Chapter XIV: The Geographical Development of Coast-Lines 393 (16)

But in earlier stages of the world's history, the geographical conditions were not of a kind to favour the accumulation of vast ice-fields. During a period of extreme eccentricity, there would probably be fluctuations of temperature in high latitudes; but nothing like the glacial and interglacial epochs of the Pleistocene could have occurred. At most, there would be a general lowering of the temperature, sufficient to render the climate of arctic seas and lands somewhat cooler, and probably to induce the appearance in suitable places of local glaciers; and, owing to precession of the equinox, these cooler conditions would be followed by a general elevation of the temperature above the normal for the geographical conditions of the period. In Palæozoic and Mesozoic times, the effects of high eccentricity of the orbit appear to have been, in a great measure, neutralised by the geographical conditions, with a possible exception in the Permian period. But in Tertiary times, when the land-masses had become more continuous, the cosmical causes of change referred to must have had greater influence. And I cannot help agreeing with Dr. Croll that the warm climates of the Arctic regions during that era were, to some extent, the result of high eccentricity.

In concluding this discussion, I readily admit that our knowledge of geographical evolution is as yet in its infancy. We have still very much to learn, and no one will venture to dogmatise upon the subject. But I hope I have made it clear that the evidence, so far as it goes, does not justify the confident assertions of Dr. Croll's opponents, that his theory is contradicted by what we know of the climatic conditions of Palæozoic, Mesozoic, and Cainozoic times. On the contrary, it seems to me to gain additional support from the very evidence to which Nordenskiöld and others have appealed.

Note.--The accompanying sketch-maps (Plate IV.) require a few words
of explanation. The geology of the world is still so imperfectly
known that any attempt at graphic representation of former
geographical conditions cannot but be unsatisfactory. The approximate
positions of the chief areas of predominant elevation and depression
during stated periods of the past may have been ascertained in a
general way; but when we try to indicate these upon a map, such
provisional reconstructions are apt to suggest a more precise and
definite knowledge than is at present attainable. For it must be
confessed that there is hardly a line upon the small maps (A, B,
C) which might not have been drawn differently. This, of course,
is more especially true of South America, Africa, Asia--of large
areas of which the geological structure is unknown. But although the
boundaries of the land-masses shown upon the maps referred to are
thus confessedly provisional, the maps nevertheless bring out the
main fact of a gradual growth and consolidation of the land-areas--a
passage from insular to continental conditions. I need hardly say
this is no novel idea. It was clearly set forth by Professor Dana
upwards of forty years ago (_Silliman's Journal_, 1846, p. 352; 1847,
pp. 176, 381), and it received some years later further illustration
from Professor Guyot, who insisted upon the insular character of
the climate during Palæozoic times (_The Earth and Man_, 1850). It
must be understood that the maps (A, B, C) are not meant to exhibit
the geographical conditions of the world at any one point of time.
In Map A, for example, the area coloured blue was not necessarily
covered by sea at any particular stage in the Palæozoic era. It
simply represents approximately the regions tended. But, as already
stated, numerous oscillations of level occurred in Palæozoic times,
so that many changes in the distribution of land and water must have
taken place down to the close of the Permian period. The land-areas
shown upon the map are simply those which appear to have been more
or less persistent through all the geographical changes referred to.
Similar remarks apply to the other maps representing the more or
less persistent land-areas of Mesozoic and Tertiary times. Thus, for
example, there are reasons for believing that Madagascar was joined
to the mainland of Africa at some stage of the Mesozoic era, but was
subsequently insulated before Tertiary times. Again, as Mesozoic
era a land-connection obtained between New Zealand and Australia.
The same naturalist also points out that a chain of islands, now
represented by numerous islets and shoals, served in Tertiary times
to link Madagascar to India. Map D shows the areas of predominant
elevation and depression. The area coloured brown represents the
great continental plateau, which extends downwards to 1000 fathoms or
so below the present sea-level. The area tinted blue is the oceanic
depression. From the present distribution of plants and animals, we
infer that considerable tracts which are now submerged have formerly
been dry land--some of these changes having taken place in very
recent geological times. And the same conclusions are frequently
suggested by geological evidence. There can be little doubt that
Europe in Tertiary times extended further into the Northern Ocean
than it does now. And it is quite possible that in the Mesozoic and
Palæozoic eras considerable land-areas may likewise have appeared
here and there in those northern regions which are at present under
water. There is, indeed, hardly any portion of the continental
plateau which is now submerged that may not have been land at some
time or other. But after making all allowance for such possibilities,
the geological evidence, as far as it goes, nevertheless leads to
the conclusion that upon the whole a wider expense of primeval
continental plateau has come to the surface since Tertiary times than
was ever exposed during any former period of the world's history.

[Mr. Marcou states (_American Geologist_, 1890, p. 229) that the idea
of a gradual growth of land-areas originated with Elie de Beaumont,
who was in the habit of showing such maps, and used them in his
lectures at Paris as early as 1836. Professor Beudant published three
of these same maps for the Jurassic, Cretaceous, and Tertiary seas in
his _Cours élémentaire de Géologie_ (1841); and Professor Carl Vogt
in his _Lehrbuch der Geologie und Petrefactenkunde_ (1845), which
was confessedly based on Elie de Beaumont's lectures during 1844-46,
gives four maps of the Carboniferous, Jurassic, Cretaceous, and
Tertiary seas.]

XIII.

The Scientific Results of Dr. Nansen's Expedition.[DN]

[DN] From _The Scottish Geographical Magazine_, 1891.

In the Appendix to his most interesting and instructive work, _The First Crossing of Greenland_, Dr. Nansen treats of the scientific results of his remarkable journey. The detailed enumeration of these results, he tells us, would have been out of place in a general account of his expedition, but will appear in due time elsewhere. Hence he confines attention in his present work to such questions as are of most obvious interest, such as the extent, outward form, and elevation of the inland-ice of Greenland. By way of introduction his readers are presented with some account of the geological history of the country, which, although it contains nothing that was not already familiar to geologists, will doubtless prove interesting to others. After indicating that Greenland would appear to be composed almost exclusively of Archæan schists and granitoid eruptive rocks, the author glances at the evidence which the Mesozoic and Cainozoic strata of the west coast have supplied as to the former prevalence of genial climatic conditions. Heer is cited to show that during the formation of the Cretaceous beds the mean temperature of north Greenland was probably between 70° and 72° F., while in later Cainozoic times it could not have been less than 55° F., in 70° N.L. These conclusions are based on the character of the fossil floras. Now the mean annual temperature on the west coast of Greenland, where the relics of these old floras occur, is about 15° F., from which it is inferred that there has been a decrease of 40° since Cainozoic times. In those times, says Dr. Nansen, "the country must have rejoiced in a climate similar to that of Naples, while in the earlier Cretaceous period it must have resembled that of Egypt." He then refers to the well-known fact that, long after the deposition of the Cainozoic beds of Greenland, intensely arctic conditions supervened, when the inland-ice of that country extended much beyond its present limits. This was the Glacial period of geologists, during which all the northern regions of America and Europe, down to what are now temperate latitudes were likewise swathed in ice. Various hypotheses have been advanced in explanation of these strange climatic vicissitudes, and some of them are very briefly discussed by Dr. Nansen. None of the suggested solutions of the problem quite satisfies him; but he appears to look with most favour on the view that great climatic revolutions in what are now polar regions may have resulted from movements of the earth's axis. He admits, however, that there are certain strong objections to this hypothesis, and concludes that we have not yet got any satisfactory explanation to cover all the facts of the case. In discussing the question of a possible wandering of the pole, the author cites certain astronomical observations to show that the position of the axis is even now slowly changing, the movement amounting to half a second in six months. This is not much; but if the change, as he remarks, were to continue at the same rate for 3600 years, the shift would amount to one degree. Thus in a period of no more than 72,000 to 108,000 years Greenland might be brought into the latitude required for the growth of such floras as those of Cainozoic and Mesozoic times. Geologists will readily concede these or longer periods if they be required, but they will have graver doubts than Dr. Nansen as to whether any such great changes in the axis are possible. The astronomical observations referred to, even if they were fully confirmed, do not show that the movement is constant in one direction. They indicated, as he mentions, a slight increase of latitude during the first quarter of 1889, followed in the second quarter of the same year by a decrease, which continued to January, 1890. Since the publication of Professor George Darwin's masterly paper on the influence of geological changes on the earth's axis of rotation, geologists have felt assured that the great climatic revolutions to which the stratified rocks bear witness must be otherwise explained than by a wandering of the pole. Indeed, the geological evidence alone is enough to show that profound climatic changes have taken place while the pole has occupied its present position. Thus, there is no reasonable grounds for doubting that during the Glacial period the pole was just where we find it to-day. For, under existing geographical conditions, could a sufficient lowering of temperature be brought about, snow-fields and ice-sheets would gather and increase over the very same areas as we know were glaciated in Pleistocene times. Still further, we have only to recall the fact that several extreme revolutions of climate supervened during the so-called Glacial period, to see how impossible it is to account for the phenomena by movements of the earth's axis.

If it be true that the great climatic changes of the Pleistocene period did not result from a wandering to and fro of the pole, then it is not at all likely that the Mesozoic and Cainozoic climates of Greenland were induced by any such movement. But does the geological evidence justify us in believing that the climates in Greenland during Cretaceous and Tertiary times really resembled those of Egypt and southern Italy? It may be strongly doubted if it does. Palæontologists, like other mortals, find it hard to escape the influence of environment. They are apt to project the actual present into the past, without, perhaps, fully considering how far they are justified in doing so. Because there occur in Cretaceous and Tertiary strata, within Arctic regions, certain assemblages of plants which find their nearest representatives in southern Italy and Egypt, surely it is rather rash to conclude that Greenland has experienced climates like those now characteristic of Mediterranean lands. All that the evidence really entitles us to assume is simply that the _winter temperature_ of Greenland was formerly much higher than it is now. That great caution is required in comparing past with present climatological conditions may be seen by glancing for a moment at the character of the flora which lived in Europe during the interglacial phase of the Pleistocene period. The plants of that period are for the most part living species, so that while dealing with these we are on safer ground than when we are treating of the floras of periods so far removed from us as those of Tertiary and Cretaceous times. Now, in the Pleistocene flora of Europe we find a strange commingling of species, such as we nowhere see to-day over any equally wide area of the earth's surface. During Pleistocene times many plants which are still indigenous to southern France flourished side by side in that area with species which are no longer seen in the same region; some of these last having retreated because unable to support the cold of winter, while others have retired to the mountains to escape the dryness of the summer. Similar evidence is forthcoming from the Pleistocene accumulations of Italy, northern France, and Germany. In a word, clement winters and relatively cool and humid summers permitted the wide diffusion and intimate association of plants which have now a very different distribution, temperate and southern species formerly flourishing together over vast areas of southern and central Europe. And similarly we find that during the same period the regions in question were tenanted by southern and temperate forms of animal life--elephants, rhinoceroses, and hippopotamuses, together with cervine, bovine, and other forms, not a few of which are still indigenous to our Continent--that ranged from the shores of the Mediterranean up to our own latitudes. We cannot doubt, indeed, that the present geographical distribution of plants and animals differs markedly from anything that has yet been disclosed by the researches of geologists. The climatic conditions of our day are exceptional as compared with those of earlier times, and the occurrence in Greenland of southern types of plants, therefore, does not justify us in concluding that climates like those of southern Italy and Egypt were ever characteristic of arctic regions. It is a low winter temperature rather than a want of great summer heat that restricts the range northward of southern floras. If Greenland could be divested of its inland-ice--if its winter temperature never fell below that of our own island--it would doubtless become clothed in time with an abundant temperate flora.

Judging from what is known of the various floras and faunas that have successively clothed and peopled the world, from Palæozoic down to the close of Cainozoic times, the general climatic conditions of the globe, prior to the Glacial period, would seem to have been prevalently insular rather than continental as they are now. The lands appear to have been formerly much less continuous, and ocean currents from southern latitudes had consequently freer access to high northern regions than is at present possible. In no other way can we account for the facts connected with the geographical distribution and extent of the fossiliferous formations. But are we to infer, from the occurrence of similar assemblages of marine organic remains in arctic, temperate, and tropical latitudes, that the shores of primeval Greenland were washed by waters as warm as those of the tropics? Surely not: an absence of very cold water in the far north is all that we seem justified in assuming. And so, in like manner, the presence in Greenland of fossil floras having the same general facies as those that occur in the corresponding strata of more southern latitudes, does not compel us to believe that conditions at all similar to what are now met with in warm-temperate and sub-tropical lands ever obtained in arctic regions. A relatively high winter temperature alone would permit the range northward of many tribes of plants which are now restricted to southern latitudes. Yet, under the most uniform insular climatic conditions that we can conceive of, there must always have been differences due to latitude--although such differences were never apparently so marked as they are now.

In order to appreciate the character of the climate which must have prevailed when the lands of the globe were much more interrupted and insular than at present, we have only to consider how greatly isothermal lines, even under existing continental conditions, are deflected by ocean-currents. In the North Atlantic, for example, the winter isotherm of 32° F. is deflected northward from the parallel of New York to that of Hammerfest--a displacement of at least 30° of latitude. The Arctic Sea now occupies a partially closed basin, into which only one considerable current enters from the south. But in earlier ages the case was otherwise, and there was often communication across what are now our continental areas. Instead of being girdled, as at present, by an almost continuous land-mass, the Arctic Sea seems to have formed with the circumjacent ocean one great archipelago. Thus freely open to the influx of southern currents, it is not difficult to believe that the seas of the far north might never be frozen, and that an "inland-ice" like that of Greenland would be impossible. The present cold summers of that country, as the late Dr. Croll has insisted, are due not so much to high latitude as to the presence of snow and ice. Could these be removed, the summers would be as warm at least as those of England. Now the occurrence in arctic regions of Palæozoic and Mesozoic marine faunas is strongly suggestive of the former presence there of genial waters having free communication with lower latitudes; and it is to the presence of these warm currents, flowing uninterruptedly through polar regions, that we would attribute the high winter temperature and uniform climate to which the fossil floras and faunas of Greenland bear testimony.

If these views be at all reasonable, it seems unnecessary to call to our aid hypothetical changes in the position of the earth's axis. It may be admitted, however, that the climate of the Arctic regions must have been from time to time more or less affected by those cosmical causes to which Croll has appealed. So long, however, as insular conditions prevailed, the changes induced by a great increase in the eccentricity of the earth's orbit would not necessarily be strongly marked. Dr. Nansen objects to Croll's well-known theory that "it cannot account for the recurrence of conditions so favourable as to explain the existence in Greenland of a climate comparable to what we now find in tropical regions." No doubt it cannot, but, as we maintain, there is no good reason for supposing that tropical or sub-tropical climates ever characterised any area within the Arctic Circle. The remarkable association in Europe, during so recent a period as the Pleistocene, of southern and temperate species of plants and animals, ought to warn us against taking the present distribution of life-forms as an exact type of the kind of distribution which characterised earlier ages. It is safe to say that were our present continental areas to become broken up into groups of larger and smaller islands, so as to allow of a much less impeded oceanic circulation, the resulting climatic conditions would offer the strongest contrast to the present. And as the lands of the globe were apparently in former times more insular than they are now, it is hazardous to compare the climates of the present with those of the past. It is reasonable to infer, from the occurrence in Greenland of fossil floras which find their nearest representatives in southern Europe and north Africa, that the winters of the far north were formerly mild and clement. But we cannot conclude, from the same evidence, that the Arctic summers were ever as hot as those of our present warm-temperate and sub-tropical zones.

But if the recent expedition has thrown no new light on the disputed question as to the cause of the high temperature which formerly prevailed in Greenland, it is needless to say that it has added considerably to our knowledge of the present physical conditions of that country. The view held by many that Greenland must be wrapped in ice has been amply justified, and we can now no longer doubt that the inland-ice covers the whole country from the 75th parallel southwards. A section of Greenland in the latitude at which it was crossed by Nansen and his comrades "gives an almost exact mathematical curve, approximating very closely to the arc of a circle described with a radius of about 6500 miles. The whole way across the surface coincides tolerably accurately with this arc, though it falls away somewhat abruptly at the coasts, and a little more abruptly on the east side than the west." Taking the observations of other Arctic travellers with his own, Nansen is led to the conclusion that "the surface of the inland-ice forms part of a remarkably regular cylinder, the radius of which nevertheless varies not a little at different latitudes, increasing markedly from the south, and consequently making the arc of the surface flatter and flatter as it advances northwards." He points out that this remarkable configuration must to a certain extent be independent of the form of the underlying land-surface, which, to judge from the character of the wild and mountainous coast-lands, probably resembles Norway in its general configuration--if, indeed it be not a group of mountainous islands. The buried interior of Greenland must in fact be a region of high mountains and deep valleys, all of which have totally disappeared under the enveloping _mer de glace_. It is obvious, as Dr. Nansen remarks, that the minor irregularities of the land "have had no influence whatever upon the form of the upper surface of the ice-sheet." That surface-form has simply been determined by the force of pressure--the quasi-viscous mass attaining its maximum thickness towards the central line of the country, where resistance to the movement due to pressure must necessarily have been greatest. Thus although the larger features of the ice-drowned land may have had some influence in determining the position of the ice-shed, it is not by any means certain that this central line coincides with the dominant ridge or watershed of the land itself. For, as Nansen reminds us, the ice-shed of the Scandinavian inland-ice of glacial times certainly lay about 100 miles to the east of the main water-parting of Norway and Sweden. Similar facts, we may add, have been noticed in connection with the old ice-sheets of Scotland and Ireland.

The greatest elevation attained by the expedition was 9000 feet. How deeply buried the dominating parts of the land-surface may be at that elevation one cannot tell. It is obvious, however, that the _mer de glace_ must be very unequal in thickness. According to Dr. Nansen the average elevation of the valleys in the interior cannot much exceed 2300 or 3300 feet, so that the ice lying above such depressions must have a thickness of 5700 to 6700 feet. It cannot, of course, lie so deeply over mountain-ridges. The eroding power of such a glacier-mass must be enormous, and Dr. Nansen does not doubt that the buried valleys of Greenland are being widened and deepened by the grinding of the great ice-streams that are ever advancing towards the sea.

The expedition met with no streams of surface-water on the inland-ice; indeed, the amount of superficial melting in the interior was quite insignificant. And yet, as is well known, many considerable streams and rivers flow out from underneath the inland-ice all the year round. It is obvious, therefore, that this water-supply does not come from superficial sources, as, according to Dr. Nansen, it is usually supposed to do. But surely it has long been recognised that such rivers as the Mary Minturn must be derived from sub-glacial melting. And the various causes to which our author attributes this melting have already frequently been pointed out. Earth-heat--the influence of pressure in lowering the melting-point of ice--and the friction induced by the movement of the ice itself have all long ago been recognised as factors tending to produce the sub-glacial water-drainage of an ice-sheet.

Dr. Nansen's speculations on the origin of the "drumlins" and "kames" of formerly glaciated areas will interest geologists, but are not so novel as he supposes. His description of what are known as "drumlins" is not quite correct. These long lenticular banks cannot be said to lie upon boulder-clay, but are merely a structural form of that accumulation. And it is hardly the case that geologists have "performed the most acrobatic feats" in trying to explain the origin of the banks in question. The usual explanation is that they have been formed underneath the ice as ground-moraine--the upper surface of which varies in configuration--being sometimes approximately even, as in broad mountain-valleys; at other times ridged and corrugated, as in open lowlands. And these modifications of surface are supposed to have resulted from the varying movement and pressure of the overlying ice-sheet. The drumlins, in fact, would appear to be analogous to the banks that accumulate in the beds of rivers. Many drumlins, indeed, are composed partly of solid rock and partly of boulder-clay, which would seem to have accumulated in the lee of the projecting rock, much in the same way as gravel and sand gather behind any large boulder in a stream-course. Dr. Nansen, apparently, to some extent confounds drumlins with "kames" and "åsar," of which certainly many strange and conflicting explanations have been hazarded. These, however, differ essentially from drumlins, for they consist exclusively, or almost exclusively, of water-worn and more or less water-assorted materials. And one widely-accepted view of their origin is that they have accumulated in tunnels underneath an ice-sheet. This is practically the same view as Dr. Nansen's. He thinks that when an ice-sheet has its under-surface furrowed by running water, the ground-moraine will tend to be pressed up into the river-channels. The water will, in this way, be compelled to hollow out the roof of its tunnel to a greater degree, and as the stream continues to work upwards the moraine will follow it, so as to partially fill the tunnel and form a ridge along the back of which the sub-glacial stream will run. The material forming the upper portion of the ridge will thus come to be composed mainly of water-worn and stratified detritus, derived from the erosion of the ground-moraine. This is an ingenious suggestion which may be of good service in some cases, but it is certainly inapplicable to most kames and åsar. If it were a complete explanation we ought to find these ridges consisting of an upper water-assorted portion and a lower unmodified morainic portion (boulder-clay). But this is not the case, for most kames consist entirely, from top to bottom, of water-assorted materials. They are found running across an even or gently-undulating surface of boulder-clay, and sometimes they rest not on boulder-clay but solid rock.

Dr. Nansen considers another geological question which has given rise to much controversy, and is still far from being settled--namely, whether the oscillations of level which have left such conspicuous traces in northern regions are in any way connected with the appearance and disappearance of great ice-sheets. Can a big ice-sheet push down the earth's crust by its weight? and does the crust rise again as the ice melts away? Could a thick ice-sheet exercise sufficient attraction upon the sea to cause it to rise upon the land, and thus explain the origin of some of the so-called raised beaches of this and other formerly glaciated lands? Can the weight of a great ice-sheet shift the earth's centre of gravity, and, if so, to what extent? Each of these questions has been answered in the affirmative and the negative by controversialists, and, until the geological evidence has been completely sifted, each, doubtless, will continue to be alternately affirmed and denied. All that need be pointed out here is that some of the movements which occurred during the Pleistocene period were on much too large a scale to be explicable by any of the hypotheses referred to.

XIV.

The Geographical Development of Coast-lines.[DO]

[DO] Presidential Address to the Geographical Section of the British Association, Edinburgh, 1892.

Amongst the many questions upon which of late years light has been thrown by deep-sea exploration and geological research, not the least interesting is that of the geographical development of coast-lines. How is the existing distribution of land and water to be accounted for? Are the revolutions in the relative position of land and sea, to which the geological record bears witness, due to movements of the earth's crust or of the hydrosphere? Why are coast-lines in some regions extremely regular, while elsewhere they are much indented? About 150 years ago the prevalent belief was that ancient sea-margins indicated a formerly higher ocean-level. Such was the view held by Celsius, who, from an examination of the coast-lands of Sweden, attributed the retreat of the sea to a gradual drying up of the latter. But this desiccation hypothesis was not accepted by Playfair, who thought it much more likely that the land had risen. It was not, however, until after Von Buch had visited Sweden (1806-1808), and published the results of his observations, that Playfair's suggestion received much consideration. Von Buch concluded that the apparent retreat of the sea was not due to a general depression of the ocean-level, but to elevation of the land--a conclusion which subsequently obtained the strong support of Lyell. The authority of these celebrated men gained for the elevation theory more or less complete assent, and for many years it has been the orthodox belief of geologists that the ancient sea-margins of Sweden and other lands have resulted from vertical movements of the crust. It has long been admitted, however, that highly-flexed and disturbed strata require some other explanation. Obviously such structures are the result of lateral compression and crumpling. Hence geologists have maintained that the mysterious subterranean forces have affected the crust in different ways. Mountain-ranges, they conceive, are ridged up by tangential thrusts and compression, while vast continental areas slowly rise and fall, with little or no disturbance of the strata. From this point of view it is the lithosphere that is unstable, all changes in the relative level of land and sea being due to crustal movements. Of late years, however, Trautschold and others have begun to doubt whether this theory is wholly true, and to maintain that the sea-level may have changed without reference to movements of the lithosphere. Thus Hilber has suggested that sinking of the sea-level may be due, in part at least, to absorption, while Schmick believes that the apparent elevation and depression of continental areas are really the results of grand secular movements of the ocean. The sea, according to him, periodically attains a high level in each hemisphere alternately, the waters being at present heaped up in the southern hemisphere. Professor Suess, again, believing that in equatorial regions the sea is, on the whole, gaining on the land, while in other latitudes the reverse would appear to be the case, points out this is in harmony with his view of a periodical flux and reflux of the ocean between the equator and the poles. He thinks we have no evidence of any vertical elevation affecting wide areas, and that the only movements of elevation that take place are those by which mountains are upheaved. The broad invasions and transgressions of the continental areas by the sea, which we know have occurred again and again, are attributed by him to secular movements of the hydrosphere itself.

Apart from all hypothesis and theory, we learn that the surface of the sea is not exactly spheroidal. It reaches a higher level on the borders of the continents than in mid-ocean, and it varies likewise in height at different places on the same coast. The attraction of the Himalaya, for example, suffices to cause a difference of 300 feet between the level of the sea at the delta of the Indus and on the coast of Ceylon. The recognition of such facts has led Penck to suggest that the submergence of the maritime regions of north-west Europe and the opposite coasts of North America, which took place at a recent geological date, and from which the lands in question have only partially recovered, may have been brought about by the attraction exerted by the vast ice-sheets of the Glacial period. But, as Drygalski, Woodward, and others have shown, the heights at which recent marine deposits occur in the regions referred to are much too great to be accounted for by any possible distortion of the hydrosphere. The late James Croll had previously endeavoured to show that the accumulation of ice over northern lands during glacial times would suffice to displace the earth's centre of gravity, and thus cause the sea to rise upon the glaciated tracts. More recently other views have been advanced to explain the apparently causal connection between glaciation and submergence, but these need not be considered here.

Whatever degree of importance may attach to the various hypotheses of secular movements of the sea, it is obvious that the general trends of the world's coast-lines are determined in the first place by the position of the dominant wrinkles of the lithosphere. Even if we concede that all "raised beaches," so-called, are not necessarily the result of earth-movements, and that the frequent transgressions of the continental areas by oceanic waters in geological times may possibly have been due to independent movements of the sea, still we must admit that the solid crust of the globe has always been subject to distortion. And this being so, we cannot doubt that the general trends of the world's coast-lines must have been modified from time to time by movements of the lithosphere.

As geographers we are not immediately concerned with the mode of origin of those vast wrinkles, nor need we speculate on the causes which may have determined their direction. It seems, however, to be the general opinion that the configuration of the lithosphere is due simply to the sinking-in and doubling-up of the crust on the cooling and contracting nucleus. But it must be admitted that neither physicists nor geologists are prepared with a satisfactory hypothesis to account for the prominent trends of the great world-ridges and troughs. According to the late Professor Alexander Winchell, these trends may have been the result of primitive tidal action. He was of opinion that the transmeridional progress of the tidal swell in early incrustive times on our planet would give the forming crust structural characteristics and aptitudes trending from north to south. The earliest wrinkles to come into existence, therefore, would be meridional or submeridional, and such, certainly, is the prevalent direction of the most conspicuous earth-features. There are many terrestrial trends, however, as Professor Winchell knew, which do not conform to the requirements of his hypothesis; but such transmeridional features, he thought, could generally be shown to be of later origin than the others. This is the only speculation, so far as I know, which attempts, perhaps not altogether unsuccessfully, to explain the origin of the main trends of terrestrial features. According to other authorities, however, the area of the earth's crust occupied by the ocean is denser than that over which the continental regions are spread. The depressed denser part balances the lighter elevated portion. But why these regions of different densities should be so distributed no one has yet told us. Neither does Le Conte's view, that the continental areas and the oceanic depressions owe their origin to unequal radial contraction of the earth in its secular cooling, help us to understand why the larger features of the globe should be disposed as they are.

Geographers must for the present be content to take the world as they find it. What we do know is that our lands are distributed over the surface of a great continental plateau of irregular form, the bounding slopes of which plunge down more or less steeply into a vast oceanic depression. So far as geological research has gone, there is reason to believe that these elevated and depressed areas are of primeval antiquity--that they ante-date the very oldest of the sedimentary formations. There is abundant evidence, however, to show that the relatively elevated or continental area has been again and again irregularly submerged under tolerably deep and wide seas. But all historical geology seems to assure us that the continental plateau and the oceanic hollows have never changed places, although from time to time portions of the latter have been ridged up and added to the margins of the former, while ever and anon marginal portions of the plateau have sunk to very considerable depths. We may thus speak of the great world-ridges as regions of dominant elevation, and of the profound oceanic troughs as areas of more or less persistent depression. From one point of view, it is true, no part of the earth's surface can be looked upon as a region of dominant elevation. Our globe is a cooling and contracting body, and depression must always be the prevailing movement of the lithosphere. The elevation of the continental plateau is thus only relative. Could we conceive the crust throughout the deeper portions of the oceanic depression to subside to still greater depths, while at the same time the continental plateau remained stationary, or subsided more slowly, the sea would necessarily retreat from the land, and the latter would then appear to rise. It is improbable, however, that any extensive subsidence of the crust under the ocean could take place without accompanying disturbance of the continental plateau; and in this case the latter might experience in places not only negative but positive elevation. During the evolution of our continents, crustal movements have again and again disturbed the relative level of land and sea; but since the general result has been to increase the land-surface and to contract the area occupied by the sea, it is convenient to speak of the former as the region of dominant elevation, and of the latter as that of prevalent depression. Properly speaking, both are sinking regions, the rate of subsidence within the oceanic trough being in excess of that experienced over the continental plateau. The question of the geographical development of coast-lines is therefore only that of the dry lands themselves.

The greater land-masses are all situated upon, but are nowhere co-extensive with, the area of dominant elevation, for very considerable portions of the continental plateau are still covered by the sea. Opinions may differ as to which fathoms-line we should take as marking approximately the boundary between that region and the oceanic depression; and it is obvious, indeed, that any line selected must be arbitrary and more or less misleading, for it is quite certain that the true boundary of the continental plateau cannot lie parallel to the surface of the ocean. In some regions it approaches within a few hundreds of fathoms of the sea-level; in other places it sinks for considerably more than 1000 fathoms below that level. Thus, while a very moderate elevation would in certain latitudes cause the land to extend to the edge of the plateau, an elevation of at least 10,000 feet would be required in some other places to bring about a similar result.

Although it is true that the land-surface is nowhere co-extensive with the great plateau, yet the existing coast-lines may be said to trend in the same general direction as its margins. So abruptly does the continental plateau rise from the oceanic trough, that a depression of the sea-level, or an elevation of the plateau, for 10,000 feet, would add only a narrow belt to the Pacific coast between Alaska and Cape Horn, while the gain of land on the Atlantic slope of America between 30° N.L. and 40° S.L. would not be much greater. In the higher latitudes of the northern hemisphere, however, very considerable geographical changes would be accomplished by a much less amount of elevation of the plateau. Were the continental plateau to be upheaved for 3000 feet, the major portion of the Arctic Sea would become land. Thus, in general terms, we may say that the coast-lines of arctic and temperate North America and Eurasia are further withdrawn from the edge of the continental plateau than those of lower latitudes.

In regions where existing coast-lines approach the margin of the plateau, they are apt to run for long distances in one determinate direction, and, whether the coastal area be high or not, to show a gentle sinuosity. Their course is seldom interrupted by bold projecting headlands or peninsulas, or by intruding inlets, while fringing or marginal islands rarely occur. To these appearances the northern regions, as every one knows, offer the strongest contrast. Not only do they trend irregularly, but their continuity is constantly interrupted by promontories and peninsulas, by inlets and fiords, while fringing islands abound. But an elevation of some 400 or 500 fathoms only would revolutionise the geography of those regions, and confer upon the northern coast-lines of the world the regularity which at present characterises those of western Africa.

It is obvious, therefore, that the coast-lines of such lands as Africa owe their regularity primarily to their approximate coincidence with the steep boundary-slopes of the continental plateau, while the irregularities characteristic of the coast-line of north-western Europe and the corresponding latitudes of North America are determined by the superficial configuration of the same plateau, which in those regions is relatively more depressed. I have spoken of the general contrast between high and low northern latitudes; but it is needless to say that in southern regions the coast-lines exhibit similar contrasts. The regular coast-lines of Africa and South America have already been referred to; but we cannot fail to recognise in the much-indented sea-board and the numerous coastal islands of southern Chile a complete analogy to the fiord regions of high northern latitudes. Both are areas of comparatively recent depression. Again, the manifold irregularities of the coasts of south-eastern Asia, and the multitudes of islands that serve to link that continent to Australia and New Zealand, are all evidence that the surface of the continental plateau in those regions is extensively invaded by the sea.

A word or two now as to the configuration of the oceanic trough. There can be no doubt that this differs very considerably from that of the land-surface. It is, upon the whole, flat or gently-undulating. Here and there it swells gently upwards into broad elevated banks, some of which have been traced for great distances. In other places narrower ridges and abrupt mountain-like elevations diversify its surface, and project again and again above the level of the sea, to form the numerous islets of Oceania. Once more, the sounding-line has made us acquainted with the notable fact that numerous deep depressions--some long and narrow, others relatively short and broad--stud the floor of the great trough. I shall have occasion to refer again to these remarkable depressions, and need at present only call attention to the fact that they are especially well-developed in the region of the western Pacific, where the floor of the sea, at the base of the bounding slopes of the continental plateau, sinks in places to depths of three and even of five miles below the existing coast-lines. One may further note the fact that the deepest areas of the Atlantic are met with in like manner close to the walls of the plateau--a long ridge, which rises midway between the continents and runs in the same general direction as their coast-lines, serving to divide the trough of the Atlantic into two parallel hollows.

But, to return to our coast-lines and the question of their development, it is obvious that their general trends have been determined by crustal movements. Their regularity is in direct proportion to the closeness of their approach to the margin of the continental plateau. The more nearly they coincide with the edge of that plateau, the fewer irregularities do they present; the further they recede from it, the more highly are they indented. Various other factors, it is true, have played a more or less important part in their development, but their dominant trends were undoubtedly determined at a very early period in the world's history--their determination necessarily dates back, in short, to the time when the great world-ridges and oceanic troughs came into existence. So far as we can read the story told by the rocks, however, it would seem that in the earliest ages of which geology can speak with any confidence, the coast-lines of the world must have been infinitely more irregular than now. In Palæozoic times, relatively small areas of the continental plateau appeared above the level of the sea. Insular conditions everywhere prevailed. But as ages rolled on, wider and wider tracts of the plateau were exposed, and this notwithstanding many oscillations of level. So that one may say there has been, upon the whole, a general advance from insular to continental conditions. In other words, the sea has continued to retreat from the surface of the continental plateau. To account for this change, we must suppose that depression of the crust has been in excess within the oceanic area, and that now and again positive elevation of the continental plateau has taken place, more especially along its margins. That movements of elevation, positive or negative, have again and again affected our land-areas can be demonstrated, and it seems highly probable, therefore, that similar movements may have been experienced within the oceanic trough.

Two kinds of crustal movement, as we have seen, are recognised by geologists. Sometimes the crust appears to rise, or, as the case may be, to sink over wide regions, without much disturbance or tilting of strata, although these are now and again more or less extensively fractured and displaced. It may conduce to clearness if we speak of these movements as regional. The other kind of crustal disturbance takes place more markedly in linear directions, and is always accompanied by abrupt folding and mashing together of strata, along with more or less fracturing and displacement. The plateau of the Colorado has often been cited as a good example of regional elevation, where we have a wide area of approximately horizontal strata apparently uplifted without much rock-disturbance, while the Alps or any other chain of highly-flexed and convoluted strata will serve as an example of what we may term axial or linear uplifts. It must be understood that both regional and axial movements result from the same cause--the adjustment of the solid crust to the contracting nucleus--and that the term _elevation_, therefore, is only relative. Sometimes the sinking crust gets relief from the enormous lateral pressure to which it is subjected by crumpling up along lines of weakness, and then mountains of elevation are formed; at other times, the pressure is relieved by the formation of broader swellings, when wide areas become uplifted relatively to surrounding regions. Geologists, however, are beginning to doubt whether upheaval of the latter kind can affect a broad continental area. Probably, in most cases, the apparent elevation of continental regions is only negative. The land appears to have risen because the floor of the oceanic basin has become depressed. Even the smaller plateau-like elevations which occur within some continental regions may in a similar way owe their dominance to the sinking of contiguous regions.

In the geographical development of our land, movements of elevation and depression have played an important part. But we cannot ignore the work done by other agents of change. If the orographical features of the land everywhere attest the potency of plutonic agents, they no less forcibly assure us that the inequalities of surface resulting from such movements are universally modified by denudation and sedimentation. Elevated plains and mountains are gradually demolished, and the hollows and depressions of the great continental plateau become slowly filled with their detritus. Thus inland-seas tend to vanish, inlets and estuaries are silted up, and the land in places advances seaward. The energies of the sea, again, come in to aid those of rain and rivers, so that under the combined action of all the superficial agents of change, the irregularities of coast-lines become reduced, and, were no crustal movement to intervene, would eventually disappear. The work accomplished by those agents upon a coast-line is most conspicuous in regions where the surface of the continental plateau is occupied by comparatively shallow seas. Here full play is given to sedimentation and marine erosion, while the latter alone comes into prominence upon shores that are washed by deeper waters. When the coast-lines advance to the edge of the continental plateau, they naturally trend, as we have seen, for great distances in some particular direction. Should they preserve that position, undisturbed by crustal oscillation, for a prolonged period of time, they will eventually be cut back by the sea. In this way a shelf or terrace will be formed, narrow in some places, broader in others, according to the resistance offered by the varying character of the rocks. But no long inlets or fiords can result from such action. At most the harder and less readily demolished rocks will form headlands, while shallow bays will be scooped out of the more yielding masses. In short, between the narrower and broader parts of the eroded shelf or terrace a certain proportion will tend to be preserved. As the shelf is widened, sedimentation will become more and more effective, and in places may come to protect the land from further marine erosion. This action is especially conspicuous in tropical and sub-tropical regions, which are characterised by well-marked rainy seasons. In such regions immense quantities of sediment are washed down from the land to the sea, and tend to accumulate along shore, forming low alluvial flats. All long-established coast-lines thus acquire a characteristically sinuous form, and perhaps no better examples could be cited than those of western Africa.

To sum up, then, we may say that the chief agents concerned in the development of coast-lines are crustal movements, sedimentation, and marine erosion. All the main trends are the result of elevation and depression. Considerable geographical changes, however, have been brought about by the silting up of those shallow and sheltered seas which, in certain regions, overflow wide areas of the continental plateau. Throughout all the ages, indeed, epigene agents have striven to reduce the superficial inequalities of that plateau, by levelling heights and filling up depressions, and thus, as it were, flattening out the land-surface and causing it to extend. The erosive action of the sea, from our present point of view, is of comparatively little importance. It merely adds a few finishing touches to the work performed by the other agents of change.

A glance at the geographical evolution of our own Continent will render this sufficiently evident. Viewed in detail, the structure of Europe is exceedingly complicated, but there are certain leading features in its architecture which no profound analysis is required to detect. We note, in the first place, that highly-disturbed rocks of Archæan and Palæozoic age reach their greatest development along the north-western and western borders of our Continent, as in Scandinavia, the British Islands, north-west France, and the Iberian peninsula. Another belt of similarly disturbed strata of like age traverses central Europe from west to east, and is seen in the south of Ireland, Cornwall, north-west France, the Ardennes, the Thüringer-Wald, the Erz Gebirge, the Riesen Gebirge, the Böhmer-Wald, and other heights of middle and southern Germany. Strata of Mesozoic and Cainozoic age rest upon the older systems in such a way as to show that the latter had been much folded, fractured, and denuded before they came to be covered with younger formations. North and north-east of the central belt of ancient rocks just referred to, the sedimentary strata that extend to the shores of the Baltic and over a vast region in Russia, range in age from Palæozoic down to Cainozoic times, and are disposed for the most part in gentle undulations--they are either approximately horizontal or slightly inclined. Unlike the disturbed rocks of the maritime regions and of central Europe, they have obviously been subjected to comparatively little folding since the time of their deposition. To the south of the primitive back-bone of central Europe succeeds a region composed superficially of Mesozoic and Cainozoic strata for the most part, which, along with underlying Palæozoic and Archæan rocks, are often highly-flexed and ridged up, as in the chains of the Jura, the Alps, the Carpathians, etc. One may say, in general terms, that throughout the whole Mediterranean area Archæan and Palæozoic rocks appear at the surface only when they form the nuclei of mountains of elevation, into the composition of which rocks of younger age largely enter.