Island Life.

[1] A small number of species belonging to the West Indies are found in the extreme southern portion of the Florida Peninsula.

[2] I cannot avoid here referring to the enormous waste of labour and money with comparatively scanty and unimportant results to natural history of most of the great scientific voyages of the various civilized governments during the present century. All these expeditions combined have done far less than private collectors in making known the products of remote lands and islands. They have brought home fragmentary collections, made in widely scattered localities, and these have been usually described in huge folios or quartos, whose value is often in inverse proportion to their bulk and cost. The same species have been collected again and again, often described several times over under new names, and not unfrequently stated to be from places they never inhabited. The result of this wretched system is that the productions of some of the most frequently visited and most interesting islands on the globe are still very imperfectly known, while their native plants and animals are being yearly exterminated, and this is the case even with countries under the rule or protection of European governments. Such are the Sandwich Islands, Tahiti, the Marquesas, the Philippine Islands, and a host of smaller ones; while Bourbon and Mauritius, St. Helena, and several others, have only been adequately explored after an important portion of their productions has been destroyed by cultivation or the reckless introduction of goats and pigs. The employment in each of our possessions, and those of other European powers, of a resident naturalist at a very small annual expense, would have done more for the advancement of knowledge in this direction than all the expensive expeditions that have again and again circumnavigated the globe.

[3] The general facts of Palæontology, as bearing on the migrations of animal groups, are summarised in my Geographical Distribution of Animals, Vol. I. Chapters VI., VII., and VIII.

[4] Since these lines were written, a fine series of specimens of this rare humming-bird has been obtained from the same locality. (See Proc. Zool. Soc. 1881, pp. 827-834.)

[5] Many of these large genera are now subdivided, the divisions being sometimes termed genera, sometimes sub-genera.

[6] The Palæarctic region includes temperate Asia and Europe, as will be explained in the next chapter.

[7] The following list of the genera of reptiles and amphibia peculiar to the Palæarctic Region has been furnished me by Mr. G. A. Boulenger, of the British Museum:—

Snakes.	Frogs and Toads.
Achalinus—China, Japan. Cœlopeltis—S. Eur., N. Af., S.W. Asia. Macroprotodon—S. Eur., N. Af. Taphrometopon—Cent. Asia.	Pelobates—Eur., S.W. Asia. Pelodytes—W. Europe. Discoglossus—S. Eur., N.W. Af. Bombinator—Eur., Temp. Asia. Alytus—Cent. and W. Eur.
Lizards.	Newts.
Phrynocephalus—Cent. and S.W. Asia. Anguis—Europe, W. Asia. Blanus—S.W. Eur., N.W. Africa, S.W. Asia. Trogonophis—N.W. Africa. Lacerta—Eur., Temp. Asia, N. Africa (one sp. in W. Af.). Psammodromus—S.W. Eur., N.W. Africa. Algiroides—S. Eur.	Salamandra—Eur., N. Af., S.W. Asia. Chioglossa—Spain and Portugal. Salamandrina—Italy. Pachytriton—East Thibet. Hynobius—China and Japan. Geomolge—E. Manchuria. Onychodactylus—Japan. Salamandrella—Siberia. Ranidens—Siberia. Batrachyperus—East Thibet. Myalobatrachus—China, Japan. Proteus—Caverns of S. Austria.

[8] Remains of the dingo have been found fossil in Pleistocene deposits but the antiquity of man in Australia is not known. It is not, however, improbable that it may be as great as in Europe. My friend A. C. Swinton, Esq., while working in the then almost unknown gold-field of Maryborough, Victoria, in January, 1855, found a fragment of a well-formed stone axe resting on the metamorphic schistose bed-rock about five feet beneath the surface. It was overlain by the compact gravel drift called by the miners "cement," and by an included layer of hard iron-stained sandstone. The fragment is about an inch and three-eighths wide and the same length, and is of very hard fine-grained black basalt. One side is ground to a very smooth and regular surface, terminating in a well-formed cutting edge more than an inch long, the return face of the cutting part being about a quarter of an inch wide. The other side is a broken surface. The weapon appears to have been an axe or tomahawk closely resembling that figured at p. 335 of Lumholtz's Among Cannibals, from Central Queensland. The fragment was discovered by Mr. Swinton and the late Mr. Mackworth Shore, one of the discoverers of the gold-field, before any rush to it had taken place, and it seems impossible to avoid the conclusion that it was formed prior to the deposit of the gravel drift and iron-stained sandstone under which it lay. This would indicate a great antiquity of man in Australia, and would enable us to account for the fossilised remains of the dingo in Pleistocene deposits as those of an animal introduced by man.

[9] These facts are taken from a memoir on The Mammals and Winter Birds of Florida, by J. A. Allen; forming Vol. II., No. 3, of the Bulletin of the Museum of Comparative Zoology at Harvard College, Cambridge, Massachusetts.

[10] The great variation in wild animals is more fully discussed and illustrated in the author's Darwinism (Chapter III.).

[11] See Ibis, 1879, p. 32.

[12] In Mr. Seebohm's latest work, Birds of the Japanese Empire (1890), he says, "Examples from North China are indistinguishable from those obtained in Greece" (p. 82).

[13] Ibis, 1879, p. 40. In his Birds of the Japanese Empire (1890), Mr. Seebohm classes the Japanese and European forms as E. schœniclus, and thinks that their range is probably continuous across the two continents.

[14] Lyell's Principles of Geology, ii., p. 369.

[15] Mr. Darwin found that the large Helix pomatia lived after immersion in sea-water for twenty days. It is hardly likely that this is the extreme limit of their powers of endurance, but even this would allow of their being floated many hundred miles at a stretch, and if we suppose the shell to be partially protected in the crevice of a log of wood, and to be thus out of water in calm weather, the distance might extend to a thousand miles or more. The eggs of fresh-water mollusca, as well as the young animals, are known to attach themselves to the feet of aquatic birds, and this is probably the most efficient cause of their very wide diffusion.

[16] Principles of Geology, 11th Ed., Vol. I., p. 258.

[17] On Limestone as an Index of Geological Time.

[18] In his Preliminary Report on Oceanic Deposit, Mr. Murray says:—"It has been found that the deposits taking place near continents and islands have received their chief characteristics from the presence of the debris of adjacent lands. In some cases these deposits extend to a distance of over 150 miles from the coast." (Proceedings of the Royal Society, Vol. XXIV. p. 519.)

"The materials in suspension appear to be almost entirely deposited within 200 miles of the land." (Proceedings of the Royal Society of Edinburgh, 1876-77, p. 253.)

[19] Geographical Evolution. (Proceedings of the Royal Geographical Society. 1879, p. 426.)

[20] Professor Dana was, I believe, the first to point out that the regions which, after long undergoing subsidence and accumulating vast piles of sedimentary deposit have been elevated into mountain ranges, thereby become stiff and unyielding, and that the next depression and subsequent upheaval will be situated on one or the other sides of it; and he has shown that, in North America, this is the case with all the mountains of the successive geological formations. Thus, depressions, and elevations of extreme slowness but often of vast amount, have occurred successively in restricted adjacent areas; and the effect has been to bring each portion in succession beneath the ocean but always bordered on one or both sides by the remainder of the continent, from the denudation of which the deposits are formed which, on the subsequent upheaval, become mountain ranges. (Manual of Geology, 2nd Ed., p. 751.)

[21] Nature, Vol. II., p. 297.

[22] Sir W. Thomson, Voyage of Challenger, Vol. II., p. 374.

[23] The following is the analysis of the chalk at Oahu:—

Carbonate of Lime	92.800	per cent.
Carbonate of Magnesia	2.385	,,
Alumina	0.250	,,
Oxide of Iron	0.543	,,
Silica	0.750	,,
Phosphoric Acid and Fluorine	2.113	,,
Water and loss	1.148	,,

This chalk consists simply of comminuted corals and shells of the reef. It has been examined microscopically and found to be destitute of the minute organisms abounding in the chalk of England. (Geology of the United States Exploring Expedition, p. 150.) Mr. Guppy also found chalk-like coral limestones containing 95 p.c. of carbonate of lime in the Solomon Islands.

The absence of Globigerinæ is a local phenomenon. They are quite absent in the Arafura Sea, and no Globigerina-ooze was found in any of the enclosed seas of the Pacific, but with these exceptions the Globigerinæ "are really found all over the bottom of the ocean." (Murray on Oceanic Deposits—Proceedings of Royal Society, Vol. XXIV., p. 523.)

The above analysis shows a far closer resemblance to chalk than that of the Globigerina-ooze of the Atlantic, four specimens of which given by Sir W. Thomson (Voyage of the Challenger Vol. II. Appendix, pp. 374-376, Nos. 9, 10, 11 and 12) from the mid-Atlantic, show the following proportions:—

Carbonate of Lime	43.93		to	79.17	per cent.
Carbonate of Magnesia	1.40		to	2.58	,,
Alumina and Oxide of Iron	6.00	?	to	32.98	,,
Silica	4.60		to	11.23	,,

In addition to the above there is a quantity of insoluble residue consisting of small particles of sanidine, augite, hornblende, and magnetite, supposed to be the product of volcanic dust or ashes carried either in the air or by ocean currents. This volcanic matter amounts to from 4.60 to 8.33 per cent. of the Globigerina-ooze of the mid-Atlantic, where it seems to be always present; and the small proportion of similar matter in true chalk is another proof that its origin is different, and that it was deposited far more rapidly than the oceanic ooze.

The following analysis of chalk by Mr. D. Forbes will show the difference between the two formations:—

	Grey Chalk, Folkestone.	White Chalk, Shoreham.
Carbonate of Lime	94.09	98.40
Carbonate of Magnesia	0.31	0.08
Alumina and Phosphoric Acid	a trace	0.42
Chloride of Sodium	1.29	—
Insoluble débris	3.61	1.10

(From Quarterly Journal of the Geological Society, Vol. XXVII.)

The large proportion of carbonate of lime, and the very small quantity of silica, alumina, and insoluble débris, at once distinguish true chalk from the Globigerina-ooze of the deep ocean bed.

[24] Notes on Reticularian Rhizopoda; in Microscopical Journal, Vol. XIX., New Series, p. 84.

[25] Proceedings of the Royal Society, Vol. XXIV. p. 532.

[26] See Presidential Address in Sect. D. of British Association at Plymouth, 1877.

[27] Geological Magazine, 1871, p. 426.

[28] In his lecture on Geographical Evolution (which was published after the greater part of this chapter had been written) Sir Archibald Geikie expresses views in complete accordance with those here advocated. He says:—"The next long era, the Cretaceous, was more remarkable for slow accumulation of rock under the sea than for the formation of new land. During that time the Atlantic sent its waters across the whole of Europe and into Asia. But they were probably nowhere more than a few hundred feet deep over the site of our continent, even at their deepest part. Upon their bottom there gathered a vast mass of calcareous mud, composed in great part of foraminifera, corals, echinoderms, and molluscs. Our English chalk, which ranges across the north of France, Belgium, Denmark, and the north of Germany, represents a portion of the deposits of that sea-floor." The weighty authority of the Director-General of the Geological Survey may perhaps cause some geologists to modify their views as to the deep-sea origin of chalk, who would have treated any arguments advanced by myself as not worthy of consideration.

[29] Introduction and Succession of Vertebrate Life in America, by Professor O. C. Marsh. Reprinted from the Popular Science Monthly, March, April, 1878.

[30] Physical Geography and Geology of Great Britain, 5th Ed. p. 61.

[31] Of late it has been the custom to quote the so-called "ridge" down the centre of the Atlantic as indicating an extensive ancient land. Even Professor Judd at one time adopted this view, speaking of the great belt of Tertiary volcanoes "which extended through Greenland, Iceland, the Faroe Islands, the Hebrides, Ireland, Central France, the Iberian Peninsula, the Azores, Madeira, Canaries, Cape de Verde Islands, Ascension, St. Helena, and Tristan d'Acunha, and which constituted as shown by the recent soundings of H.M.S. Challenger a mountain-range, comparable in its extent, elevation, and volcanic character with the Andes of South America" (Geological Mag. 1874, p. 71). On examining the diagram of the Atlantic Ocean in the Challenger Reports, No. 7, a considerable part of this ridge is found to be more than 1,900 fathoms deep, while the portion called the "Connecting Ridge" seems to be due in part to the deposits carried out by the River Amazon. In the neighbourhood of the Azores, St. Paul's Rocks, Ascension, and Tristan d'Acunha are considerable areas varying from 1,200 to 1,500 fathoms deep, while the rest of the ridge is usually 1,800 or 1,900 fathoms. The shallower water is no doubt due to volcanic upheaval and the accumulation of volcanic ejections, and there may be many other deeply submerged old volcanoes on the ridge; but that it ever formed a chain of mountains "comparable in elevation with the Andes," there seems not a particle of evidence to prove. It is however probable that this ridge indicates the former existence of some considerable Atlantic islands, which may serve to explain the presence of a few identical genera, and even species of plants and insects in Africa and South America, while the main body of the fauna and flora of these two continents remains radically distinct.

In my Darwinism (pp. 344-5) I have given an additional argument founded on the comparative height and area of land with the depth and area of ocean, which seems to me to add considerably to the weight of the evidence here submitted for the permanence of oceanic and continental areas.

[32] In a review of Mr. T. Mellard Reade's Chemical Denudation and Geological Time, in Nature (Oct. 2nd, 1879), the writer remarks as follows:—"One of the funny notions of some scientific thinkers meets with no favour from Mr. Reade, whose geological knowledge is practical as well as theoretical. They consider that because the older rocks contain nothing like the present red clays, &c., of the ocean floor, that the oceans have always been in their present positions. Mr. Reade points out that the first proposition is not yet proved, and the distribution of animals and plants and the fact that the bulk of the strata on land are of marine origin are opposed to the hypothesis." We must leave it to our readers to decide whether the "notion" developed in this chapter is "funny," or whether such hasty and superficial arguments as those here quoted from a "practical geologist" have any value as against the different classes of facts, all pointing to an opposite conclusion, which have now been briefly laid before them, supported as they are by the expressed opinion of so weighty an authority as Sir Archibald Geikie, who, in the lecture already quoted says:—"From all this evidence we may legitimately conclude that the present land of the globe, though formed in great measure of marine formations, has never lain under the deep sea; but that its site must always have been near land. Even its thick marine limestones are the deposits of comparatively shallow water."

[33] Antiquity of Man, 4th Ed. pp. 340-348.

[34] The Great Ice Age and its Relation to the Antiquity of Man. By James Geikie, F.R.S. (Isbister and Co., 1874.)

[35] This view of the formation of "till" is that adopted, by Dr. Geikie, and upheld by almost all the Scotch, Swiss, and Scandinavian geologists. The objection however is made by many eminent English geologists, including the late Mr. Searles V. Wood, Jun., that mud ground off the rocks cannot remain beneath the ice, forming sheets of great thickness, because the glacier cannot at the same time grind down solid rock and yet pass over the surface of soft mud and loose stones. But this difficulty will disappear if we consider the numerous fluctuations in the glacier with increasing size, and the additions it must have been constantly receiving as the ice from one valley after another joined together, and at last produced an ice-sheet covering the whole country. The grinding power is the motion and pressure of the ice, and the pressure will depend on its thickness. Now the points of maximum thickness must have often changed their positions, and the result would be that the matter ground out in one place would be forced into another place where the pressure was less. If there were no lateral escape for the mud, it would necessarily support the ice over it just as a water-bed supports the person lying on it; and when there was little drainage water, and the ice extended, say, twenty miles in every direction from a given part of a valley where the ice was of less than the average thickness, the mud would necessarily accumulate at this part simply because there was no escape for it. Whenever the pressure all round any area was greater than the pressure on that area, the débris of the surrounding parts would be forced into it, and would even raise up the ice to give it room. This is a necessary result of hydrostatic pressure. During this process the superfluous water would no doubt escape through fissures or pores of the ice, and would leave the mud and stones in that excessively compressed and tenacious condition in which the "till" is found. The unequal thickness and pressure of the ice above referred to would be a necessary consequence of the inequalities in the valleys, now narrowing into gorges, now opening out into wide plains, and again narrowed lower down; and it is just in these openings in the valleys that the "till" is said to be found, and also in the lowlands where an ice-sheet must have extended for many miles in every direction. In these lowland valleys the "till" is both thickest and most wide-spread, and this is what we might expect. At first, when the glaciers from the mountains pushed out into these valleys, they would grind out the surface beneath them into hollows, and the drainage-water would carry away the débris. But when they spread all over the surface from sea to sea, and there was little or no drainage water compared to the enormous area covered with ice, the great bulk of the débris must have gathered under the ice wherever the pressure was least, and the ice would necessarily rise as it accumulated. Some of the mud would no doubt be forced out along lines of least resistance to the sea, but the friction of the stone-charged "till" would be so enormous that it would be impossible for any large part of it to be disposed of in this way.

[36] That the ice-sheet was continuous from Scotland to Ireland is proved by the glacial phenomena in the Isle of Man, where "till" similar to that in Scotland abounds, and rocks are found in it which must have come from Cumberland and Scotland, as well as from the north of Ireland. This would show that glaciers from each of these districts reached the Isle of Man, where they met and flowed southwards down the Irish Sea. Ice-marks are traced over the tops of the mountains which are nearly 2,000 feet high. (See A Sketch of the Geology of the Isle of Man, by John Horne, F.G.S. Trans. of the Edin. Geol. Soc. Vol. II. pt. 3, 1874.)

[37] The Great Ice Age, p. 177.

[38] These are named, in descending order, Hessle Boulder Clay, Purple Boulder Clay, Chalky Boulder Clay, and Lower Boulder Clay—below which is the Norwich Crag.

[39] "On the Climate of the Post-Glacial Period." Geological Magazine, 1872, pp. 158, 160.

[40] Geological Magazine, 1876, p. 396.

[41] Early Man in Britain and his Place in the Tertiary Period, p. 113.

[42] Heer's Primæval World of Switzerland Vol. II., pp. 148-168.

[43] Dr. James Geikie in Geological Magazine, 1878, p. 77.

[44] This subject is admirably discussed in Professor Asa Gray's Lecture on "Forest Geography and Archæology" in the American Journal of Science and Arts, Vol. XVI. 1878.

[45] In a letter to Nature of October 30th, 1879, the Rev. O. Fisher calls attention to a result arrived at by Pouillet, that the temperature which the surface of the ground would assume if the sun were extinguished would be -128° F. instead of -239° F. If this corrected amount were used in our calculations, the January temperature of England during the glacial epoch would come out 17° F., and this Mr. Fisher thinks not low enough to cause any extreme difference from the present climate. In this opinion, however, I cannot agree with him. On the contrary, it would, I think, be a relief to the theory were the amounts of decrease of temperature in winter and increase in summer rendered more moderate, since according to the usual calculation (which I have adopted) the differences are unnecessarily great. I cannot therefore think that this modification of the temperatures, should it be ultimately proved to be correct (which is altogether denied by Dr. Croll), would be any serious objection to the adoption of Dr. Croll's theory of the Astronomical and Physical causes of the Glacial Epoch.

The reason of the theoretical increase of summer heat being greater than the decrease of winter cold is because we are now nearest the sun in winter and farthest in summer, whereas we calculate the temperatures of the glacial epoch for the phase of precession when the aphelion was in winter. A large part of the increase of temperature would no doubt be used up in melting ice and evaporating water, so that there would be a much less increase of sensible heat; while only a portion of the theoretical lowering of temperature in winter would be actually produced owing to equalising effect of winds and currents, and the storing up of heat by the earth and ocean.

[46] Dr. Croll says this "is one of the most widespread and fundamental errors within the whole range of geological climatology." The temperature of the snow itself is, he says, one of the main factors. (Climate and Cosmology, p. 85.) But surely the temperature of the snow must depend on the temperature of the air through which it falls.

[47] In an account of Prof. Nordenskjöld's recent expedition round the northern coast of Asia, given in Nature, November 20th, 1879, we have the following passage, fully supporting the statement in the text. "Along the whole coast, from the White Sea to Behring's Straits, no glacier was seen. During autumn the Siberian coast is nearly free of ice and snow. There are no mountains covered all the year round with snow, although some of them rise to a height of more than 2,000 feet." It must be remembered that the north coast of Eastern Siberia is in the area of supposed greatest winter cold on the globe.

[48] Dr. Croll objects to this argument on the ground that Greenland and the Antarctic continent are probably lowlands or groups of islands. (Climate and Cosmology, Chap. V.)

[49] "On the Glacial Epoch," by James Croll. Geol. Mag. July, August, 1874.

[50] "The general absence of recent marks of glacial action in Eastern Europe is well known; and the series of changes which have been so well traced and described by Prof. Szabó as occurring in those districts seems to leave no room for those periodical extensions of 'ice-caps' with which some authors in this country have amused themselves and their readers. Mr. Campbell, whose ability to recognise the physical evidence of glaciers will scarcely be questioned, finds quite the same absence of the proof of extensive ice-action in North America, westward of the meridian of Chicago." (Prof. J. W. Judd in Geol. Mag. 1876, p. 535.)

The same author notes the diminution of marks of ice-action on going eastward in the Alps; and the Altai Mountains far in Central Asia show no signs of having been largely glaciated. West of the Rocky Mountains, however, in the Sierra Nevada and the coast ranges further north, signs of extensive old glaciers again appear; all which phenomena are strikingly in accordance with the theory here advocated, of the absolute dependence of glaciation on abundant rainfall and elevated snow-condensers and accumulators.

[51] I have somewhat modified this whole passage in the endeavour to represent more accurately the difference between the views of Dr. Croll and Sir Charles Lyell.

[52] For numerous details and illustrations see the paper—"On Ocean Currents in Relation to the Physical Theory of Secular Changes of Climate"—in the Philosophical Magazine, 1870.

[53] See Darwin's Naturalist's Voyage Round the World, 2nd Edition, pp. 244-251.

[54] The influence of geographical changes on climate is now held by many geologists who oppose what they consider the extravagant hypotheses of Dr. Croll. Thus, Prof. Dana imputes the glacial epoch chiefly, if not wholly, to elevation of the land caused by the lateral pressure due to shrinking of the earth's crust that has caused all other elevations and depressions. He says: "Now, that elevation of the land over the higher latitudes which brought on the glacial era is a natural result of the same agency, and a natural, and almost necessary, counterpart of the coral-island subsidence which must have been then in progress. The accumulating, folding, solidification, and crystallisation of rocks attending all the rock-making and mountain-making through the Palæozoic, Mesozoic, and Cenozoic eras, had greatly stiffened the crust in these parts; and hence in after times, the continental movements resulting from the lateral pressure necessarily appeared over the more northern portions of the continent, where the accumulations and other changes had been relatively small. To the subsidence which followed the elevation the weight of the ice-cap may have contributed in some small degree. But the great balancing movements of the crust of the continental and oceanic areas then going forward must have had a greatly preponderating effect in the oscillating agency of all time—lateral pressure within the crust." (American Journal of Science and Arts, 3rd Series, Vol. IX. p. 318.)

"In the 2nd edition of his Manual of Geology, Professor Dana suggests elevation of Arctic lands sufficient to exclude the Gulf Stream, as a source of cold during glacial epochs. This, he thinks, would have made an epoch of cold at any era of the globe. A deep submergence of Behring's Strait, letting in the Pacific warm current to the polar area, would have produced a mild Arctic climate like that of the Miocene period. When the warm current was shut out from the polar area it would yet reach near to it, and bring with it that abundant moisture necessary for glaciation." (Manual of Geology, 2nd Edition, pp. 541-755, 756.)

[55] Dana's Manual of Geology, 2nd Edition, p. 540.

[56] Dr. Croll says that I here assume an impossible state of things. He maintains "that the change from the distant sun in winter, and near sun in summer to the near sun in winter and distant sun in summer, aided by the change in the physical causes which this would necessarily bring about, would certainly be sufficient to cause the snow and ice to disappear." (Climate and Cosmology, p. 106.) But I demur to his "necessarily." It is not the direct effect of the nearer sun in winter that is supposed to melt the snow and ice, but the "physical causes," such as absence of fogs and increase of warm equatorial currents. But the near sun in winter acting on an ice-clad surface would only increase the fogs and snow, while the currents could only change if a large portion of the ice were first melted, in which case they would no doubt be modified so as to cause a further melting of the ice. Dr. Croll says: "The warm and equable conditions of climate which would then prevail, and the enormous quantity of intertropical water carried into the Southern Ocean, would soon produce a melting of the ice." (Loc. cit. p. 111.) This seems to me to be assuming the very point at issue. He has himself shown that the presence of large quantities of ice prevents "a warm and equable climate" however great may be the sun-heat; the ice therefore would not be melted, and there would be no increased flow of intertropical water to the Southern Ocean. The ocean currents are mainly due to the difference of temperature of the polar and equatorial areas combined with the peculiar form and position of the continents, and some one or more of these factors must be altered before the ocean currents towards the north pole can be increased. The only factor available is the Antarctic ice, and if this were largely increased, the northward-flowing currents might be so increased as to melt some of the Arctic ice. But the very same argument applies to both poles. Without some geographical change the Antarctic ice could not materially diminish during its winter in perihelion, nor increase to any important extent during the opposite phase. We therefore seem to have no available agency by which to get rid of the ice over a glaciated hemisphere, so long as the geographical conditions remained unchanged and the excentricity continued high.

[57] In the Geological Magazine, April, 1880, Mr. Searles V. Wood adduces what he considers to be the "conclusive objection" to Dr. Croll's excentricity theory, which is, that during the last glacial epoch Europe and North America were glaciated very much in proportion to their respective climates now, which are generally admitted to be due to the distribution of oceanic currents. But Dr. Croll admits his theory "to be baseless unless there was a complete diversion of the warm ocean currents from the hemisphere glaciated," in which case there ought to be no difference in the extent of glaciation in Europe and North America. Whether or not this is a correct statement of Dr. Croll's theory, the above objection certainly does not apply to the views here advocated; but as I also hold the "excentricity theory" in a modified form, it may be as well to show why it does not apply. In the first place I do not believe that the Gulf Stream was "completely diverted" during the glacial epoch, but that it was diminished in force, and (as described at p. 144) partly diverted southward. A portion of its influence would, however, still remain to cause a difference between the climates of the two sides of the Atlantic; and to this must be added two other causes—the far greater penetration of warm sea-water into the European than into the North American continent, and the proximity to America of the enormous ice-producing mass of Greenland. We have thus three distinct causes, all combining to produce a more severe winter climate on the west than on the east of the Atlantic during the glacial epoch, and though the first of these—the Gulf Stream—was not nearly so powerful as it is now, neither is the difference indicated by the ice-extension in the two countries so great as the present difference of winter-temperature, which is the essential point to be considered. The ice-sheet of the United States is usually supposed to have extended about ten, or, at most, twelve, degrees further south than it did in Western Europe, whereas we must go twenty degrees further south in the former country to obtain the same mean winter-temperature we find in the latter, as may be seen by examining any map of winter isothermals. This difference very fairly corresponds to the difference of conditions existing during the glacial epoch and the present time, so far as we are able to estimate them, and it certainly affords no grounds of objection to the theory by which the glaciation is here explained.

[58] Dr. Croll objects to this argument, and adduces the case of Greenland as showing that ice may accumulate far from sea. But the width of Greenland is small compared with that of the supposed Antarctic ice-cap. (Climate and Cosmology, p. 78.)

[59] The recent extensive glaciation of New Zealand is generally imputed by the local geologists to a greater elevation of the land; but I cannot help believing that the high phase of excentricity which caused our own glacial epoch was at all events an assisting cause. This is rendered more probable if taken in connection with the following very definite statement of glacial markings in South Africa. Captain Aylward in his Transvaal of To-day (p. 171) says:—"It will be interesting to geologists and others to learn that the entire country, from the summits of the Quathlamba to the junction of the Vaal and Orange rivers, shows marks of having been swept over, and that at no very distant period, by vast masses of ice from east to west. The striations are plainly visible, scarring the older rocks, and marking the hill-sides—getting lower and lower and less visible as, descending from the mountains, the kopjies (small hills) stand wider apart; but wherever the hills narrow towards each other, again showing how the vast ice-fields were checked, thrown up, and raised against their Eastern extremities."

This passage is evidently written by a person familiar with the phenomena of glaciation, and as Captain Aylward's preface is dated from Edinburgh, he has probably seen similar markings in Scotland. The country described consists of the most extensive and lofty plateau in South Africa, rising to a mountain knot with peaks more than 10,000 feet high, thus offering an appropriate area for the condensation of vapour and the accumulation of snow. At present, however, the mountains do not reach the snow-line, and there is no proof that they have been much higher in recent times, since the coast of Natal is now said to be rising. It is evident that no slight elevation would now lead to the accumulation of snow and ice in these mountains, situated as they are between 27° and 30° S. Lat.; since the Andes, which in 32° S. Lat. reach 23,300 feet high, and in 28° S. Lat. 20,000, with far more extensive plateaus, produce no ice-fields. We cannot, therefore, believe that a few thousand feet of additional elevation, even if it occurred so recently as indicated by the presence of striations, would have produced the remarkable amount of glaciation above described; while from the analogy of the northern hemisphere, we may well believe that it was mainly due to the same high excentricity that led to the glaciation of Western and Central Europe, and Eastern North America.

These observations confirm those of Mr. G. W. Stow, who, in a paper published in the Quarterly Journal of the Geological Society (Vol. XXVII. p. 539), describes similar phenomena in the same mountains, and also mounds and ridges of unstratified clay packed with angular boulders; while further south the Stormberg mountains are said to be similarly glaciated, with immense accumulations of morainic matter in all the valleys. We have here most of the surface phenomena characteristic of a glaciated country, only a few degrees south of the tropic; and taken in connection with the indications of recent glaciation in New Zealand, and those discovered by Dr. R. von Lendenfeld in the Australian Alps between 6,000 and 7,000 feet elevation (Nature, Vol. XXXII. p. 69), we can hardly doubt the occurrence of some general and wide-spread cause of glaciation in the southern hemisphere at a period so recent that the superficial phenomena are almost as well preserved as in Europe. Other geologists however deny that there are any distinct indications of glacial action in South Africa; but the recent discovery by Dr. J. W. Gregory, F.G.S., of the former extension of glaciers on Mount Kenya 5,000 feet below their present limits, renders probable the former glaciation of the South African Highlands.

[60] The astronomical facts connected with the motions and appearance of the planet are taken from a paper by Mr. Edward Carpenter, M.A., in the Geological Magazine of March, 1877, entitled, "Evidence Afforded by Mars on the Subject of Glacial Periods," but I arrive at somewhat different conclusions from those of the writer of the paper.

[61] In an article in Nature of Jan. 1, 1880, the Rev. T. W. Webb states that in 1877 the pole of Mars (? the south pole) was, according to Schiaparelli, entirely free of snow. He remarks also on the regular contour of the supposed snows of Mars as offering a great contrast to ours, and also the strongly marked dark border which has often been observed. On the whole Mr. Webb seems to be of opinion that there can be no really close resemblance between the physical condition of the Earth and Mars, and that any arguments founded on such supposed similarity are therefore untrustworthy.

[62] London, Edinburgh and Dublin Philosophical Magazine, Vol. XXXVI., pp. 144-150 (1868).

[63] Climate and Time in their Geological Relations, p. 341.

[64] Nature, Vol. XXI., p. 345, "The Interior of Greenland."

[65] Prof. J. W. Judd says: "In the case of the Alps I know of no glacial phenomena which are not capable of being explained, like those of New Zealand, by a great extension of the area of the tracts above the snow-line which would collect more ample supplies for the glaciers protruded into surrounding plains. And when we survey the grand panoramas of ridges, pinnacles, and peaks produced for the most part by sub-aërial action, we may well be prepared to admit that before the intervening ravines and valleys were excavated, the glaciers shed from the elevated plateaux must have been of vastly greater magnitude than at present." (Contributions to the Study of Volcanoes, Geological Magazine, 1876, p. 536.) Professor Judd applies these remarks to the last as well as to previous glacial periods in the Alps; but surely there has been no such extensive alteration and lowering of the surface of the country since the erratic blocks were deposited on the Jura and the great moraines formed in North Italy, as this theory would imply. We can hardly suppose wide areas to have been lowered thousands of feet by denudation, and yet have left other adjacent areas apparently untouched; and it is even very doubtful whether such an extension of the snow-fields would alone suffice for the effects which were certainly produced.

[66] Geological Magazine, 1876, p. 392.

[67] Colonel Fielden thinks that these trees have all been brought down by rivers, and have been stranded on shores which have been recently elevated. See Trans. of Norfolk Nat. Hist. Soc., Vol. III., 1880.

[68] Geological Magazine, 1876, "Geology of Spitzbergen," p. 267.

[69] The preceding account is mostly derived from Professor Heer's great work Flora Fossilis Arctica.

[70] Geological Magazine, 1875, p. 531.

[71] Geological Magazine, 1876, p. 266. In his recent work—Climate and Cosmology (pp. 164, 172)—the late Dr. Croll has appealed to the imperfection of the geological record as a reply to these arguments; in this case, as it appears to me, a very unsuccessful one.

[72] It is interesting to observe that the Cretaceous flora of the United States (that of the Dakota group), indicates a somewhat cooler climate than that of the following Eocene period. Mr. De Rance (in the geological appendix to Capt. Sir G. Nares's Narrative of a Voyage to the Polar Sea) remarks as follows: "In the overlying American Eocenes occur types of plants occurring in the European Miocenes and still living, proving the truth of Professor Lesquereux's postulate, that the plant types appear in America a stage in advance of their advent in Europe. These plants point to a far higher mean temperature than those of the Dakota group, to a dense atmosphere of vapour, and a luxuriance of ferns and palms." This is very important as adding further proof to the view that the climates of former periods are not due to any general refrigeration, but to causes which were subject to change and alternation in former ages as now.

[73] Mr. S. B. J. Skertchley informs me that he has himself observed thick Tertiary deposits, consisting of clays and anhydrous gypsum, at Berenice on the borders of Egypt and Nubia, at a height of about 600 feet above the sea-level; but these may have been of fresh-water origin.

[74] By referring to our map of the Indian Ocean showing the submarine banks indicating ancient islands (Chap. XIX.), it will be evident that the south-east trade-winds—then exceptionally powerful—would cause a vast body of water to enter the deep Arabian Sea.

[75] In his recently published Lectures on Physical Geography, Professor Haughton calculates, that more than half the solar heat of the torrid zone is carried to the temperate zones by ocean currents. The Gulf Stream itself carries one-twelfth of the total amount, but it is probable that a very small fraction of this quantity of heat reaches the polar seas owing to the wide area over which the current spreads in the North Atlantic. The corresponding stream of the Indian Ocean in Miocene times would have been fully equal to the Gulf Stream in heating power, while, owing to its being so much more concentrated, a large proportion of its heat may have reached the polar area. But the Arctic Ocean occupies less than one-tenth of the area of the tropical seas; so that, whatever proportion of the heat of the tropical zone was conveyed to it, would, by being concentrated into one-tenth of the surface, produce an enormously increased effect. Taking this into consideration, we can hardly doubt that the opening of a sufficient passage from the Indian Ocean to the Arctic seas would produce the effects above indicated.

[76] For an account of the resemblances and differences of the mammalia of the two continents during the Tertiary epoch, see my Geographical Distribution of Animals, Vol. I. pp. 140-156.

[77] Professor Haughton has made an elaborate calculation of the difference between existing climates and those of Miocene times, for all the places where a Miocene flora has been discovered, by means of the actual range of corresponding species and genera of plants. Although this method is open to the objection that the ranges of plants and animals are not determined by temperature only, yet the results may be approximately correct, and are very interesting. The following table which summarizes these results is taken from his Lectures on Physical Geography (p. 344):—

	Latitude.	Present Temperature.	Miocene Temperature.	Difference.
1. Switzerland	47°.00	53°.6 F	69°.8 F	16°.2 F
2. Dantzig	54°.21	45°.7 ,,	62°.6 ,,	16°.9 ,,
3. Iceland	65°.30	35°.6 ,,	48°.2 ,,	12°.6 ,,
4. Mackenzie River	65°.00	19°.4 ,,	48°.2 ,,	28°.8 ,,
5. Disco (Greenland)	70°.00	19°.6 ,,	55°.6 ,,	36°.0 ,,
6. Spitzbergen	78°.00	16°.5 ,,	51°.8 ,,	35°.3 ,,
7. Grinnell Land	81°.44	1°.7 ,,	42°.3 ,,	44°.0 ,,

It is interesting to note that Iceland, which is now exposed to the full influence of the Gulf Stream, was only 12°.6 F. warmer in Miocene times, while Mackenzie River, now totally removed from its influence was 28° warmer. This, as well as, the greater increase of temperature as we go northward and the polar area becomes more limited, is quite in accordance with the view of the causes which brought about the Miocene climate which is here advocated.

[78] The objection has been made, that the long polar night would of itself be fatal to the existence of such a luxuriant vegetation as we know to have existed as far as 80° N. Lat., and that there must have been some alteration of the position of the pole, or diminution of the obliquity of the ecliptic, to permit such plants as magnolias and large-leaved maples to flourish. But there appears to be really no valid grounds for such an objection. Not only are numbers of Alpine and Arctic evergreens deeply buried in the snow for many months without injury, but a variety of tropical and sub-tropical plants are preserved in the hot-houses of St. Petersburg and other northern cities, which are closely matted during winter, and are thus exposed to as much darkness as the night of the Arctic regions. We have besides no proof that any of the Arctic trees or large shrubs were evergreens, and the darkness would certainly not be prejudical to deciduous plants. With a suitable temperature there is nothing to prevent a luxuriant vegetation up to the pole, and the long continued day is known to be highly favourable to the development of foliage, which in the same species is larger and better developed in Norway than in the south of England.

[79] Geological Magazine, 1873, p. 320.

[80] Geological Magazine, 1877, p. 137.

[81] Manual of Geology, 2nd Ed. p. 525. See also letter in Nature, Vol. XXIII. p. 410.

[82] Nature, Vol. XVIII. (July, 1878), p. 268.

[83] "On the Comparative Value of certain Geological Ages considered as items of Geological Time." (Proceedings of the Royal Society, 1874, p. 334.)

[84] Trans. Royal Society of Edinburgh, Vol. XXIII. p. 161. Quarterly Journal of Science, 1877. (Croll on the "Probable Origin and Age of the Sun.")

[85] Philosophical Magazine, April, 1853.

[86] It has usually been the practice to take the amount of denudation in the Mississippi valley, or one foot in six thousand years, as a measure of the rate of denudation in Europe, from an idea apparently of being on the "safe side," and of not over-estimating the rate of change. But this appears to me a most unphilosophical mode of proceeding and unworthy of scientific inquiry. What should we think of astronomers if they always took the lowest estimates of planetary or stellar distances, instead of the mean results of observation, "in order to be on the safe side!"? As if error in one direction were any worse than error in another. Yet this is what geologists do systematically. Whenever any calculations are made involving the antiquity of man, it is those that give the lowest results that are always taken, for no reason apparently except that there was, for so long a time, a prejudice, both popular and scientific, against the great antiquity of man; and now that a means has been found of measuring the rate of denudation, they take the slowest rate instead of the mean rate, apparently only because there is now a scientific prejudice in favour of extremely slow geological change. I take the mean of the whole; and as this is almost exactly the same as the mean of the three great European rivers—the Rhone, Danube, and Po—I cannot believe that this will not be nearer the truth for Europe than taking one North American river as the standard.

[87] "On the Height of the Land and the Depth of the Ocean," in the Scottish Geographical Magazine, 1888.

[88] These figures are merely used to give an idea of the rate at which denudation is actually going on now; but if no elevatory forces were at work, the rate of denudation would certainly diminish as the mountains were lowered and the slope of the ground everywhere rendered flatter. This would follow not only from the diminished power of rain and rivers, but because the climate would become more uniform, the rainfall probably less, and no rocky peaks would be left to be fractured and broken up by the action of frosts. It is certain, however, that no continent has ever remained long subject to the influences of denudation alone, for, as we have seen in our sixth chapter, elevation and depression have always been going on in one part or other of the surface.

[89] The following statement of the depths at which the Palæozoic formations have been reached in various localities in and round London was given by Mr. H. B. Woodward in his address to the Norwich Geological Society in 1879:—

Deep Wells through the Tertiary and Cretaceous Formations.
Harwich	at	1,022	feet	reached	Carboniferous Rock.
Kentish Town	,,	1,114	,,	,,	Old Red Sandstone.
Tottenham Court Road	,,	1,064	,,	,,	Devonian.
Blackwall	,,	1,004	,,	,,	Devonian or Old Red Sandstone.
Ware	,,	800	,,	,,	Silurian (Wenlock Shale).

We thus find that over a wide area, extending from London to Ware and Harwich, the whole of the formations from the Oolite to the Permian are wanting, the Cretaceous resting on the Carboniferous or older Palæozoic rocks; and the same deficiency extends across to Belgium, where the Tertiary beds are found resting on Carboniferous at a depth of less than 400 feet.

[90] Geological Magazine, Vol. VIII., March, 1871.

[91] Mr. C. Lloyd Morgan has well illustrated this point by comparing the generally tilted-up strata denuded on their edges, to a library in which a fire had acted on the exposed edges of the books, destroying a great mass of literature but leaving a portion of each book in its place, which portion represents the thickness but not the size of the book. (Geological Magazine, 1878, p. 161.)

[92] Professor J. Young thinks it highly probable that—"the Lower Greensand is contemporaneous with part of the Chalk, so were parts of the Wealden; nay, even of the Purbeck a portion must have been forming while the Cretaceous sea was gradually deepening southward and westward." Yet these deposits are always arranged successively, and their several thicknesses added together to obtain the total thickness of the formations of the country. (See Presidential Address, Sect. C. British Association, 1876.)

[93] Mr. John Murray in his more careful estimate makes it about 51½ millions.

[94] As by far the larger portion of the denuded matter of the globe passes to the sea through comparatively few great rivers, the deposits must often be confined to very limited areas. Thus the denudation of the vast Mississippi basin must be almost all deposited in a limited portion of the Gulf of Mexico, that of the Nile within a small area of the Eastern Mediterranean, and that of the great rivers of China—the Hoang Ho and Yang-tse-kiang, in a small portion of the Eastern Sea. Enormous lengths of coast, like those of Western America and Eastern Africa, receive very scanty deposits; so that thirty miles in width along the whole of the coasts of the globe will probably give an area greater than that of the area of average deposit, and certainly greater than that of maximum deposit, which is the basis on which I have here made my estimates. In the case of the Mississippi, it is stated by Count Pourtales that along the plateau between the mouth of the river and the southern extremity of Florida for two hundred and fifty miles in width the bottom consists of clay with some sand and but few Rhizopods; but beyond this distance the soundings brought up either Rhizopod shells alone, or these mixed with coral sand, Nullipores, and other calcareous organisms (Dana's Manual of Geology, 2nd Ed. p. 671). It is probable, therefore, that a large proportion of the entire mass of sediment brought down by the Mississippi is deposited on the limited area above indicated.

Professor Dana further remarks: "Over interior oceanic basins as well as off a coast in quiet depths, fifteen or twenty fathoms and beyond, the deposits are mostly of fine silt, fitted for making fine argillaceous rocks, as shales or slates. When, however, the depth of the ocean falls off below a hundred fathoms, the deposition of silt in our existing oceans mostly ceases, unless in the case of a great bank along the border of a continent."

[95] From the same data Professor Haughton estimates a minimum of 200 million years for the duration of geological time; but he arrives at this conclusion by supposing the products of denudation to be uniformly spread over the whole sea-bottom instead of over a narrow belt near the coasts, a supposition entirely opposed to all the known facts, and which had been shown by Dr. Croll, five years previously, to be altogether erroneous. (See Nature, Vol. XVIII., p. 268, where Professor Haughton's paper is given as read before the Royal Society.)

[96] See Geological Magazine for 1877, p. 1.

[97] In his reply to Sir W. Thomson, Professor Huxley assumed one foot in a thousand years as a not improbable rate of deposition. The above estimate indicates a far higher rate; and this follows from the well-ascertained fact, that the area of deposition is many times smaller than the area of denudation.

[98] Dr. Croll and Sir Archibald Geikie have shown that marine denudation is very small in amount as compared with sub-aërial, since it acts only locally on the edge of the land, whereas the latter acts over every foot of the surface. Mr. W. T. Blanford argues that the difference is still greater in tropical than in temperate latitudes, and arrives at the conclusion that—"If over British India the effects of marine to those of fresh-water denudation in removing the rocks of the country be estimated at 1 to 100, I believe that the result of marine action will be greatly overstated" (Geology and Zoology of Abyssinia, p. 158, note). Now, as our estimate of the rate of sub-aërial denudation cannot pretend to any precise accuracy, we are justified in neglecting marine denudation altogether, especially as we have no method of estimating it for the whole earth with any approach to correctness.

[99] Agassiz appears to have been the first to suggest that the principal epochs of life extermination were epochs of cold; and Dana thinks that two at least such epochs may be recognised, at the close of the Palæozoic and of the Cretaceous periods—to which we may add the last glacial epoch.