 
 these companions of the white man. The horse had become ex- 
 tinct in S. America before the arrival of Europeans : it there 
 flourishes now in such fashion as to make its extinction 
 unintelligible, unless light be thrown on it by the fact that 
 it cannot gain a footing in Paraguay, an insect in that region 
 attacking the navel of new-born foals. 
 
 e. Excess of numbers within an area is sometimes spoken 
 of as equal in importance with those already mentioned. But 
 this opinion seems to rest on an overstrained analogy of what 
 happens in the case of man and some other social animals. 
 Where there is organization of labour, where each individual 
 has his place assigned, it may be possible for the excessive 
 population to be received for a time into what may be called 
 the social interstices. But after a time the surplus must 
 remove, or the whole community will suffer. In the bee-hive 
 there are no interstices, and the surplus is added suddenly 
 and in mass; the swarm must be cast at once. In highly 
 civilised human communities the surplus is constantly added, 
 and as constantly drained off, while the swarming process 
 is only a feature of a very imperfect state of civilisation, 
 when tribes are isolated, and when, as in Asia at the present 
 day, the traveller cannot journey alone, but must wait for a 
 caravan. Among animals, say the carnivores, each individual 
 requires a certain area for the satisfaction of his wants; and if 
 he is restricted in this enjoyment he must fight his encroach- 
 ing neighbour, and the weaker must either go or die. The 
 slow expansion of this specific area is therefore indistinguish- 
 able in kind from the sudden migration in mass. Even the 
 lemming wanders forth, not as an emigrant in the human 
 sense, but as a migrant like the swallow, for the lemming 
 swarm precedes a severe winter, that is diminished food 
 supply. 
 
 410. Results of Migration.— If circumstances are favour- 
 able, the wandering may mean only extension of unaltered 
 forms over a wider area. By unfavourable conditions the 
 
SELECTION OF SPECIES. 333 
 
 wanderers may fail to obtain a footing, perishing by climate, 
 by want of food, or the opposition of enemies. But the 
 extinction may not be rapid, and in the struggle to retain 
 their place, the species may become modified in various ways. 
 Modifications also may take place even where a species easily 
 holds its ground, for there are few areas identical in their 
 conditions throughout. 
 
 411. Analogy of Distribution of Species in Space and 
 Time. — Observation has shoAvn that, in the majority of cases, 
 species «.rise in areas which coincide both in space and time 
 with those of other closely allied species, and that, as has 
 been already said, varieties are proj)ortioned in number to 
 local difierences of conditions. It is further known that 
 man can cultivate varieties, that is, when a desirable varia- 
 tion of form or colour appears in plant or animal under 
 domestication, man can, by careful selection of the offspring, 
 perpetuate the variety. If the efibrt is suspended, the 
 descendants gi'adually revert to the original form, but never 
 actiially resume identity therewith. It is further known by 
 observation, that species which have once disappeared do not 
 recur, their place being taken by others; that members of 
 the same genus differ the more the further the strata in 
 which they occur are separated from each other in time ; 
 and that, finally, if at the present we travel from the nor- 
 thern to the southern temperate zone, Vv'e gradually diminish 
 the number of identical forms, the species changing as we 
 advance, till at the antipodes a very small number of identi- 
 cal forms would be met with. 
 
 412. Natural and Artificial Selection : Survival of the 
 Fittest. — By these two phrases is meant that the organic 
 world is never in a state of equilibrium, but that as there is 
 constant fluctuation in the physical surroundings of organised 
 beings, there is a corresponding struggle among these beings 
 themselves to maintain their existence. The one phrase calls 
 attention to the result of the struggle, the other to the 
 machinery by which it is carried on. Whether the changes 
 which result in permanent species are due to plasticity of the 
 animal or vegetable organism, whereby external influences 
 lead directly to changes of structure, or these changes are 
 gradually accumulated after having come into existence 
 
334 PHYSICAL GEOGRAPHY. 
 
 under some influences whose operation we know nothing of, 
 it is immaterial here to discuss. Suffice it that these changes 
 will be perpetuated if they are beneficial to the animal; that 
 just as particular varieties are " selected" artificially by man, 
 so in nature a stronger or better endowed animal is selected, 
 and survives and mnltiplies, its weaker or worse endowed 
 associates disappearing. Continuing to use the impersonal 
 language hitherto employed, the illustration taken by Mr. 
 Huxley is a happy one, when he says that the wind selects 
 the successively finer sand on the dunes, and that frost selects 
 the weak plants as the gardener would select them. 
 
 413. Variations : How Beneficial. — The modifications 
 which constitute " the fittest/' may be either of an obviously 
 useful kind, as the length of neck which extends the girafie's 
 browsing ground, the length of limb which increases the 
 swiftness of a carnivore, or which gives its prey better chance 
 of escape. In artificial selection that may be desirable which 
 to the wild animal would be fatal, as the shortness of limb 
 which gave to the Ancon sheep of Massachusetts their value, 
 because they could not leap fences. But the benefit may be 
 indirect, as when a beetle (Psox) has the colour and appear- 
 ance of dust, and thus escapes detection; when a leaf — or a 
 stick — insect is indistinguishable from the vegetable matter 
 amid which it lives; when one butterfly comes to resemble 
 a butterfly of another species, so as to acquire the immunity 
 from the pursuit of birds enjoyed by that one which it 
 resembles; or when a moth and a humming bird have a 
 suj)erficial likeness protective to both. 
 
 414. Mimicry: Protective Resemblances: Independent 
 Resemblances. — All these modifications are in reality uncon- 
 scious on the part of the animals, though the poverty of 
 language renders it difficult to avoid the suggestion of voli- 
 tion in the last -mentioned cases. But the occurrence of 
 resemblances which cannot be protective, demonstrates the 
 unconscious character of the change, and points to similarity 
 of endowment of living bodies as the source of the accidental 
 similarity which is seen, for example, in the curious likeness 
 of marine shells in one part of the world, to land shells in 
 another. Homogenetic has been proposed for resemblances 
 which luay be traced to common ancestry, homoplastic for 
 
HOMOTAxia. 335 
 
 cliaiiges proclucecl by similar influences on similarly enclovred 
 tissues. 
 
 415. Representative Species. — In distant regions identical 
 forms are met with, tlie more frequently the lower are the 
 organisms, wide distribution in space and time being propor- 
 tioned to the simplicity of the forms. But among the higher 
 plants and animals, the independent origin in both localities 
 of the identical forms is rendered improbable when it is 
 found that, in the one case, the individuals are so numerous 
 as to indicate that tliis is their proper habitat, while, in the 
 other, their paucity indicates that they are outliers, and 
 sometimes the intervening area may yield fossils proving the 
 former continuity of the sj)ecies. Moreover, on the assump- 
 tion that all species reach their present localities by migra- 
 tion, the imj^robability of identical variations occurring far 
 apart is very gi^eat, since the influences camiot be absolutely 
 identical, and perfect reversion is unknown. But closely 
 allied forms may arise, and these are truly representative 
 species, since they set forth the changes produced by similar 
 conditions in the descendants from a common stock. 
 
 416. Dangers Incident to Migration. — The diminish- 
 ing number of the travellers might be expected, when it is 
 remembered that every change of locality may yield change 
 of temperature, of food, of rivals, and of enemies. The 
 prospect, therefore, of great extension of a species, and of 
 its numerical superiority, is a complicated problem, to which 
 must be added another factor, the number of individuals 
 which may have started on the route. 
 
 The names in the following table must be regarded as the 
 ends of branches which are given off successively from pre- 
 vious branches. It is not necessary in this volume to state 
 the points of divergence. It is only desired to suggest that 
 the succession Is no more linear than is that of the branches 
 of a tree, the highest branches of which diverge from the axis 
 equally with the lowest. 
 
 417. Homotaxis. — It is often said that, in the silurian 
 times, there was a greater uniformity of life over the earth 
 
 \ than there is now, the deposits of that epoch, wherever 
 examined, revealing closely similar forms. But all stratii 
 fire assumed to be silurian which are the earliest fossiliferous 
 
:36 
 
 PHYSICAL GEOGRAPHY. 
 
 
 ?3 
 
 
 ^ 
 
 
 
 o 
 
 
 To 
 
 
 m 
 
 $L( 
 
 
 xi 
 
 
 
 xn 
 O 
 
 03 
 
 ^ 
 
 
 u 
 
 
 o 
 
 
 ■ • 
 
 P4 
 
 g 
 
 
 
 t 
 
 M 
 
 >5 
 
 1— ( 
 
 
 O 
 
 o 
 
 >> 
 
 o 
 o 
 
 
 1 
 
 
 TO 
 
 0) 
 
 rS 
 
 p^ 
 
 f>» 
 
 
 o 
 
 
 O 
 
 
 •^ 
 
 
 o 
 
 
 g 
 
 w 
 
 o 
 
 0) 
 
 1^ 
 
 m 
 m 
 
 
 O 
 
 
 fcH 
 
 
 r^ 
 
 O 
 
 ^ o so 
 
 t> 
 P 
 
 ^ I « 
 
 ^ . . g !^ 
 
 <i 
 o 
 
 I— » 
 
 PQ g 
 
 jf «3 
 
 a> j-J C« -rf 
 
 ;; CS ™ ja Q) 
 
 
 o 
 
 
 
 
 
 p: 
 
 O 
 
 
 ^ 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 
 • iH 
 
 
 
 
 
 
 
 
 rd 
 
 
 
 
 
 
 
 
 O 
 
 
 
 
 
 
 
 
 t 
 
 
 s 
 
 
 
 
 
 . 
 
 ^ 
 
 
 r§ 
 
 03 
 
 
 
 
 e3 
 
 
 
 'o 
 
 •+-> 
 
 
 
 
 O 
 
 
 • 
 
 M 
 
 c3 
 
 
 
 
 
 
 o 
 
 < 
 
 o 
 
 
 
 
 1^ 
 
 
 N 
 
 
 -4J 
 
 
 
 
 '^ 
 
 
 ^ 
 
 
 rt 
 
 
 
 
 S 
 
 1 
 .1 
 
 75 
 
 
 O 
 
 »2 
 O 
 
 to 
 
 o 
 
 C/2 
 
 o 
 
 8 
 
 
 
 o 
 o 
 
 
 c3 
 
 
 _c3 
 
 1 
 
 t-3 
 << 
 
INSULAR FAUNAS AND PLOI^AS. 337 
 
 sti^ta of each region, an arbitrary basis is assumed, and tlie 
 resemblance of these rocks in different regions is exaggerated, 
 it being forgotten that as these strata everywhere rest on 
 metamorphic rocks (where their lowest portions are visible), 
 it is impossible to say to what extent the records of pre- 
 existent zoological provinces have been destroyed. In later 
 deposits, as the marine portions of the carboniferous series, 
 community of fossils in distant localities has been taken to 
 indicate contemporaneity of deposits, and the existence of 
 identical species at the same time in these localities. Eeason- 
 ing from the small number of identical species at distant 
 localities at present, and from oiu' imperfect knowledge of 
 the strata whose organic contents are said to be identical, 
 the opinion fii'st announced by Godwin Austen seems correct, 
 that identity of species indicates lajDse of time sufficient to 
 have admitted of migration from one locality to another. 
 Professor ^uxley proposed the term homotaxis as expressive 
 of the relation of such fossil groups, insisting that strata, 
 characterised by similar remains, show a similar but not 
 identical sequence in each locality; and this doctrine, that 
 geological formations represent geographical 2:)rovinces rather 
 than periods of time, is in accordance with Professor Ham- 
 say's investigations into the distribution of former continental 
 areas. 
 
 418. Insular Faunas and Floras. — In an earlier chapter 
 the population of islands was said to differ from that of ad- 
 jacent lands in proportion to the antiquity of its isolation. 
 The Atlantic islands illustrate this generalisation. The 
 Madeira, Canary, and Cape Yerde groups differ from the 
 mainland and from each other. There are no indigenous 
 mammals except bats; but such quadrupeds as have been 
 introduced prosper remarkably. Madeira has, on the other 
 hand, ninety-nine species of bii'ds, of which one is peculiar, 
 two are common to it and the Canaries, the rest are Euro- 
 pean ; and it is noticeable that the number of these is greater 
 on the eastern than on the western islands, the wind-driven 
 birds first alighting on the eastern. Every island seems to 
 have its own peculiar insect forms, two-thirds of the Madeira 
 beetles being as yet unknown elsewhere. The flora of these 
 islands has the same American aspect as the miocene flora 
 23 y 
 
338 PHYSICAL GEOGRAPHY. 
 
 of central Europe, and it is probable tbat this resemblance 
 is due to the transport, from adjacent miocene land to these 
 islands, of seeds which might be carried by birds or drift 
 wood, the transport being perhaps facilitated by intermediate 
 islands now submerged. That the isolation of these islands 
 is anterior to the glacial period follows from Hooker's obser- 
 vation, that whereas in all continental mountains boreal 
 forms are found, the fragments of an expelled flora, the 
 main body of which has returned northwards, the mountain 
 flora of these islands is simply that of the low grounds. The 
 antiquity of these islands is further shown by the distinct- 
 ness of the pliocene fossil shells in the adjacent islands of 
 Madeira and Porto Santo, and by their equal distinctness at 
 the present time. The Galapagos Islands have a fauna of 
 land birds, all belonging to S. American types, yet four- 
 fifths are peculiar, whereas an equal proportion of the aquatic 
 birds are of S. American species. In the Atlajitic islands 
 the birds are tolerably equally diffused, probably by aid of 
 winds ; but the calms of the Galapagos area explain the re- 
 striction of some of the birds to separate islands. These 
 facts are explica,ble on the assumption that the isolation of 
 these islands is of great antiquity, and that the presence of 
 forms belonging to adjacent islands is in proportion to their 
 means of transit, as in the case of birds and bats, or to the 
 likelihood of their being drifted by help of ice, floating wood, 
 or in the mud adhering to the feet of birds ; that these im- 
 portations are not of frequent occurrence, and that the forms 
 thus isolated have varied so as to lose their identity with the 
 I'tarent stocks, as has happened in the case of Australia. 
 The student will find in Mr. Wallace's book on the Malay 
 Archipelago, an admirable statement of the method of inves- 
 tigating the differences of life in adjacent areas, and of the 
 geographical evidence which may be elicited from them. 
 
 419. Hypothesis of Lost Continents. — As our knowledge 
 of the various means by which plants and animals may be 
 transported increases, it is less necessary to assume the former 
 evidence of hypothetical land connections. A miocene At- 
 lantis, across which American plants migrated to Europe, is 
 especially unnecessary, since the existence of miocene plants 
 in the far north suggests their passage in that region, and 
 
MARINE PROVINCES. 330 
 
 likewise explains the common characters of the vegetation of 
 the then northern continents; and in the second place the mio- 
 cene corals, common to Europe and the West Indies, required 
 open sea in the Atlantic, or at least a more or less scattered 
 archipelago, a supposition quite in accordance with the facts 
 above stated. The absence from Madagascar of widely dis- 
 tributed African mammals has suggested the former exist- 
 ence of a continent in the Indian Ocean; but the presence of 
 lemurs on the African shore, seems to indicate a former land 
 connection with Africa at a period so remote that time has 
 sufficed for the variation of the isolated species, so that 
 nearly all are now peculiar. Moreover, the presence of a 
 lemur in Europe in miocene times renders more likely the 
 existence of dry land where the Mozambique Channel now 
 is. The land connection with the Pacific area is supported 
 by the affinity of the insectivore centetes to the American 
 genus solenodon, and by the presence in Madagascar of 
 American types of serpents and insects. 
 
 420. Marine Provinces. — Continuous as the sea appears, 
 it is mapped out into provinces as distinct, though more ex- 
 tensive, than those of the land. The fishes have free powers 
 of locomotion, but they are as much addicted to particular 
 localities as the birds, while it is less easy to recognise the 
 conditions on which their limitation depends. The species 
 associated in the bathymetrical zones, already mentioned, are 
 as local as the inhabitants of subaeiial plains, and after the 
 limit of influence of surface temperatures is passed, the species 
 become more extended in range, just as the mountains, even 
 to the equator, support boreal forms. The littoral species 
 are the most varied, they being most under the influence of 
 seasonal vicissitudes, and diversity of soil and food. 
 
 The provinces of terrestrial life have their analogues in the 
 
 421. N. Atlantic. — The North Atlantic province includes 
 a large part of the Ai'ctic shores of the Old and New Worlds, 
 and extends far down the American coast on the one side, 
 the European shores on the other : the Mediterranean and 
 Aralo-Caspian lakes belong naturally to it, and its southern 
 limit is the equatorial drift. Over all this area the living 
 and the most recent tertiary species of mollusca agree, and 
 
34:0 PHYSICAL GEOGRAPHY. 
 
 tlie cliaracters whicli unite so large a tract under one desig- 
 nation are due to the dispersive power of the glacial cold, 
 whereby boreal forms are now located in the Mediterranean, 
 and the aifinities of that sea with the Indian Ocean have 
 been greatly reduced as compared with what they were in 
 pliocene times. The adherence of cod, salmon, and herring, 
 to the coast lines is as marked as that of the molluscs, and 
 confirms the view that the community of species on either 
 side, amounting to nearly one hundred among the molluscs 
 alone, is due to the existence of a former land connection in 
 the far north. 
 
 422. Caribbean Province. — This area extends from Concep- 
 cion on the west coast of S. America, to the outflow of the 
 Gulf Stream. Its molluscs, by which chiefly the province 
 is determined, merge into each other, the Falkland Isles 
 (Malvinas) proving their antiquity by the possession of 73 
 per cent, of peculiar species. The fauna of the Mexican 
 Gulf includes a large number of corals, but its molluscs have 
 about 3 per cent, of species common to the western side of 
 Darien, and some of the common species are found also in 
 Western Africa. The transport of these species by the 
 equatorial drift, and their migration westward during the 
 patency of a channel through the Isthmus of Darien, is an 
 event of the same kind, probably, as that by which a few 
 species have been carried by the Gulf Stream to the shores 
 of Britain. But when it is remembered that of 522 Sicilian 
 species, thirty-five are common to the West Indies, and 
 twenty-eight to the west coast of Africa, it is possible that 
 the extension of species from Panama to Africa may have 
 occurred at the same time, and by the same means, that 
 Central Europe and the West Indies shared their coral 
 fauna. 
 
 423. Indo-Pacific Province. — This, the largest marine 
 area, is still imperfectly known. Following the coast line, 
 a remarkable continuity of molluscan forms can be traced, 
 and such agreement as the Pacific Islands present with the 
 American continent, ceases at the Peninsula of California. 
 The sea lions range throughout this great area, and even 
 beyond it, manifesting as little agreement in detail with the 
 regions marked out bv terrestrial life as do the other ani- 
 
DEEP SEA FAUNAS. 341 
 
 mals possessed of various powers of locomotion, with the 
 areas whose limits have been marked out by Mr. Wallace. 
 The area is predominantly that of corals, and the specific 
 variety of the mollusca is due to the multiplicity of local con- 
 ditions arising from the submergence and re-emergence of 
 lands whose prominent peaks have been seized on by the 
 corals (see page 66 for the axes of Pacific lands). In this 
 region, too, are foimd the remains of the decadent group of 
 cephalopods, which filled the mesozoic strata with ammonites, 
 but is now rediiced to the nautilus. 
 
 424. Australian Province. — The large number of peculiar 
 genera, and of genera which here attain their maximum, 
 gives as marked a character to Australia as it owes to its 
 terrestrial fauna, more especially as mesozoic types are con- 
 spicuous. But the common genera are also of interest, since 
 they are, with unimportant exceptions, found in the southern 
 land masses of America and Africa, as Avell as on the shores 
 of the Antarctic continent. 
 
 425. Western S. America. — The Gulf of California 
 belongs to the great province of which Panama is the centre, 
 and whose southern limit is the margin of the easterly or 
 Antarctic drift. The community of forms on either side of 
 Darien has already been mentioned; it only remains to say 
 that the total difference of the molluscan fauna of the Indo- 
 Pacific from that of this portion of America, renders the 
 region thus limited a very natural one. . The Galapagos 
 Islands maintain in their marine fauna the preponderance 
 of peculiar forms seen in their terrestrial population. 
 
 426. Pelagic Foims. — The conditions of migration of 
 some families of animals are unknown. Cuttle fishes are 
 found in most regions, but the species, though dwellers in 
 the high seas for the most part, are very conservative of 
 their customary tracks. If it is difiicult to follow the move- 
 ments of an animal like the herring Avlien it approaches the 
 coast, it is impossible to say how the pelagic species travel, 
 and our marine provinces are therefore defined entirely by 
 reference to the coast faunas. 
 
 427. Deep Sea Faunas: Continuity of the Cretaceous 
 Epoch. — Only in the Atlantic Basin have soundings as yet 
 given systematic information regarding the inhabitants of a 
 
342 PHYSICAL GEOGRAPnr. 
 
 depth, beyond 200 fathoms, and even there, great as has been 
 the addition to our knowledge, the observations are few and 
 isolated. It appears that the oaze, often referred to, repre- 
 sents the chalk of the cretaceous formation, its large amount 
 of diffased silica not having yet been segTegated in the flint 
 nodules of the old rock; that the area occupied by the deep 
 water species is wider than that of the shallow water forms; 
 and that while the propriety of including within one area 
 the regions mentioned in a previous paragraph is confirmed 
 by the presence of miocene and pliocene Sicilian forms, in 
 others a remarkable affinity is shown to the cretaceous forms. 
 The continuance of these forms, belonging to types supposed 
 to be extinct, in association with a deposit so peculiar, one 
 which from theoretical considerations we should expect to 
 fi.nd, as it is known to be a deep water accumulation, has 
 given grounds for the generalisation that the chalk is con- 
 tinued to the present time. Whatever exception may be taken 
 to the phrases in which this idea is embodied, the fact is, 
 that the cretaceous deposits persist with very much of their 
 characteristic fauna, and that thus two provinces co-exist, 
 which, judged by the pal^eontological standard, would be 
 regarded as belonging to distinct periods of time. It further 
 appears, as Dr. Wyville Thomson points out, that, "the 
 gasteropods range from the shore to a depth of 100 to 
 200 fathoms, the lamellibranch molluscs become scarce at a 
 slightly greater depth, while some orders of brachiopods, 
 Crustacea, echinoderms, sponges, and foraminifera descend in 
 scarcely diminished numbers to 10,000 feet. In fact, the 
 hatliymetrical range of various groiqys in modern seas corre- 
 sponds remarkably loith their vertical range in ancient strata." 
 The sentence which we have italicised contains a generalisa- 
 tion which accords with that of Agassiz as to the order of 
 the genera of corals in a reef, the representatives of the 
 groups whose maximum was in the remote past being near 
 the bottom, the modern species towards the surface of the reef. 
 428. Extinction and Replacement of Species. — The dis- 
 appearance of species is very slow, and the final extinction 
 can very rarely be determined. ' The lists of fossils given in 
 works on geology show that a species may not occur in every 
 successive series of strata in one region; but its absence in 
 
PEESISTENT TYPES OF SPECIIIS. 343 
 
 any particular layer may be due to its migi'ation elsewMtlier, 
 or to our imperfect knowledge of the layer, as well as to its 
 extinction; and one of the former is clearly the explanation 
 when it occurs in the layer above. The colonies of M. 
 Barrande show how extensive migrations may be, and how 
 long an interval may separate the earlier and later occupancy 
 of a region by the same species. As there is no reason to 
 believe in the sudden and entire destruction of a species, which, 
 so far as evidence goes, could only take place with a decay- 
 ing species, whose numbers and area were therefore already 
 gTeatly reduced, the sudden sharp cessation of, say, the meso- 
 zoic types in Britain at the close of the chalk, is due to 
 migration and to the nonpreservation of the highest strata of 
 the chalk series. For the same reason the replacement of 
 species is a process not traceable, even if it were effected only 
 by migration; still less can the steps be followed when modi- 
 fication is instrumental in the changed aspect, 
 
 429. Persistent Types; Progressive Development.— It 
 is impossible to enumerate the influences by which plants and 
 animals are altered, destroyed, or displaced ; but to the 
 external agents must be added one whose power we cannot 
 estimate, namely, inherent tendency to decay. If there is 
 in species a natural period of existence, as there is in indi- 
 viduals a limit to functional activity, it is not probable that 
 the full period of duration is often attained, since species 
 are subjected to ever-varying external influences. It is not 
 necessary to argue against theories of progressive development, 
 as that term was understood some years ago. It is now well 
 known that the earliest fossils with Avhich we are acquainted 
 belong to types as well developed as those of the present day; 
 and that while new orders have made theii' appearance, they 
 have done so, for the most part, at remote dates. Vertebrates, 
 fishes, are found in the Silurian series, molluscs of various 
 groups, the lamellibranchs, gasteropods, and cephalopods, are 
 all present in the same strata : brachyurous and macrurous 
 crustaceans (of which the crab and lobster are the familiar 
 living examples) appear in the carboniferous rocks, along 
 with sharks and highly-organised labyi'inthodont amphibians. 
 In the triassic strata all the orders of reptiles, except serpents, 
 are represented. Birds are found early in the Jurassic series; 
 
344 PHYSICAL GEOGRAPHY. 
 
 and the lower tertiaries contain nearly, if not all, the orders 
 of mammalia. But these higher types not only appear 
 very early, they also persist for long periods alongside of the 
 lower types. Thus trilobites are still found in the carboni- 
 ferous rocks with the decapod crustaceans mentioned; the 
 labyrinthodonts still flourished in the triassic times, along 
 with the reptiles; the marsupials, which still exist, commenced 
 their career early in the mesozoic series. Among the inverte- 
 brates similar cases are found. The classes, crustaceans, 
 insects, myiiapods, and arachnids, are all found in the carbon- 
 iferous, and of the orders of insects several have appeared in 
 strata lower than the cretaceous, deposits; and in the shores of 
 the chalk sea air-breathing gasteropods of the genera Physa 
 and Paludina, lived in lakes and lagoons, over which dragon- 
 flies hovered, while beetles (Bu2)restes) crept through the 
 decaying plants on the margin. The orders, then the types 
 of structure, are of great antiquity, and those which have 
 disappeared have given place to more specialised forms, in 
 which, that is to say, the adaptation to particular conditions 
 is more complete. But genera are also very persistent, many 
 dating from the early Silurians, as lingula, rhynchonella ; 
 and others appear at successive periods, lasting, with little 
 change, to the present time. On the theory of the origin of 
 species by descent with modification, the early appearance of 
 so many highly organised forms is exjolicable by their origin 
 in pre-existent common ancestors, not by their evolution each 
 out of the other. But it is clear that, in the first place, 
 this requires the ea-rliest plants and animals to have been in 
 existence long before the cambrian, perhaps even before the 
 lanrentian; and thus we should be without fossil evidence of 
 a population as important, and, for aught we know, numeri- 
 cally as great, as that of the present time. In the second 
 place, there is still greater proof, in the fact stated, of the 
 existence of zoological provinces distinct from each other even 
 at the earliest times. The provinces, as they existed in 
 the later palaeozoic and earlier mesozoic periods, have been 
 sketched out, but the consideration of them, of their differ- 
 ences from those already given, belongs more properly to 
 descriptive geology, in connection with which the relations of 
 this triassic continent will be discussed. 
 
CHAPTEE X. 
 
 HISTORY AND DISTRIBUTION OF MAN. 
 
 Diffusion of Man — Unity or Plurality of Mankind — Are Races Species? 
 — Bases of Classification — Language — Form of Skull — Hair: 
 Woolly and Straight Haired Races : ISIelanoi, Xanthomelanoi, 
 Xanthochroi, Melanochroi, Negroes, Negritos — Antiquity of 
 Man — Evidence of Antiquity — Prehistoric Periods — Stone Period 
 — Metallic Age — Primitive Home of Man — Hypothesis of Lost 
 Continents. 
 
 430. Diffusion of Man. — The diffusion of man over the 
 surface of the globe, and the permanent footing he has secured 
 in all regions — save those polar and tropical localities in which 
 cold on the one hand, drought on the other, forbid the pre- 
 sence of animals and vegetables in sufficient quantities, or for 
 long enough periods, to supply his wants — are sometimes 
 referred to as if they were characteristics of man alone, and 
 implied an inherent power of adaptation to various external 
 conditions. But his history, and the analogy of other animals, 
 justify the opinion that this power of endurance and of 
 adaptation is an acquisition requiring long time for its full 
 development; and even now there are certain limits beyond 
 which his skill and energy cannot carry him in the struggle 
 with nature. The Bengali and the Esquimaux cannot occupy 
 in permanence each the other's country; and though the 
 native of temperate regions has greater tenacity than he who 
 is subjected to extreme conditions of temperature, there is no 
 instance of the permanence of a race from temperate regions 
 in tropical countries, in which intermarriage with natives 
 and a steady stream of immigrants have not shared in bring- 
 ing about the result ; while the change which the descendants 
 of the British colonists in North America and Australia have 
 undergone, shows that the permanent occupation of a country 
 does not imply the maintenance of the original character of 
 the people. Mr. Bates has given strong reasons for believing 
 in the comparatively recent occupation of South America by 
 
346 PHYSICAL GEOGRAPHY. 
 
 tlie tribes wMcIi we speak of as native; but the facility with 
 •wbich the Negro has become acclimatised in the same area 
 shows that the pov/er of adaptation does not depend on the 
 relative elevation of a race, the Negro and the South American 
 being nearly on the same level. The difficulty of determin- 
 ing the conditions of the problem is increased by the fact that, 
 while certain anatomical differences can be recognised between 
 the races in difierent regions, none of them account for the 
 physiological dissimilarities which we know to exist. 
 
 431. Unity or Plurality of Mankind. — Two opposite 
 opinions are held regarding the origin of man. According to 
 the one, the varieties which now people the world are the 
 descendants of a single pair; the other assigns the varieties 
 to a plurality of ancestors. The latter view may be set aside 
 as not resting on sufficient evidence. The monogenists, as 
 the advocates of the former view are called, believe that the 
 ancestors of mankind were specially created in one locality, 
 and were most probably a single pair, or that they were 
 developed by evolution out of lower forms, and that they 
 were a pair or several, but in either case appeared in one 
 locality. It is erroneous and unjust to assume, as is too 
 frequently done, thsd, this last view excludes the influence of 
 that higher power which is more conspicuously appealed to 
 by the hypothesis of special creation. The hypothesis of 
 evolution has been adopted in the case of plants, and animals 
 other than man; and it is here again adopted because it affi)rds 
 adequate explanation of a larger number of phenomena which 
 are related to each other as antecedent and consequent than 
 does any other hypothesis. 
 
 432. Are Races Species ? Bases of Classification. — This 
 is not the place to enter into the discussion of the question, 
 Are races of man varieties or species ? The grounds on which 
 they have been regarded as sub-species, that is, as presenting 
 variations more constant than is usual with varieties, but 
 v/hich cannot be certainly regarded as of sj^ecific value, are 
 chiefly: (1) that the variations are not universal in each race, 
 intermediate forms being found which might be referred to 
 other races ; (2) that there is no certainty that any races are 
 mutually infertile; (3) intellectually, all the races a.re so equally 
 endowed as to establish the probability of their identity. 
 
LANGUAGE. 347 
 
 The characters on which the varieties are established are, 
 the form of the skull, the complexion, and the quality of the 
 hair, points which, though not themselves of absolute value, 
 nevertheless characterise very well the great groups. 
 
 433. Language. — There is no reason to believe that lan- 
 guage differed as to its origin from man's other characteristics, 
 or that its development was other than gradual. But this 
 analogy with physical j)henomena leads us to anticipate that 
 philology should be an uncertain guide to the original rela- 
 tionships of races. 
 
 At the present time, apart from the facility with which somo 
 tribes, who have no written language, acquii'e the language 
 of their neighbours or invaders ; apart from the evidence we 
 possess that a national speech may be washed over by others, 
 so that language alone Vv'-ould fail to reveal the original rela- 
 tions ; setting aside also the fact, important though it is, 
 that the ultimately dominant speech is not always that of 
 the dominant race ; that, in fact, whereas the spread of 
 physical characters in the animal and vegetable Avorld is in 
 the ratio of the numbers of indi^'iduals who present them, 
 speech is extended sometimes by the weaker in numbers and 
 tenacity — setting all these considerations aside, we find that 
 among uncivilized tribes superstition plays an important part 
 in modifying the language, the taboo of the chiefs, and. that 
 consequent on death or misfortune, leading to the abolition 
 of words which are replaced by arbitrary symbols, at least 
 this is the case in Africa. The roots, therefore, of the earlier 
 speech are not certainly transmitted ; and their couiparative 
 fixity at later periods is chiefly helpful in tracing migrations. 
 In other words, the physical difierences between the Aryan 
 and Semitic races are less than the linguistic, the distinction 
 having been perhaps exaggerated by some such influence as 
 that just mentioned. It will appear, from the subsequent 
 paragraphs, that the general statement regarding the westerly 
 tendency of nations is true only for the latest disj^ersion of 
 the Xanthochroic races : that a jiortion of that stock migrated 
 towards the south-east, and that the diffusion of that great 
 section of mankind has had its movement largely determined 
 by circumstances. Thus the westward movement from Asia 
 was a necessity, smce the east had been already occupied by 
 
348 PHYSICAL GEOGRAPHY. 
 
 the descendants of the people who first tenanted that southern 
 continent, of which fragments only survive. Alternate east- 
 ward and westward movements are historically known among 
 the Mongols. The Malay has spread to south-east and 
 south-west. And at the present day the European nations 
 are throwing off their surplus population in all directions, 
 but chiefly where the indigenous people are not strong 
 enough to hold their own, where the climate is not bad 
 enough to kill the immigrant within a few years, and the 
 soil or native productions are not poor -enough to forbid the 
 hope of great gain with moderate labour. 
 
 But if the material of language does not give positive evi- 
 dence as to the ancestry of nations, its structure has been 
 found to indicate their grade of development. Languages 
 were grouped by Humboldt as isolating, agglutinating, and 
 inflectional; the simple ajDposition of the roots, as in the old 
 Chinese, being the lowest of a series which culminates in the 
 highly complicated grammar of European tongues, in which 
 prefixes and suffixes are now merely symbols of particular 
 kinds of relation, are merely the faint shadows of words 
 which had primarily an independent existence. These words 
 have undergone two kinds of change, a logical and a physi- 
 cal. The latter, called by Miiller phonetic decay, is a con- 
 venient index of the former, or at least of the extent to 
 which it has gone. It must be rememl3ered that what he 
 denominates decay, ought on the hypothesis of evolution to 
 be called progressive adaptation; for that hypothesis includes 
 language among those phenomena which are of slow develop- 
 ment and of changing type, while Miiller starts with phonetic 
 types or roots which he thinks exist by nature, change of 
 which is a departure from the standard, and in his view decay. 
 But without accepting the interpretation, Miiller's summary 
 of the facts is a useful one, and is here subjoined in tabular 
 form, Humboldt's classification being placed alongside of it. 
 
 Midler. Humholdt. 
 
 1. ER Chinese ^ = Isolating. 
 
 2. R + g ) rt g^ g + E ) J.i g + R+ g ) -^ S^ ( = Agglutinative. 
 
 > >>;« r s ^ \ 's '^ { 
 
 S. rg X^'Sgr \j^r grg \ ^% I - Inflectional. 
 [4. II + g + r (?) Bask and American, polysyuthetic. = Incapsulating]. 
 
FORM OP SKULL. 349 
 
 In this table, E, = symbol of a root wliicb bas suffered no 
 phonetic change; r = symbol of a root which has lost its 
 primary significance, but without having undergone phonetic 
 change; while g represents roots which have undergone both 
 processes, and in respect of their loss of meaning relative to 
 the objects vfhich they primarly indicated, may be called 
 dead or empty, though as the exponents of new relations (for 
 they multiply as the life of nations becomes complex), they 
 have acquired new vitality. The student will find the ample 
 discussion of these topics, and of the origin of language, in 
 Max Miiller's writings, especially the Stratification of Lan- 
 guages; in Farrar's Essa]/s; and in the works of Sir John 
 Lubbock and Mr. E. B. Tyler, on Primitive Civilization and 
 Primitive Culture of Man. 
 
 434. Form of Skull. — The terms used in the subjoined 
 table, as descrijDtive of the skull forms, have reference to the 
 proportions of length to breadth; the length being taken as 
 100; thus an extreme transverse diameter of 7 inches, with 
 a length of 9 inches, would give the proportion, 9:7 or 
 
 100: 77-7 = index •77.'"^ 
 
 I. Br.ACHYCEPHALIC. 
 
 Round Skulls. Cephahc Index '80 or mora. 
 
 Brachistocephahc. Index at or above '85 
 Eury cephalic. ..> ., "85 to "80 
 
 II. Dolichocephalic. 
 
 Long Skulls. Cephalic Index below '80 
 
 a. Oval Skulls. 
 Sub-brachycephalic. Index "80 — "77 
 
 Orthocephahc. ,, -77 — "74 
 
 Mecocephahc. ,, '74 — '71 
 
 b. Oblong Skulls. 
 Mecistocephalic. Index '71 and less. 
 
 This mensuration does not give information regarding 
 the facial angle, or inclination of the facial profile to a hori- 
 zontal plane drawn from the anterior end of the cranial axis, 
 which is 67^ in the young Orang, 70^ in the Negro, 85° in 
 the European. Neither does it tell of that relation of the 
 incisors which furnished Retzius with the subordinate char- 
 * Prehistoric Remains in Caithness. Laing and Huxley, p. 85, 
 
350 PHYSICAL GEOGRAPHY.' 
 
 acters, orLliognatlious wlien the canines are vertical, pro- 
 gnathous when they are inclined. These features accompany 
 any proportion of skull, the dolichocephalic Negro, and the 
 brachycephalic Mongol being prognathous, the dolichocephalic 
 Aryans and the brachycephalic Peruvian, orthognathous. 
 
 435. Hair: the Distinction between the Woolly and 
 Straight Haired Races. — TJlotrichous and Leiotrichous, 
 ti'ivial as the distintion may appear to be, correspond to 
 important differences in character, capacity, and geographical 
 distribution. The following table shows the results of the 
 application of the tests, skull form, and hair character. 
 
 DolkliocepTialk, Brachycephalic. 
 
 s 
 
 § ( Wooly haired, (Negroes; Bushmen.-. 
 
 g ( Crisp haired, ( Negritos ; Andaman Islanders. 
 
 U 
 
 ( Melanoi \ -^^^tralians ♦ 
 
 ' "' * ' { Dravidians ; Ancient Egyptians. ? 
 
 Xanthoclu-oi, j f^tli^I^ll^f/gt^rans. Scandinavian. 
 
 ■» yr 1 T . \ Silures : Iberians : 
 
 Melanocliroi, | park Aryans. 
 
 436. Melanous Races. — The aboriginal Australians are 
 among the lowest, both phj^sically and mentally. Their 
 prognathous skulls are small and long, the prominent super- 
 ciliary ridges are solid. Their skin is dark chocolate- 
 coloured; the hair black and wavy; the eyes dark. The 
 nose, though broad, is not flattened, and the lips, though 
 thick, are mobile, as compared with those of the Negro. 
 When first known to Europeans, they were ignorant of agri- 
 culture, the manufacture of pottery was unkno^vu to them. 
 Even in weapons they were deficient, not possessing the 
 bow and arrow, those commonest weapons of the savage; 
 but, on the other hand, they are almost the sole possessors 
 of the boomerang, while the Esquimaux share with them the 
 use of the thro^ving stick. With an extensive sea-bord, they 
 
 
XANTH03IELAN0US RACES. 351 
 
 had very inferior canoes, and only the ^vestern Australians 
 had fishing hooks and nets. Their unity of descent is not 
 incompatible with the great dialectic differences T\^hich 
 Friederich Miiller points out as separating the tribes. These 
 are rather to be expected among people long resident in an 
 isolated continent, and grouped in tribes, which a not over- 
 plentiful sustenance kept in unceasing hostility. 
 
 The primitive inhabitants of southern India, the Dra vid- 
 ians, belong to this physical type. Largely displaced by the 
 Aryan invaders from the north-west, they still number over 
 32,000,000, a fifth part of the population of India. Five 
 linguistic divisions are recognised: Tamil, which is spoken 
 over the Carnatic and north Ceylon j Telegu, misnamed 
 Gentoo, on the east coast; Kannadi, at Mysore; Malayalma, 
 on the Malabar coast; and the decadent Talu of Mangalor. 
 
 The people who occupied south and wesfc Europe in pre- 
 historic times presented the same physical characters, so far 
 as may be inferred from their remains, and were either pure 
 representatives of the Melanous group, or were the earliest 
 members of that gi'oup, the Melanochroi, to which the 
 Xanthochroi and Melanoi contributed. In either case, there 
 was, on physical evidence, a continuous band of uniform 
 characters from Australia to western Europe, and it is note- 
 worthy that this band manifests likewise a remarkable 
 uniformity of megalithic structures from the Penrose Islands 
 to Scandinavia. 
 
 437. Xanthomelanous Races. — The north-east of Europe, 
 Asia north of the Caucasus and Himalayas, large part of 
 the Pacific Islands, and the whole of America, are occupied 
 by races which have a yello^^ish or reddish-brown tint of 
 skin, dark eyes, and black haii% usually long and straight. 
 The Asiatic t}^)es, inhabitants of the fertile plain of China in 
 the east, and of the deserts to the v.^est, the Mongol, the 
 Thibetan, and the Chinese, agree in the brachycephalic 
 character of the skull, in the squareness of their fiices, the 
 breadth of the cheek-bones, the usually oblique position of 
 the eyes, and their long, straight black hair. Some of these 
 characters become less conspicuous as we go southwards along 
 the east shores of Asia, and in the Malayan Peninsula we 
 find a race agreeing in complexion, in colour of hair, and in 
 
252 PHYSICAL GEOGRAPHY. 
 
 features, Avith tne Mongol, and connecting that group with 
 the inhabitants of the Pacific Islands as far south as New 
 Zealand. The Esquimaux or Innuit, in the north of 
 America, and with him goes the Greenlander, represents 
 the dolichocephalic type, to which the Samoiedes and Tun- 
 guses of northern Asia may be referred. They are of small 
 stature, broad faced, dark olive brown, with black hair, and 
 the skulls are low as well as long. These latter considerably 
 resemble the Mongols, even to the obliquity of the eyes, 
 while their language is structurally related to that of the 
 North Americans. 
 
 The other inhabitants of the New World, setting aside the 
 European, African, and Chinese colonist, agree generally in 
 having a complexion which is some shade of brown, or even 
 olive ; in having dark eyes and straight black hair. The 
 skull is broad, high, and long, the length being greater in 
 the northern and eastern tribes than in those to the south. 
 The face is broad, and the beard is slight. It is impossible 
 to recognise any important physical distinction between the 
 native peoj)les of America. 
 
 The Fuegians are at present the most wretched of people. 
 Their chief food is molluscs, fish, which they eat i-aw, bones, 
 and scales, and blubber, when it can be procured, whether 
 fresh or putrid. "With very scanty clothing at best, they 
 and their sucklings are often naked during snowfall. They 
 have no pottery, and scarcely any weapons. There is reason 
 to suspect that they are brachycej)halic, and that the 
 difference which exists between them and the other South 
 American tribes, may be due to divergence from a common 
 stock. 
 
 438. Xanthochroic Races. — The combination of abimdant 
 fair hair, blue eyes, and a pale skin, with a short skull, is 
 found in the Scandinavians, with a long skull in the Bel- 
 gians, South Germans, Swiss, Finns, and Slavonians. These 
 peoples at an early period occupied the centre of Asia, 
 between the Siberian low grounds and the Himalayas, from 
 the confines of China westwards by successive migrations 
 till they reached the extreme confines of Europe, the shores 
 of the Atlantic, while southwards they encroached at inter- 
 vals Qii the civilization of Greece and the power of Home. 
 
NEGROES, NEGRITOS. 353 
 
 At present they occupy Europe north of the Alps, mingling 
 in Russia with the Mongol, and in Britain with Melanochroi. 
 
 439. Melanochroic Races. — From the plains of India to 
 Britain, there may be traced along the track of the races 
 referred to under Art. 436, a series of peoples agreeing in 
 that they possess a pale complexion, with dark hair and eyes, 
 and for the most part dolichocephalic skulls. These peoples 
 occupy Persia, western Asia, and both shores of the Medi- 
 terranean, being mingled in France and Britain with the 
 preceding stock, which invaded and partly rej^laced them. 
 In former times they occupied Spain, south France, and 
 probably large part of western Europe, including Britain. 
 It is obvious that this mode of grouping races breaks up the 
 Aryan peoples into two distinct groups, the Xanthochroic and 
 Melanochroic, and while one division of the Celts belongs to 
 the former, the latter includes the dark Celts and the Semitic 
 tribes. Mr. Huxley is inclined to regard the Melanochroi 
 as an area of population which owes its character to the 
 spread of the Xanthochroi over earlier Australoid races. 
 
 440. Negroes, Negritos. — From Tasmania by New Cale- 
 donia, Papua, and Madagascar, a broken line of Ulotrichous 
 peoples may be traced, and on the other side of the Mozam- 
 bique Channel the same type extends over Africa south of the 
 Sahara. But again, in Africa, as in Australia, the antiquity 
 of the continent, and the isolation of its people, have given 
 rise to great diversity in speech, the dialectic varieties being 
 numerous and important. The woolly hair of the Negro is 
 evenly distributed over the head, and, taken in conjunction 
 with his black skin and prognathism, which gives his thick 
 immobile lips remarkable prominence, gives him an aspect 
 which is more characteristic than that of any other stock, 
 the oblique-eyed Mongol alone excepted. The Negro is a 
 cultivator of the soil; within the last few hundred years he 
 has acquired the art of dealing with metal : his weapons are' 
 the spear, bow and arrow, and are always carefully made, 
 often elegantly finished and ornamented. 
 
 The yellow-skinned Bushman has his hair disposed in locks, 
 not uniformly distributed. Inferior in size and strength to the 
 Negro, he also occupies a lower place as regards his civilization. 
 
 The Negritos of Tasmania were dark, prognathous, and 
 23 S 
 
354 PHYSICAL GEOGRAPHY. 
 
 dolicKoceplialic, while their frizzly hair separated them from 
 the Australian. Their civilization was even lower than that 
 of the Australian, since they had no canoes, very imperfect 
 fishing apparatus, and not even the throwing stick "as a 
 weapon. The Papuan, as described by Wallace, is of a sooty 
 brown or black, nearly but not so jet black as the Negro. 
 The harsh, dry, tufted hair, at first in short curls, forms in 
 the adult a thick mass projecting from the head. The nose is 
 prominent, broad, but not flattened. Apart from the volatile 
 restless energy of the Papuan, characters in which he con- 
 trasts with the Malay, the distinctness of the races is evidenced 
 by the fact that members of the Negrito stock now form the 
 population of the interior of many of the islands, and stand to 
 the Malay of the coast in the same relation as the Dravidian 
 to the Aryan, save that the antagonism between the former 
 two stocks is greater, and has resulted in the more depressed 
 state of the earlier peoj)le, driven into the fastnesses of the 
 island for safety from their conquerors. 
 
 441. Antiquity of Man. — What was the date of man's 
 first appearance, it is impossible to say. In Europe, human 
 remains are found under circumstances which prove him to 
 have been the contemporary of the reindeer, elk, rhinoceros, 
 mammoth {Elephas primigenius), and other animals which 
 are now either extinct, or are the characteristic inhabitants 
 of other regions. Thus the west of Europe was already 
 occupied by man long before the Xanthochroic races spread 
 from central Asia. At Hoxne, in Suflfolk, flint weapons 
 have been found in a deposit which lies on a hill slope, 
 and which is evidently the remainder of a formation largely 
 denuded during the excavation of the adjacent valley, thu^ Aj 
 indicating a remarkable change in the physical features of 
 south England since man first occupied it. At present, the 
 earliest indications of man are preserved in the deposits 
 formed during the amelioration of climate after the glacial 
 period. Of miocene man there is no certain evidence, neither 
 can it be held as demonstrated that in America he co- 
 existed with the mastodon. But, bearing in mind what has 
 been said in previous chapters, the student will see that if I 
 man lived in central Europe, in the valley of the Somme, i 
 when ice covered the river for a large part of the year, there I 
 
STONE PERIOD. 355 
 
 is no reason to doubt, from tlie analogy of other animals, 
 he must have lived elsewhere at earlier periods. What that 
 locality may have been, we shall immediately enquire. 
 
 442. Evidence of Antiquity. — The evidence requisite to 
 prove man's co-existence with extinct animals is: — 1. The 
 association of the remains of both in undisturbed deposits, 
 under circumstances which leave no doubt that both had 
 been laid do^Ti simultaneously. 2. The absence of rolling, 
 or those siinis of friction which would leave no doubt that 
 bodies so soft as bones had been transported for some distance, 
 and might therefore have come from more ancient accumu- 
 lations. 3. In the case of articles of alleged human manu- 
 facture, the appearances must be such that no accidental 
 fractures or other markings could be supposed to coincide. 
 
 443. Prehistoric Periods. — Successive epochs of man's 
 history are recognised by the character of the weapons he 
 employed. Three principal ej)ochs are: the Stone Period, 
 the Bronze Period, and the Iron Period, the material used 
 corresponding to the progress he had made in civilization. 
 But these periods, like those of the geological tables, are only 
 true for each country, or each race; they do not mark the 
 same chronological stages for all. Thus the Esquimaux is 
 at the present day in the stone period, while the Hottentot 
 has reached the iron, and these unequal grades are found 
 close to a higher civilization of European origin. Nor is the 
 material necessarily a guide to the degree of civilization, 
 since weapons of the stone age type have been found in 
 America fashioned out of copper, which was easily wrought, 
 but was not used as a metal. 
 
 444. Stone Period. — Stone weapons of a rude kind are 
 found associated with the extinct animals above mentioned; 
 while more elegant and beautifully polished weapons of 
 flint and other kinds of stone were carefully prepared at a 
 later date. The periods at which these weapons were in 
 common use arc distmguished : the earlier as the palaeolithic, 
 or arcliKiolithic, or drift period, the later as the neolithic or 
 polished stone period. The localities in which i^alaeolithic 
 remains are found are limited, and the men who used these 
 primitive tools must have led a life similar to that of the 
 Esc^uimaux. The river gravels, or drift, or quaternaiy 
 
356 PHYSICAL GEOGRAPHY. 
 
 deposits, the early bone caves, tlie kjokkenmoddings or sliell 
 mounds of Denmark, and tlie lowest portions of the peat- 
 mosses have yielded the chief remains of this early age. 
 These primitive races were, unlike the Esquimaux, without 
 domesticated animals; yet in some localities, as in Perigord, 
 the designs roughly carved on their instruments show some 
 artistic sense. The palseolithic men seem to have been con- 
 fined largely to the river valleys, and to the vicinity of the 
 sea, from which food could be more easily obtained, in which 
 it was also more abundant than on the land. 
 
 The neolithic age was that in which the tumuli or ancient 
 burial mounds were formed, the Pfahlbauten or lake dwell- 
 ings of Switzerland were erected, the shell mounds were in 
 part collected, and in which the later cavemen lived. The 
 celts (Lat. celtis, a chisel) are most frequently of flint, but 
 other kinds of stone were used, kinds which are not known 
 to exist in some of the countries where the wrought speci- 
 mens are found. Thus jade in Europe, mica in the Mississippi 
 valley, represent probably barter, and suggest that the tribes, 
 which thus obtained more valuable materials than those at 
 hand, had already attained some mercantile sagacity. The 
 monuments already referred to, as also showing, like the 
 stone instruments, uniformity over a large part of the earth's 
 surface, are the menhirs or standing stones, cromlechs or 
 stone circles, dolmens or stone chambers, and tumuli or bar- 
 rows of various kinds. These last belong to two distinct 
 types, long and round, and the skulls differ according to 
 the form of the barrow, thus 82 per cent, of long barrow 
 skulls had a cephalic index of "63 to "73, while 63 per 
 cent, of those from the round barrows were brachisto- 
 cephalic. The lakemen of Switzerland belonged to two 
 periods, those of the neolithic having left most abundant 
 remains in eastern Switzerland, the west and central districts 
 containing the habitations belonging to the bronze period, 
 while those of the iron period are represented only on the 
 lakes of Bienne and Neufchatel. Crannoges, as the pile 
 dwellings of Irish and Scotch mosses are called, are found 
 under circumstances which suggest that this mode of securing 
 cattle may have been adopted in later troublous times, but 
 the majority are ^rehistorip constructions, The existing 
 
PRIMITIVE HOME OP MAN. 357 
 
 equivalents ai*e the platform dwellings on the Orinoco, and 
 in New Guinea. It is not possible to determine the races 
 of the neolithic period in Europe; but as we approach the 
 later part of that epoch, it is noteworthy that the size of the 
 weapon handles increases, indicating a corresponding differ- 
 ence in the manufacturers. The shell mounds of Denmark 
 and west Scotland commence in the earlier part of the neo- 
 lithic 2^Griod, and extend down to the bronze, while the 
 Tasmanian and Fuegian have within historic times prepared 
 exactly similar piles. 
 
 445. Metallic Age. — The gi-adual amelioration in the con- 
 dition of the early Europeans, shown by the discovery and 
 increasing dexterity in the use of metals, by the introduction 
 of domesticated animals, and by the greater attention to agri- 
 culture, presents a history far too complicated to be more 
 than alluded to here. The range and importance of that 
 history may be inferred from the change of meaning whereby 
 Max Mllller shows Aryan roots indicating copper were trans- 
 ferred to iron, and the word which originally stood for fir 
 became quercus (oak) while the Greek word (pijyoe (oak) is 
 represented in Latin by fagus (beech) ; these terms, taken 
 in the order mentioned, corresponding to the order in which 
 the trees succeed each other from below upwards in the 
 Danish peat mosses. Hesiod tells of a period when cojoper 
 only, not iron, was used, and alludes to a time when weapons 
 did not exist ; that is, when metals were unknown. The pas- 
 sage, therefore, from the stone period to the epoch of Greek 
 civilization may have been witnessed by those whose traditions, 
 slowly accumulated, constituted the mythology of early Greece. 
 
 446. Primitive Home of Man. — It is clear, then, that man 
 had occupied Europe long, we know not how long, before 
 that period the tradition of which represents the human 
 family as diverging from central Asia. That dispersion was, 
 in fact, the latest event in the history of the race. If our 
 interpretation of events in the palaeolithic age is correct, we 
 have evidence of a mode of life not higher than that of the 
 Esquimaux, while every succeeding epoch shows some im- 
 provement, some sign of progress. To the question, Did a 
 higher civilization exist at any other part of the earth's 
 siuface % it is impossible to reply positively. Following up 
 
358 PHYSICAL GEOGRAPHY. 
 
 the line of argument adopted in tlie previous chapter, Ihe' 
 dispersion of tribes from a common centre must coincide with 
 changes both in the travellers and those vrho remain; and 
 change means progress in the case of man. If, therefore, we 
 find that man in western Europe had not at first risen beyond 
 the conditions of the Esquimaux, it is fair to ask, how low 
 in condition was he in his primitive home ? or had he retro- 
 graded 1 The student must, in all such inquiries, bear in mind 
 that retrogression is an uncertain phrase; that the Englishman 
 who spends a winter in the Arctic region retrogrades, inas- 
 much as his habits are no longer those of his home, but 
 suitable to his surroundings; thus he adopts the custom of 
 eating his food raw, that being apparently most conducive to 
 health. It is open to doubt whether man had not reached 
 Europe before the glacial epoch, and whether the river-gTavel 
 folk were not, like the reindeer, driven back from their more 
 northern homes, carrying with them the customs and habits 
 suitable to a rigorous climate. On their retreat northward 
 (for their extinction at that period is an unnecessary assump- 
 tion), they were seemingly overlapped by the advancing wave 
 of brachycephalic men of the bronze age, and thus the pi-ogi'ess 
 of western Europe was made up of that which might result 
 from intelligent efforts to deal with surroundmg difficulties, 
 and of that which was borrowed from the invaders. Eor it 
 must be remembered that man passed through regions of 
 unlike conditions, and his exijeriences thus varying, his 
 inventions and discoveries would likewise vary. The flint 
 arrow-head was tied on the shaft as the shark's tooth, from 
 which it was copied, was tied on, and the spread of this 
 pattern over inland America is, like the exchange of copper 
 and shells between Lake Superior and the coast, another 
 evidence of the early intercommunication of these primitive 
 folk. But when the comparative facilities for migration 
 offered by Europe and America did not exist, there man Avas 
 arrested, so to speak, in his intellectual development. His 
 needs were, in the tro2:)ical islands, easily supplied; and, in 
 the absence of competition from without, excessive popula- 
 tion w^ould be checked, as at the present day in the Pacific 
 and Fuegia, by killing the children and the aged, while the 
 dominant family or families are protected from starvation by 
 
Hypothesis op lost continents. 359 
 
 tlie taboo wliicli they lay on the cocoa-nut tree, which there- 
 after it is a capital offence to touch. Other things being 
 equal, the occupants of an isolated country might be expected 
 to have more and more varied aj)pliances, if they had tra- 
 velled to it, than if they were isolated after long residence 
 there. The Australian and the Tasmanian are feebly endowed 
 with the appliances for physical comfort : the more migratory 
 Kegrito, in addition to the necessity imposed on him by his 
 changes of place, was brought in contact with other tribes, 
 and profited by the contact; hence civilization is greater 
 among that stock nearer the equator than it is farther south. 
 If the student follows up this suggestion, and compares the 
 condition of different peoples, he will find reason to believe 
 that while in all cases a tendency to improve has been shown, 
 and while in very few cases has this tendency failed to pro- 
 duce fruit, the amount of improvement depends chiefly on the 
 necessities imposed by the surroundings of each nation, and 
 thus we have many different stages representing the condition 
 of the people when they first reached localities that did not 
 cultivate their inventive powers. 
 
 447. Hypothesis of Lost Continents in Southern Hemi- 
 sphere. — It is clear that, unless the palaeolithic men were 
 created in western E jpe, they must have spread from some 
 centre which lay beyond the limits of Europe, and which lay 
 to the south of Asia. The Australian is cut ofi" from his 
 Dravidian kindred by the Negritos, which stretch from New 
 Caledonia round by New Guinea, across the Indian Ocean to 
 Madagascar. The Malay has moved south-eastwards, if we 
 take language as the guide, for the Arabic element disappears 
 in that direction; but did he not first remove northwards'? 
 There is abundant physical proof of elevations and sub- 
 sidences in that great area which lies between Africa and 
 S, America, in the middle of which is Australia, a fragment 
 of a continent homomorphic with the other two. The Tas- 
 manian is a Negrito who cannot have reached his present 
 home by sea, else the tradition of his canoe would have 
 survived, nor was there any system of slavery, before the 
 Europeans arrived, by which the process of iransplanting 
 might have been efiected. The reasons for asserting the 
 existence of continuous land in the Southern Ocean^ at a 
 
360 t>HYSICAL GEOGRAtnl^. 
 
 geologically recent period, were stated in last chapter : from 
 this land, now to a large extent submerged, man jprobably 
 spread on all sides, not necessarily in every case over con- 
 tinuous land; for navigation undoubtedly bad an early 
 sbare in the dispersal of tribes. The Australoid and the 
 woolly-haired races may be regarded as presenting the two 
 earliest divergences from the common stock, the latter, how- 
 ever, reaching sooner to the limits of their progress than 
 the descendents of the former. The distribution of the 
 woolly-haired peoples is singularly close in its parallelism 
 to that of the marsupials, lemurs, struthious birds, and 
 anthropoid apes, the Sahara sea having apparently cut them 
 off entirely from contact with the primitive dolichocephalic 
 (probably Melanous) races of the Mediterranean shores. The 
 community and diversity of characters between the Mongol 
 and the American are intelligible on this hypothesis. The 
 Malayan type does not appear in southern India, the lines 
 along which the tribes travelled being thus so far parallel, 
 not coincident. The glacial period doubtless suspended the 
 spread and development of man during a considerable period 
 prior to his advent in western Europe. To the period of 
 climatal improvement we ought probably to assign that later 
 dispersion from central Asia which flooded India with pale 
 faces, whose influence is manifested in the mixed stock, 
 the Melanochroic, by which Dravidian tribes are isolated. 
 To this time perhaps belongs the Malayan reflux which has 
 given so great uniformity to the Polynesian area. It is clear 
 that the terms, Aryan or Indo-European, are suggestive of 
 theories on which we cannot decide, for the Xanthochroi 
 may be primarily derived from Asia, east and north of the 
 Himalayas, and Indo-Euroj)ean may thus be only true, in 
 a secondary sense, for their Melanochroic descendants. Be 
 this as it may, the fact is that, while the Xanthochroi are 
 not traced beyond the Himalayas, the Melanochroi — all those 
 peoples with whom religion, science, and art, have attained 
 their earliest and greatest development — all those people to 
 whom grand conceptions, whether in cycloj)ean masonry, 
 or in gigantic empires, seem natural — extend to the shores of 
 the Indian Ocean. 
 
COMPARATIVE TABLE OF FAHRENHEIT AND 
 CENTIGRADE DEGREES. 
 
 Boiling point on the Fahrenheit scale is 212°, or 180° F. above the 
 freezing point (32°), and as boiling point on the Centigrade scale is 
 100° above 0°, it follows that 180° F. are equal to 100° C, that is, 
 9°F. are equal to 5°C. The result may be formulated thus : — 
 
 Degrees F. X 5 ^ p (.- 
 
 9 
 
 But as the Fahrenheit scale commences 32° F. below freezing point, 
 the Fahrenheit scale may be reduced to the Centigrade, or Centigrade 
 to Fahrenheit, thus : — ' 
 
 ,1 ) 5(TemperatoeF.-32°) ^ T,„p„^t„, 0. 
 
 Ex. : 5(55'r. -32°) _ 12.70c. 
 ^, J 9 Temperature C.+ 32° ^ Temperature F. 
 
 
 c 
 
 
 
 
 
 
 
 
 (^' 
 
 ^%^ 
 
 d 
 
 fe| 
 
 (** jiS 
 
 d 
 
 fe' 
 
 fe'®(2 
 
 d 
 
 60 
 
 
 to 
 
 CD 
 
 n 
 
 ft 
 
 
 to 
 
 n 
 
 
 -la 
 
 ft 
 
 32 
 
 p 
 
 fe 
 
 
 
 
 
 
 S, 
 
 
 
 
 00 
 
 32 
 
 
 
 
 
 51 
 
 19 
 
 10-5 
 
 31 
 
 1 
 
 - 0-5 
 
 33 
 
 1 
 
 0-5 
 
 52 
 
 20 
 
 111 
 
 30 
 
 2 
 
 - 11 
 
 34 
 
 2 
 
 11 
 
 53 
 
 21 
 
 11-6 
 
 29 
 
 3 
 
 - 1-6 
 
 35 
 
 3 
 
 1-6 
 
 54 
 
 22 
 
 12-2 
 
 28 
 
 4 
 
 - 2-2 
 
 36 
 
 4 
 
 2 2 
 
 55 
 
 23 
 
 12-7 
 
 27 
 
 5 
 
 - 2-7 
 
 37 
 
 5 
 
 2-7 
 
 56 
 
 24 
 
 13-3 
 
 26 
 
 6 
 
 - 3-3 
 
 38 
 
 6 
 
 3-3 
 
 67 
 
 25 
 
 13-8 
 
 25 
 
 7 
 
 - 3-8 
 
 39 
 
 7 
 
 3-8 
 
 58 
 
 26 
 
 14-4 
 
 24 
 
 8 
 
 - 4-4 
 
 40 
 
 8 
 
 4-4 
 
 59 
 
 27 
 
 150 
 
 23 
 
 9 
 
 - 50 
 
 41 
 
 9 
 
 5 
 
 68 
 
 36 
 
 20 
 
 22 
 
 10 
 
 - 5-5 
 
 42 
 
 10 
 
 5-5 
 
 77 
 
 45 
 
 25 
 
 21 
 
 11 
 
 - 61 
 
 43 
 
 11 
 
 61 
 
 86 
 
 54 
 
 30 
 
 20 
 
 12 
 
 - 6-6 
 
 44 
 
 12 
 
 6-6 
 
 95 
 
 63 
 
 35 
 
 19 
 
 13 
 
 - 7-2 
 
 45 
 
 13 
 
 7-2 
 
 104 
 
 72 
 
 40 
 
 18 
 
 14 
 
 - 7-7 
 
 46 
 
 14 
 
 7-7 
 
 113 
 
 81 
 
 45 
 
 17 
 
 15 
 
 - 8-3 
 
 47 
 
 15 
 
 8-3 
 
 122 
 
 90 
 
 50 
 
 16 
 
 16 
 
 - 8-8 
 
 48 
 
 16 
 
 8-8 
 
 212 
 
 180 
 
 100 
 
 15 
 
 17 
 
 - 9-4 
 
 49 
 
 17 
 
 9-4 
 
 . . fl 
 
 
 . • . 
 
 14 
 
 18 
 
 -100 
 
 50 
 
 18 
 
 100 
 
 
 
 ... 
 
INDEX. 
 
 Aesorptioit, of heat, 244; of liglit, 251; 
 bands in spectram, 252. 
 
 Africa, Mountains of, 83; plateaux of, 02; 
 north current, 131; rivers of, 159-161; 
 lakes, 173; winds of, 231; niousoous 
 of, 269. 
 
 Alluvium, 36, iTT. 
 
 Alps, 80, 215. 
 
 America, Plateaux of, 93 ; rivers of, 
 153-155 ; rivers. Pacific slope, 164 ; 
 winds of North, 264; winds of South, 
 265 ; monsoons of, 263; biological 
 provinces, 341; races of, 352. 
 
 Amygdaloid, 35. 
 
 Aualj'-sis of Thames "Water, 109; of lake 
 waters, 174; of sjiring water, 1S3. 
 
 Andes, 84, 215. 
 
 Antarctic Drift, 131, 132. 
 
 Anticline, 45. 
 
 Antitaurus, 80, 161. 
 
 Aqueous Vapour, 195; condensation of, 
 196, 203; transport of, 203; pressure 
 of, 195, 240; influence on tempera- 
 ture, 243. 
 
 Aralo- Caspian Area, 92. 
 
 Arctic Current, 128. 
 
 Artesian Wells, 181; yield of, 1S3. 
 
 Aryan Races, 350, 353, 360. 
 
 Asia, Central, Mountains of, 82 ; pla- 
 teaux of, 92. 
 
 Asia, Eastern, Rivers of, 103. 
 
 Asia Minor, Mountains of, SO; rivers of, 
 161. 
 
 Astronomy, Relation, to Physical Geo- 
 graphy, 10. 
 
 Atlantic, 102; soundings, 104; tempera- 
 ture of, 114, 126 ; warm and cold 
 areas in, 128, 325; winds of, 260; 
 marine provinces, 339, 341. 
 
 Atmosphere, Solid Particles in, 238; water 
 in, 194; saturation of, 196; composi- 
 tion of, 238; density of, 239; height 
 of, 239 ; temperature of, 240, 245 ; 
 temperature affected by liumiditj', 
 243 ; diathermancy, 243 ; cooled by 
 expansion, 243; decrease of tempera- 
 ture with height, 246, 289; colour 
 
 and transparency, 250; polarization 
 of, 254; movements of, 258; constant 
 westerly current, 260; course of cur- 
 rents, 269; electricity of, 280. 
 
 Aurora Borealis, 280. 
 
 Australia, Currents, 133 ; rivers of. 164 ; 
 terrestrial life, 328, 329; marine, 341. 
 
 Auvevgna, Springs, 184, 188. 
 
 Avalanche, 226. 
 
 Axis of Earth, stability, inclination, 14. 
 
 Axis of Elevation, 70,85; of mountain 
 chains, 78, 307, 
 
 Backwater, 144. 
 
 Baltic CuiTent, 130. 
 
 Barometric Pressure, Influence on Cli- 
 mate, 239, 289. 
 
 Barrande's Colonies, 43. 
 
 Bathymetrical Zones, 324. 
 
 Beaufort's Scale of "Winds, 273. 
 
 Bergschrund, 224. 
 
 Biological Provinces, unequal, 326. 
 
 Bischof's Classification of Siniugs, ISS. 
 
 Black Sea Current, 130. 
 
 Bora, 261. 
 
 Bore, 123. 
 
 B:)tanical Provinces, 322. 
 
 Botany, Relation to Physical Geograi^hy, 
 9. 
 
 Boulder Clay, upper and lower, 232 ; 
 deposit of, 233 ; moraine i^rofonde of 
 ice sheet, 232. 
 
 Brahmapootra, 82, 161. 
 
 Brazilian Current, 131. 
 
 British Islands, 157; climate of, 2SS; inha- 
 bitants of, 353. 
 
 Buchan, 290. 
 
 Calcareous Springs, 1S4. 
 Calms, Regions of, 263, 204. 
 Canons, 161. 
 
 Caribbean Sea, Biological Province, 340. 
 Cataracts, 150. . " . - 
 
 Caspian, 92, 171. 
 Caucasus, 80, 215. 
 
 Caves, Formation of, 120 ; enguMng 
 rivers, 190; contents of, 192. 
 
INDEX. 
 
 8G3 
 
 Celts, i?eople, 353; instruments, 356. 
 
 Cephalic Index, 349. 
 
 China : Monsoons, 268 ; typhoons, 274 ; 
 people of, 351. 
 
 Cirrus, 200. 
 
 Cleavage, 34. 
 
 Cliffs, <J0. 
 
 Climate, 284; continental and insular, 
 287 ; influence of currents on, 287; 
 of British Isles, 288; of lake regions, 
 289; of marsh areas, 289; influence 
 of surface features, 289 ; influence 
 of vegetation, 291 ; cycles of, 291 ; 
 influence on distribution, 322, 331. 
 
 Clouds, 200; velocity of, 201; colour of, 
 202. 
 
 Coal, Formation of, 31; varieties of, 31. 
 
 Coast Ice, 235. 
 
 Coast Line, 56. 
 
 Colonies, Barrande's, 43, 343. 
 
 Colour of Water, 116; of clouds, 202; of 
 snow, 213; of atmosphere, 250. 
 
 Condensation of Vapour, 196, 203, 
 
 Conduction, 113, 257; of electricity, 281. 
 
 Contemporaneity, 40, 42, 335. 
 
 Continental Rivers, 165; lakes, 173. 
 
 Continents, Tapering Southwards, 13, 56; 
 area of, 55; evolution of, 59; dryness 
 of, 207; lost, hvpothesis of, 338, 359; 
 Atlantic, 60; Pacific, 65. 
 
 Convection, 113, 216, 246, 257. 
 
 Coral Reefs, 29, 68, 341. 
 
 Co-tidal Lines, 124. 
 
 Crannoges, 356. 
 
 Cretaceous Epoch, Continuity of, 341. 
 
 Crevasses, 217, 224. 
 
 CroU on Cycle of Climate, 292. 
 
 Cromlechs, 356. 
 
 Cumulus, 200. 
 
 Currents, Marine, 124; origin of, 125: St. 
 Roque, 125 ; Arctic, 128 ; Labrador, 
 129 ; Rennel's, 129 ; Mediterranean, 
 130: Gibraltar, 130; Black Sea, 130; 
 Baltic, 130 ; IBrazilian, 131 ; equa- 
 torial, 132 ; Guinea, 132 ; equatorial 
 Pacific, 133 ; Humboldt's or Pem- 
 vian, 132 ; Australian, 133 ; Indian 
 Ocean, 134; Red Sea, 135; south con- 
 necting, 132; Natal or Mozambique, 
 135; influence of, on climate, 287. 
 
 Currents, Atmospheric, theories of, 258; 
 constant westerly, 200; course of, 269. 
 
 Cutch. Runn of, 78. 
 
 Cyclones, 210, 274. 
 
 Dead Sea, Saltness, 110; in an area of 
 
 subsidence, 172. 
 Deep Sea, Soundings, 53, 103 ; life in, 
 
 117, 341, 
 Deltas, 95 ; of Po, 96 ; of Mississippi, 96 ; 
 
 of Rhone, 97; of Nile, 97; formed in 
 
 valleys, 150; in lakes, 177. 
 Denudation, 36 ; marine, 71 : plains of, 
 
 94; by rivers, 146; glacial, 222. 
 
 Deserts, Rainless, 98. 
 
 Development, Progressive, 343. 
 
 Dew, 197; compared with fog, 202. 
 
 Diallage, 34, 
 
 Diathermancy, 243. 
 
 Dip, 44. 
 
 Distribution of Life, Laws of, 320; hori- 
 zontal and altitudinal, 323; in siJacQ 
 and time, 333. 
 
 Doldniras, 264. 
 
 Dolerite, 35. 
 
 Dolmens, 356. 
 
 Dolomite, 34. 
 
 Dove's Law of Temperature, 235. 
 
 Drainage Areas, 139; relation to oceans, 
 grouping of, 152. 
 
 Dravidian, 351, 359. 
 
 Drifts, 124; equatorial, 125, 132, 133; 
 Antarctic, 131, 132. 
 
 Dry Land, Relief of, 70. 
 
 Dunes, 77. 
 
 Dust Storms, 277. 
 
 Earth, Diameter and Form, 13; axisof, 14; 
 cubic contents, 14; specific gravity, 
 14, 15; internal fluidity, 15-17; com- 
 position of crust, 19, 
 
 Earthquakes, 306; wave, 308-310: co-seis- 
 mic lines of, 308; area of, 311; distri- 
 bution of, 311; and volcanoes, causes 
 of, 311 ; Sir W. Thomson on, 311 ; 
 periodicity of, 315; influence of moon, 
 315. 
 
 Ecliptic, 14; obliquity of, 14; inflvience on 
 climate, 293. 
 
 Electricity, Atmospheric, 280; and heat, 
 281, 
 
 Elevation and subsidence, 315. 
 
 England, Valleys of, 88, 139 ; thermal 
 waters, 186, 
 
 Equatorial Drift, 125, 132; Pacific, 133. 
 
 Equinoxes, Precession of, 17 
 
 Erratics, 234. 
 
 Escariiments, 90. 
 
 Eskers, 167. 
 
 Esquimaux, 352. 
 
 Etesian Winds, 261. 
 
 Ether, 239; vibrations of, 250. 
 
 Ethiopian Region, 327, 329. 
 
 Ethnology. See j\Ian. 
 
 Euro)ie, JPlateaux of, 92 ; rivers of, 150- 
 15;). 
 
 Evaporation, 195; its effects, 105; from 
 soils and plants, 190; from glacier, 
 224, 243. 
 
 Evolution, 343, 34G. 
 
 Facial Angle, 346. 
 
 Fauna, 317; relation of living to extinct, 
 318 ; mesozoic compared with Aus- 
 tralian, 318; Wealden, 318; insular, 
 C4, 337; deep sea, 117, 34!.. 
 
 Faults, 46. 
 
 Felstoue, 35, 
 
864 
 
 INDEX. 
 
 Fens, 99. 
 
 Ferruginous Springs, 1S6. 
 
 Fiords, 75. 
 
 Fireclay, 32. 
 
 Firn, 218. 
 
 Flints, .354, 355. 
 
 Floods, 145. 
 
 Flora, 317; relation of living to extinct, 
 318; number of species, 319; insular, 
 337. 
 
 Fluidity, Internal, of Earth, 15-17. 
 
 Fog, 198; of Newfoundland, IPS; of cities, 
 199; of hills. 199; height of, 199; 
 compared with dew, 202 ; of Peru- 
 vian coast, 266. 
 
 FiJhn, 262; in New Zealand, 263. 
 
 Formations, Geological, 20; unconformity 
 of, 22 ; sedimentary, not all marine, 
 22 ; definition of 52 ; red colour of, 
 177. 
 
 Fossils, Preservation of, 49. 
 
 France, Springs of, 187. 
 
 Freezing Point of Water, 112, 216, 
 
 Puegians, 352. 
 
 Fulgurites, 283. 
 
 Ganges, 149, 161. 
 
 Geikie, A., 85; 87, 147. 
 
 Geikie, J., 34. 
 
 Geography, Botanical and Zoological, 
 322. 
 
 Geology, its Relation to Physical Geo- 
 graphy, 9, 12. 
 
 German Ocean, 102. 
 
 Geysers, Theory of, 814. 
 
 Gibraltar Current, 130. 
 
 Glacial Period, 231, 294. 
 
 Glaciation, Signs of, 230, 236. 
 
 Glacier, Formation of Lakes by, 168 ; 
 motions of, 217, 221, 223; plasticity 
 of, 217; genesis of, 219; structure of, 
 220 ; erosion of channel, 222 ; con- 
 trasted with river, 223; bifurcation 
 of, 223 ; crevasses, 217, 224 ; dirt 
 bands, 224 ; diminution of, 224 ; 
 dimensions of, 225; stream beneath, 
 229; at sea-level, 230; compared with 
 ice-sheet, 232. 
 
 Gneiss, 33. 
 
 Granite, 33 ; sometimes volcanic, 83 ; 
 various ages of, 34 ; contraction of, 
 315. 
 
 Ground Ice, 285. 
 
 Gregale, 261. 
 
 Guinea Current, 132. 
 
 Gulf Stream, 126 ; temperature of 126 ; 
 cause of, 127; influence of, 288, 292. 
 
 Hadley's Theory of Currents, 258. 
 Hail, 236; structure of, 236; relation to 
 
 Btorms, 237. 
 Hair, a character of race, 350. 
 Harmattan, 261, 274. 
 
 Heat, distribution, 14; internal, 15, 17 
 amount received, IS; specific heat, 
 113; compared with electricity, 2S1 
 absorption of, 244 ; by gases, 244 
 by perfumes, 244; effects of, 247 
 equivalent of, 247; analogies of, 248 
 dark rays of light, 249; transmission 
 of, 257; reflection of, 257. 
 
 Hills, 76; fogs of, 199. 
 
 Himalayas, SO, 205, 215.- 
 
 Hoar-frost, 213. 
 
 Homogenetic resemblances, 334. 
 
 Homomorphisra, 56, 205. 
 
 Homoplastic resemblances, 334. 
 
 Homotaxis, 42, 335. 
 
 Hornblende, 34. 
 
 Horse Latitudes, 268. 
 
 Humboldt's current, 132. 
 
 Humidity, see vapour, aqueous. 
 
 Hurricanes, West Indies, 263, 274; 
 Indian Ocean, 274. 
 
 Huxley, on Ethnology, 349 ; on homo- 
 taxis, 337. 
 
 Hypersthene, 34. 
 
 Hypogene, Rocks, S3; phenomena, 297. 
 
 Iceberg, 226, 234, 235. 
 
 Ice, geoiogital importance of, 215; influ- 
 ence of pressure on, 217; plasticity 
 of, 217; passage of snow into, 218 
 lake and glacier, 221; pack, 235 
 foot, 235; coast, 235; ground, 235 
 floe, 235. 
 
 Iceland, springs of, 187. 
 
 Ice-sheet, definition of, 232. 
 
 Indian Ocean, soundings of, 107, 150; 
 temperature of, 115; currents, 134; 
 winds of, 266. 
 
 Indian Province, 328, 329. 
 
 India, Southern plateau, 162; races, 351. 
 
 Indo-Pacific, biological province, 340. 
 
 Indraughts, 124. 
 
 Indus, 82, 161. 
 
 Innuit, 352. 
 
 Insular period, hypothesis of, 61; faunas 
 and floras, 64, 337. 
 
 Ii'on ores, varieties of, 32. 
 
 Islands, definition of, 63; classification 
 of, 64; Pacific, 64; volcanic, 67, 09; 
 coral, 67. 
 
 Isocheimal lines, 286. 
 
 Isotheral lines, 286. 
 
 Isothermal lines, 280, 
 
 Joints, 46. 
 
 Kames, 107. 
 Karroo, Africa, 83. 
 Kitchen middens, 356. 
 Kloofs, Africa, 83. 
 
 Labrador Current, 129. 
 Lagging of tides, 122. 
 Lagoons, 107. 
 
INDEX. 
 
 S65 
 
 Lake dwellings, 356. 
 
 Lakes, fiUingup of, 41, 177; classification 
 of, 166 ; obstruction of, 167, 172; of 
 glacier erosion, 16S; theory of origin 
 of, 166; Ramsay on, 169; of -west 
 Scotland, 170; of subsidence, 171; 
 pitch, 172, 190; continental, 173; 
 African, 173; analysis of Avaters of, 
 17-1; salt, 175; ancient, 175; old red 
 sandstone, 176; ice of, 220; lake re- 
 gions, climate of, 289. 
 
 Landes, 95. 
 
 Land, relief of, 70; proportion of to 
 water, 102; specific heat of, 113; 
 temperature of, 290. 
 
 Landslips, 193. 
 
 Language, Ethnological Value, 347; Hum- 
 boldt's classification, 347 ; Muller"s 
 classification, 347 ; development of, 
 347. 
 
 Lava, 299; texture of, 301; composition 
 of, 301; underground course of, 803. 
 
 Leiotrichous, 350. 
 
 Levanter, 262. 
 
 Light, distribution, 14; analogies of, 24?; 
 velocity of, 250; absorption of, 251; 
 refraction of, 251; cbeniical rays of, 
 249; heat rays, 249; polarization, 253. 
 
 Lightning, 283. 
 
 Limestones, 30; bituminous and fetid, 
 30; crystalline, 34; magnesian, 34. 
 
 Loess, 228. 
 
 London, Avater supply of, 178. 
 
 Madagascar, 327, 353. 
 
 Magnetism, terrestial, 278; magnetic 
 poles, 278; ec[uator, 279; variation, 
 279; intensity, 279; storms, 280. 
 
 Malay, 351, 360. 
 
 Man, Geographical influences on, 62 ; 
 345 ; power of endurance, 345 ; 
 Esquimaux,352; unity and plurality, 
 346;Monogenists, 364: species (?), 346; 
 skull, proportions of, 349; brachyce- 
 phalic, 349 ; dolichocephalic, 349 ; 
 facial angle, 349; orthognathism, 349; 
 prognathism, 349; Melanous, 350; 
 Xanthomelanous, 351; Xanthochroic, 
 352; Melanochroic, 353; Negro, 353; 
 contemporaneous with extinct ani- 
 mals, 354 ; evidences of antiquity, 
 355; stone age, 355; metallic age, 
 357; primitive home of, 357; dis- 
 persion of, 360. 
 
 ]\rarmora, Sea of, 103. 
 
 IVlaury's Theory of Cixrrents, 253. 
 
 Mediterranean, 10;^; soundings of, 106; 
 temperature of, 114, 115; currents, 
 130; winds of, 261. 
 . Megaliths, 351. 
 
 Melanochroic races, 353. 
 
 Melanous Races, 350. 
 
 Melaphyre, 35. 
 
 Menhir, 356. 
 
 Metamorphism, 33, 297. 
 Meyen, Zones of vegetation, 323 
 Migration of species, 330; results of, 282f 
 
 dar-gers of, 335. 
 Mimicry, 334. 
 Minerals, 24. 
 Mineral Springs, 184. 
 Mississip])i, Delta of, 9Q; ancient valley 
 
 of, 154. 
 Mist, 198. - 
 Mistral, 261. 
 Mongol, 351. 
 Monsoons, 265, 269; Chinese, 268; 
 
 African, 269; American, 268. 
 Moon, 18, 19; influence on weather, 295; 
 
 influence on volcanoes and earth- 
 
 qiiakes, 315. 
 Moraines, 78, 226; galets of, 228; moraine 
 
 profonde, 232. 
 Mountaius, 58, 76 ; E. de Beaumont's 
 
 theory, 86; forms of, 89; table of 
 
 axes, 307. 
 Mozambique Current, 135. 
 
 Natal Current, 135. 
 
 Natural Selection, 333. 
 
 Negrito, 353. 
 
 Negro, 353. 
 
 Neolithic Age, 356. 
 
 Neotropical Region, 327. 
 
 Neve, 218. 
 
 Newfoundland, Fogs of, 19S. 
 
 New Zealand, Winds of, 263. 
 
 Niagara, 151. 
 
 Niger, 160. 
 
 Nile, 159; delta of, 97. 
 
 Nullahs, 101. 
 
 Oasis, 98. 
 
 Oaze, 30, 38, 108. ' 
 
 Oceans, 102; Pacific, 102; Atlantic, 102 
 German, 102 ; Mediterranean, 103 
 form of floor, 107 ; deposits in, 108 . 
 specific gravity of, 110; temperature 
 of, 112, 126, 290; luminosity of, 118; 
 density at freezing point, 112, 216. 
 
 Old Red Sandstone, Lacustrine, 176. 
 
 Orbit, Eccentricity of Earth's, 17; influ- 
 ence on climate, 292. 
 
 Orthognathism, 349. 
 
 Ourals, 79, 270. 
 
 Outcrop. 41. 
 
 Ozone, 143. 
 
 Pacific, Islands, 64; ocean, 102;soundinga 
 of, 107 ; drift, 132 ; winds of, 2C5; 
 races of, 353, 360. 
 
 Palfearctic Province, 329 
 
 Palaiolithic Age, 355. 
 
 Palaeontology, Relation to Physical Geo- 
 graphy, 10. 
 
 Peat Mosses, 99. 
 
 Periodic Rainfall, 208, 209» 
 
366 
 
 INDEX. 
 
 Peruvian Cuvrentj 132; fogs, 26G. 
 
 Petroleum, 190. 
 
 Plahlbauten, 356. 
 
 Phosphorescence of Sea, 113. 
 
 Phvsical Geography, its Scope, 9-12. 
 
 Pitch Lakes, 172, 190. 
 
 Plains, 93; of deposit, 93; of denudation, 
 94; salt, 165; aUuvial, 177. 
 
 Planets, Disturbing Inftiieuce of, 18. 
 
 Plants, Evaporation from, 196; species, 
 319; usefu] to man, 323. 
 
 Plasticity of Ice, 217. 
 
 Plateau, 91 ; of Asia, 92 ; Europe, 92'; 
 Africa, 92 ; America, 93 ; volcanic, 
 91; Thibetan, 92, 161; southern In- 
 dia, 102. 
 
 Po, Delta of, 90. 
 
 Polarization, 253. 
 
 Polders, 99. 
 
 Pollution of ^Yater, 130. 
 
 Porjphyry, 35. 
 
 Portage of Rivers, 152. 
 
 Port, Establishment of, 12i. 
 
 Prairies, 98. 
 
 Prehistoric Period, 355. 
 
 Pressure of Water, 112; of aqueous va- 
 pour, 195, 210; influence on freezing 
 point, 216. 
 
 Pressure of atmosphere, 239; relation to 
 storms, 271 ; influence on climate, 
 2S9. 
 
 Priming of Tides, 122. 
 
 Prognathism, 319. , 
 
 Provinces, Botanical and Zoological, 
 non-coincidence of, 322; aquatic and 
 subaerial, 322; how determined, 324; 
 biological, unequal, 326 ; Sclater's, 
 320; neotropical region, 327; Ethio- 
 pian, 327; Indian, 328; Australian, 
 328; Paloearotic, 329; marine, 324, 
 339; N. Atlantic, 330; Caribbean, 
 340 ; Indo-Pacific, 340 ; Australian, 
 341; Western S. America, 341. 
 
 Pyrenees, SO. 
 
 QuAQUAVErvSAL Watershed, 139. 
 
 Races of Man, 346; Dolichocephalic, 349; 
 Brachycephalic, 349 ; Melanous, 350 ; 
 Xanthomelaiious, 351; Xanthochroic, 
 352; jMelanocluoic, 353 ; Negro, 353; 
 Negrito, 353; Celts, 353; Aryan, 353, 
 360; Semitic, 353; Malayan, 351, 300; 
 Pol3'nesiau, 353 ; Dravidian, 351, 
 359. 
 
 Radiation, Solar, 18, 115, 244, 245; Ter- 
 restrial. 243-245. 
 
 Rainfall. 201; conditions of, 204; depend- 
 ent on winds, 205, 207; affected by 
 high grounds, 205; not incessant, 206; 
 aliected by vegetation, 207; from clear 
 sky, 208; periodic, 208, 209; variable 
 and constant, 208; table of 209, 210; 
 and cyclones, 210. 
 
 Rainbow, 202; lunar, 203. 
 
 Rain, 203-211; compared with snow, 213. 
 
 Rainless Regions, 98. 
 
 Rainy Days, nixmber of, 206. 
 
 Ramsay on Lakes, 29, 169, 175. 
 
 Rapids, 150. 
 
 Red Sea, 102; temperatirre of, 115; cur- 
 rc.ifs, 135; winds of, 267. 
 
 Refraction of Light, 203, 251; of Sound, 
 256. 
 
 Regelation, theory of, 217. 
 
 Rennel's Current, 129. 
 
 Resemblances, pvoteciive and iudeiiend- 
 ent, 334. 
 
 Rhone, Delta of, 97. 
 
 River Gravels, Human Remains in, 356. 
 
 Rivers, Origin of, 71, 143; courses affected 
 by disturbance, 143; motions of, 144; 
 speed of, 115; slope of channel. 145; 
 limit of denudation by, 147; sedi- 
 ments in, 147; divisions of course, 
 149; deltas of, 95-97, 150; portage of, 
 152; American, 153-155; Pacific slope, 
 164; European, 156-159; of British 
 Islands, 157; African, 159-161; of 
 Asia Minor, 161; table of principal, 
 163; Asia, Eastern, 163; Australian, 
 164; continental, or of inland drain- 
 age, 165 ; underground, 190 ; con- 
 trasted with glacier, 223. 
 
 Roaring Forties, 263. 
 
 Roches moutonnees, 231. ( 
 
 Rocks, Grouping of, 24 ; mechanically 
 formed, 26 - 28 ; sedimentary, 26 ; 
 ajolian, 28; Subaerial, 28; chemically 
 formed, 28; calcareous, 28; siliceous 
 sinters, 29; salt, 29; organic, 29; cal- 
 careous, of organic origin, 29; carbon- 
 aceous, composition of. Dr. Percy's 
 table, 31; hypogene, 33; porosity of, 
 180. 
 
 Rocky Mountains, 84, 215. 
 
 Rotation of Earth, Influence on Currents, 
 259. 
 
 Sahaea, 92; springs in, 182. 
 
 Salt, Rock, 29; proportion of in water, 
 110; in Dead Sea, 110; plains, 165; 
 lakes, 175. 
 
 Samiel, 278. 
 
 Sargasso Seas, 131, 132, 259. 
 
 Scandinavia, Moujitains of, 79. 
 
 Schists, varieties of, 33. 
 
 Sclater's Provinces, 326. 
 
 Seasons, N. and S. Hemisphere, 280. "l 
 
 Sections, true only for locality, 21 ; com- 
 parative in Britain, 21; Scotland, 22. 
 
 Sedimentary Strata, formation of, 35 ; 
 disturbances of, 39 ; curvature, 45 ; 
 unconformity, 46 ; faults, 46. 
 
 Sediments, their arrangement, 38 ; in 
 rivers, 147. 
 
 Seismic Movements, 308. 
 
 Selection, natural and artiQcial^ 3^3. 
 
INDEX. 
 
 367 
 
 Belvas, 9S. 
 
 Sferacs, 221. 
 
 Serpentine, o4. 
 
 Shales, oil, 32. 
 
 Shoals, 69. 
 
 Skull, Proportions of Human, 349; index, 
 349. 
 
 Siliceous Springs, 1S4;1 deposits of, 29, 
 190. 
 
 Silurians, unconformities in, 48, 52. 
 
 Simoom, 277. 
 
 Sirocco, 260. 
 
 Slate, 34. 
 
 Snow, 212; crystals, 212; flakes, 213; tex- 
 ture and colour, 213; compared with 
 rain, 213; limit of perpetual, 213; 
 converted into ice, 218. 
 
 Soils, saturation of, ISO; evaporation 
 from, 196. 
 
 Solstices, 2S6. 
 
 Sound, Analogies of, 248; vaves of, 249; 
 velocity of, 254 ; in diflerent sub- 
 stances, 255; intensity of, 255; refrac- 
 tion and reflection of, 256; resonance, 
 256. 
 
 Soundings, Deep Sea, 53; how taken, 103; 
 Atlantic, 104; Mediterz-anean, 100; 
 Indian Ocean, 107, 150; Pacific, 107. 
 
 South Connecting CuiTent, 132. 
 
 Spain, Mountains of, 79; springs of, 187. 
 
 Species, Origin by modification, 325; va- 
 riation of, 325; analogous or repre- 
 sentative, 330,335; migration of, 330, 
 335; results of, 332; influence of cli- 
 mate on, 322, 331; dLstribution in 
 space and time, 333; variations bene- 
 ficial, 334 ; mimicry, 334 ; resem- 
 blances of homogenetic and homo- 
 plastic, 334; pelagic, 341; extinction 
 and replacement, 342 ; persistent 
 types, 342; progressive development, 
 343; Barrandes colonies, 43, 343; 
 
 . . man, 346. 
 
 Specific Centres, 325. 
 
 Spectrum, 252; absorption bands, 252. 
 
 Springs, 178; origin of, 178; saturation of 
 soil by, ISO; conditions of their ex- 
 istence, 181; their yield, 183; of Sa- 
 hara, 182; mineral and thermal, 184; 
 siliceous, 184; calcareous, 184; tem- 
 perature of, 184; at Yellowstone, 185; 
 English thermal, 180; ferruginous, 
 ISO; intermittent, 187, 314; periodic, 
 187; of France, 187; of Spain, 187; 
 Dischof's classification of, 188; of 
 Auvergne, 188; petroleum, 190; rela- 
 tion to volcanoes, 313. 
 
 Steppes, 97; Pamir, 92. 
 
 Storms, 273; rotatory, 274; velocity and 
 area of, 276; tornadoes, 274, 277 ; ty- 
 phoons, 274; cjclones, 210, 274 ; pres- 
 sure during,274; waves, 276; dust, 277; 
 magnetic, 280; hail, 237; thunder, 
 
 Stratification, 35-41. 
 
 Stratus, 200. 
 
 Streams, 124; Gulf, 120; beneath glacier, 
 
 229. 
 Strike, 44. 
 
 Subsidence and Elevation, 315. 
 Subterranean 3*Iovements, 71, 315, 
 Surface Wash, 225. 
 Survival of Fittest, 333. 
 Syenite, 34. 
 Syncline, 45. 
 
 Tasmaxiax, 353. 
 
 Taurus, 80, 101, 
 
 Tehuantepecer, 266. 
 
 Temperature, undergi'ound, 15; of ocean, 
 112, 290 ; of Mediterranean, 114, 115; 
 of Atlantic, 114, 126, 128 ; of Indian 
 Ocean, 115; Red Sea, 115; and depth, 
 116; of Gulf Stream, 120; in Atlantic, 
 126; of springs, 184; and density of 
 freezing water, 215; of atmosphere, 
 240, 243, 245; decrease with height, 
 246; relation of, to latitude, 284 ; 
 Dove's law, 2S5. 
 
 Thermal Springs, 184. 
 
 Thibetan Plateau, 92, 161; people, 351. 
 
 Thomson, James, on Glacier Motion, 217. 
 
 Thomson, Sir W., on Geological Time, 
 16 ; on internal heat, 311. 
 
 Thunderbolts, 283. 
 
 Thunderstorms, 282. 
 
 Tides, 121; priming and lagging of, 122; 
 spring, 122; neap, 122; co-tidal lines, 
 124. 
 
 Till, 233. 
 
 Tornadoes, 274, 277. 
 
 Trade-winds, Atlantic,2C0; Indian Ocean, 
 206. • - 
 
 Tramontana, 261. 
 
 Trachyte, 35. ■ 
 
 Traps, contemporaneous, 49; intrusive, 
 49, 304. 
 
 Tundras, 99. 
 
 Turkestan, winds of, 270. 
 
 Twilight, 251. 
 
 Typhoons, 274. 
 
 Ulotrichous races, 3^0. 
 
 Unity and Plurality of Man, 346. 
 
 Valleys, formation of by rivers, 72 ; 
 transverse and longitudinal, 73, 139; 
 forms, 74; submerged, 75; of Eng- 
 land. 88, 139; antiquity of, 139; 
 obliterated, 150; of subsidence, 172. 
 
 Vapour, aqueous, 195; pressme ol^ 195, 
 240; condensation of, 196, 203; trans- 
 port of, 203; influence on tempera- 
 ture, 243. 
 
 Vegetation, influence on rainfall, 207 ; 
 influence on climate, 291. 
 
 Variation of Species, 325; beneficial, 334, 
 ■\'crsaut, 139, 
 
368 
 
 INDEX. 
 
 Vibrations, see Waves. 
 
 Volcanoes, 298; insular, 67, 69; materials 
 emitted from, 3t^ structure of cones, 
 299; ashes of, 300; Java, 299; dormant 
 and extinct, 304; distribution of, 304; 
 and earthq^uakes, causes of, 311 ; con- 
 nection of with ocean, 312; periodi- 
 city of, 315; influtftice of moon, 315. 
 
 Waste, see Denudation. 
 
 Water, cycle of, 100; contrasted with air, 
 100; proportion to land, 102; specific 
 gravity, 109; analysis of, 109, 174, 
 183; pressure of, 112; freezing point of, 
 112, 218; specific heat of, 113; colour 
 of, 116; transparency of, 117; move- 
 ments of, 118; functions of, 136; sub- 
 terranean, 178; in atmosphere, 194; 
 in clouds, 201; colour of, 116. 
 
 Waterfalls, 150. 
 
 Watershed or Waterparting, 139; q.ua- 
 quaversal, 139. 
 
 Waterspouts, 277. 
 
 Waves, Defioition of, 118; in water, 118; 
 force of, 120; height of, 120; raised 
 by wind, 120; of translation, 123; 
 flood, 145; of heat, light, and sound, 
 
 249; of ether, 250; storm, 276; earth- 
 quake, 308-310. 
 
 Wealden, 318. 
 
 Weather, 284 ; prognostics, 295 ; influ- 
 ence of moon on, 295. 
 
 Wells, Artesian, 181, 183 
 
 Whirlwinds, 277. 
 
 Winds, caiise of currents, 125 ; affecting 
 rainfall, 205, 207; North Atlantic, 
 260; trade, N. E., 260; S. E., 264; 
 Mediterranean, 261; Africa, 261, 269; 
 North America, 264; South America, 
 265; Indian Ocean, 266; Red Sea, 
 267; Turkestan, 270; land and sea, 
 272; velocity of. 272; table, 273; si- 
 moom, 277; samiel, 277. 
 
 Woolly-haired Races, 350. 
 
 Xakthochroic, 352. 
 Xanthomelanous races, 351. 
 
 Yellowstone valley springs, 185. 
 
 Zambesi, Falls of, 151; river, 160. 
 Zoology, Relation to Physical Geography 
 
 9. 
 Zoological Provinces, 322. 
 
 €31 
 
 WILLIAM COIiLINS & COMPANY, PRI5?TEKS, GLASGOW. 
 
1"^' 
 
 % 
 
 "ho" 
 
 
 v> - V ' « 
 
 
 vV c •^- 
 
 5> -U. 
 
 
 ^ 
 
 '^^ 
 
 -^^ 
 
 .-^ 
 
 
 ^ c " ^ '' * . '^O 
 
 ■^■,^* ". 
 
 v^ o 
 
 
 '^ 
 
 V. A^^' 
 
 
 
 
 .0- 
 
 .0 o 
 
 ^^'^' 
 
 nV '•>. 
 
LIBRARY OF CONGRESS 
 
 029 708 025 8