The Origin of Continents and Oceans

Front matter

THE ORIGIN OF CONTINENTS AND OCEANS

BY ALFRED WEGENER PROFESSOR OF GEOPHYSICS AND METEOROLOGY AT THE UNIVERSITY OF GRAZ, AUSTRIA

TRANSLATED FROM THE THIRD GERMAN EDITION BY J. G. A. SKERL, M.Sc.

Chapter 1

THE ORIGIN OF CONTINENTS AND OCEANS

I. THE ESSENTIALS OF THE DISPLACEMENT THEORY

CHAPTER I

THE DISPLACEMENT THEORY

He who examines the opposite coasts of the South Atlantic Ocean must be somewhat struck by the similarity of the shapes of the coast-lines of Brazil and Africa. Not only does the great right-angled bend formed by the Brazilian coast at Cape San Roque find its exact counterpart in the re-entrant angle of the African coast-line near the Cameroons, but also, south of these two corresponding points, every projection on the Brazilian side corresponds to a similarly shaped bay in the African, and conversely each indentation in the Brazilian coast has a complementary protuberance on the African. Experiment with a compass on a globe shows that their dimensions agree accurately.

This phenomenon was the starting-point of a new conception of the nature of the earth’s crust and of the movements occurring therein; this new idea is called the theory of the displacement of continents, or, more shortly, the displacement theory, since its most prominent component is the assumption of great horizontal drifting movements which the continental blocks underwent in the course of geological time and which presumably continue even to-day.

According to this idea, to take a particular case, millions of years ago the South American continental plateau lay directly adjoining the African plateau, even forming with it one large connected mass. This first split in Cretaceous time into two parts, which then, like floating icebergs drifted farther and farther apart. Similarly, North America was close to Europe; and, at least from Newfoundland and Ireland northward, they formed with Greenland one connected block, which broke up by a forked rift near Greenland at the end of Tertiary time and farther north even in the Quaternary era; whereupon the constituent blocks moved apart from one another. The shelves, the portions of the continental masses overflowed by shallow seas, will always be considered in this book as parts of the blocks, the boundaries of which for great distances are not given by the coast-lines, but by the steep descent to the deep sea floor.

Similarly, it will be assumed that Antarctica, Australia and India lay adjoining South Africa, and with the latter and South America formed, until the beginning of the Jurassic period, a single large—even if partly submerged at times by shallow water—continental area, which in the course of Jurassic, Cretaceous and Tertiary time split and crumbled into smaller blocks which drifted away from each other in all directions. The three maps of the earth reproduced in Figs. 1 and 2 show these developments during the Upper Carboniferous, Eocene and Lower Quaternary periods. The case of India is somewhat different: it was originally connected by a long continental tract, mostly, it is true, covered by shallow seas, to the Asiatic continent. After, the separation of India from Australia on one side (in the Lower Jurassic) and from Madagascar on the other (during the transition from the Chalk to Tertiary) this long connecting portion was more and more folded together through the continuous gradual approach of India to Asia and constitutes to-day the mightiest mountain folds of the earth, the Himalayas and the numerous folded ranges of the high lands of Asia.

In other regions also the displacement of the blocks occurs in causal connection with the origin of the mountain systems. By the westward drift of the two Americas their anterior margin was folded together to form the mighty range of the Andes (which stretches from Alaska to Antarctica) as a result of the opposition of the ancient well-cooled and therefore resistant floor of the Pacific. A similar case is that of the Australian block, which includes New Guinea since it is only separated therefrom by a shelf. The recent high ranges of New Guinea occur on the side which is anterior with reference to the movement; as our map shows, this direction of movement was different before the break from Antarctica, for the present east coast was then the front side. Next the mountains of New Zealand lying immediately in front of this coast were folded, later becoming detached as festoons of islands by the altered direction of movement and then lagging behind. The present cordilleras of East Australia originated in a still older period; they were formed (at the same time as the more ancient folds, “Pre-cordilleras,” in South and North America, which are the foundations of the Andes) on the front margin of the continental masses, which were drifting as a whole before the separation.

Besides this westward wandering we also see to a large extent a striving towards the equator of the continental blocks. With this is connected the formation of the great Tertiary belt of folding stretching from the Himalayas to the Alps and Atlas mountains which were then in the equatorial zone.

The previously mentioned separation from the Australian block of the former coastal ranges of New Zealand, forming later a festoon of islands, leads us up to the phenomenon that smaller portions of the blocks are left behind by the westward wandering of the larger blocks. In this manner the marginal ranges of the East Asiatic continental coast separated as festoons of islands. The Lesser and Greater Antilles lag behind the movement of the Central American block, and similarly the so-called arc of the South Antilles between Patagonia and West Antarctica. Indeed all blocks tapering in a meridional direction show a curve of their points towards the east on account of this lag. The latter is well shown in the south point of Greenland, the submarine shelf of Florida, Tierra del Fuego and Graham Land, and the manner in which Ceylon has broken away from India.

It will easily be seen that this complete and extensive conception of the displacement theory must emanate from a definite acceptance of the relation of the oceans to the continental blocks. In fact, it is assumed that these two phenomena are fundamentally distinct, that the continental blocks with a thickness of about 100 km. swim in a magma out which they only project about 5 km. and which is uncovered in the floor of the oceans.

Thus the outermost lithosphere no longer completely covers the entire earth (whether it ever did can be left undecided), but has become smaller and smaller by continued folding and compression during the course of geological time, thereby increasing in thickness and splitting ultimately into more and more separated smaller continental blocks. The latter to-day cover but a quarter of the earth. The floors of the oceans form the free surface of the next layer of the body of the earth which is also assumed to exist under the continental blocks. The existence of this involves the geophysical side of the displacement theory.

The detailed establishment of this new hypothesis will form the major part of the book. Some historical remarks, however, should be given beforehand.

The first notion of the displacement of continents came to me in 1910 when, on studying the map of the world, I was impressed by the congruency of both sides of the Atlantic coasts, but I disregarded it at the time because I did not consider it probable. In the autumn of 1911 I became acquainted (through a collection of references, which came into my hands by accident) with the palæontological evidence of the former land connection between Brazil and Africa, of which I had not previously known. This induced me to undertake a hasty analysis of the results of research in this direction in the spheres of geology and palæontology, whereby such important confirmations were yielded, that I was convinced of the fundamental correctness of my idea. I first brought forward the idea on January 6th, 1912, in a lecture to the Geological Association of Frankfort-on-Main entitled “Die Herausbildung der Grossformen der Erdrinde (Kontinente und Ozeane) auf geophysikalischer Grundlage.” This lecture was followed on January 10th by a second on “Horizontalverschiebungen der Kontinente” to the Society for the Advancement of Science of Marburg. In the same year (1912), also, both of the first publications on the theory took place.

Afterwards the participation in the traverse of Green

under J. P. Koch of 1912/13 and later war-service hindered me from further elaboration of the theory. In 1915, however, I was able to use a long sick-leave to give a somewhat detailed description of the theory in the Vieweg series under the title of this book. As a second edition of this was necessary after the close of the war, the publishers generously consented to transfer the book from the Vieweg to the Wissenschaft series, whereby the possibility was given for a considerably enlarged work. The present edition is again virtually rewritten, as the process of the grouping of the data which affect the question according to the view-point of the new theory has meanwhile made further progress and an extensive recent literature about the subject has appeared.

During the above-mentioned work of examining the literature I several times chanced on views concordant with my own by older authors. Thus a rotation of the entire crust of the earth—but whose parts, however, did not alter their relative positions—had already been assumed by many authors, as Löffelholz von Colberg, Kreichgauer, Sir John Evans and others. H. Wettstein has written a remarkable book, in which, however, among many absurdities, a leaning towards great relative horizontal displacements of the continents is shown. The continents, (the submarine shelves of which however he did not consider) undergo, according to him, not only displacement but also deformation; they wander collectively westwards drawn by the tidal forces of the sun on the viscous body of the earth (as also E. H. L. Schwarz assumed in the Geogr. Journ., 1912, pp. 284–299). But the oceans were considered by him as sunken continents, and he expressed fantastic ideas about the so-called geographical homologies and other problems of the face of the earth, which we will pass over. Like the present writer, Pickering in a work on the similarities of the South Atlantic coasts has expressed the supposition that America was torn off from Europe-Africa and was dragged across the breadth of the Atlantic Ocean. But he did not consider that all the facts of the geological history of both these continents necessitated the assumption of an earlier connection up to the Cretaceous period, and thus he places the connection in the very remote past, and thought that the breaking away was connected with the theory of G. H. Darwin that the moon was thrown off the earth. Traces of this he believed are still to be seen in the Pacific basin.

F. B. Taylor approaches the sphere of the displacement theory in another way. In a work which first appeared in 1910 he assumes not unimportant horizontal displacements of the individual continents in Tertiary time and brings them into connection with the great Tertiary systems of folding. As an example he comes to practically the same view as that of the displacement theory to explain the separation of Greenland from North America. It is true that in the case of the Atlantic Ocean he assumes that only a portion of its width is due to the dragging away of the American block, whilst the remainder has been submerged and forms the elevation in the floor of the Middle Atlantic. Like Kreichgauer, Taylor sees in the drift of the land from the poles the guiding principle in the disposition of the great mountain ranges, whilst the displacement of continents plays but a minor rôle and indeed is only briefly treated.

As already mentioned, I became acquainted with all these works only when the displacement theory in its main outline had already been worked out, and with many others considerably later still. The possibility is not ruled out that in the course of time still further works will be discovered which will be in accordance with the displacement theory or which will anticipate this or that point. On this subject a historical examination has not yet been instituted and is not intended in the present book.

Chapter 2

CHAPTER II

Geology has not yet completely freed itself from the idea of a shrinkage of the earth. This contraction theory was especially advocated by Dana, Albert Heim and Eduard Suess, but still prevails as one of the fundamentals in geological text-books, as for example in those of E. Kayser and Kober. Just as a drying apple becomes wrinkled by folds on its surface through the evaporation from its interior, so should the earth through cooling and the consequent shrinkage form folded mountain chains on its surface. Suess coined the short phrase: “Der Zusammenbruch des Erdballes ist es, dem wir beiwohnen” (“The breaking up of the earth, that is what we see”). The historical service of this theory cannot be denied; it formed for a considerable period a sufficient epitome of our geological knowledge. As a result of this long period, during which the contraction theory has yielded such logical results from a large amount of varied research work, the immense simplicity of its basic principle and the multiplicity of its applications still give it a strong hold. Nevertheless, there can be no doubt that the contraction theory stands in direct contradiction to all the recent conclusions of geophysics, and that the course of geological research tends to deviate more and more from it.

The already difficult explanation of the mountain chains by reason of a shrinkage of the earth has been made to appear even more unsatisfactory by the discovery of the imbricated sheet folding or over-thrusting in the Alps. This new conception of the structure of the Alps and of numerous other mountain ranges, developed by the work of Bertrand, Schardt, Lugeon and others, involves far greater amounts of contraction than do previous suppositions. Whilst Heim, according to the latter, calculated a shortening of one-half for the Alps, he finds to-day, on the basis of the present generally accepted sheet-folding structure, a shortening to one-quarter or one-eighth.

Since the present breadth amounts to about 150 km., there would thus have been thrust together a portion of the crust of about 600 to 1200 km. (5 to 10 degrees of latitude). Any attempt to deduce such a decrease in diameter from a lowering of the temperature of the earth’s interior is bound to fail. E. Kayser remarks that a contraction of 1200 km. constitutes only 3 per cent. of the circumference of the earth, so that the radius must also be decreased about 3 per cent. These figures are significant only when the temperatures to which they correspond are calculated. On the basis of the average value of the four linear coefficients of expansion of nickel (0.000013), iron (0.000012), calcite (0.000015) and quartz (0.000010) giving 0.0000125, a loss of heat of about 2400° C. is required merely to explain the Tertiary folding. For more ancient periods, when tectonic movements were much more universally in operation, still higher values will be required. This, however, is in direct contradiction to the results of theoretical physics, for Lord Kelvin calculated from the present weak flow of heat from the centre of the earth to the surface that a temperature of the body of the earth so much higher in former periods is quite out of the question. Rudzki has, of course, drawn attention to the fact that in Lord Kelvin’s calculation no allowance is made for the work done in compression under the force of gravitation, in consequence of which, in spite of the loss of heat, the temperature remains nearly constant, and contraction is almost all that occurs; but nevertheless Rudzki hastens to add the conjecture that the above-quoted coefficients of expansion might probably be considerably decreased on account of the high pressure existing in the earth, so that Lord Kelvin’s calculations would again be correct. To sum up, it may be said that theoretical geophysics on this aspect of the subject does not yet lend itself to definite conclusions. The researches on radium appear to offer more unequivocal results. Great quantities of heat are generated through the automatic disintegration of radium. This element in any case exists in all rocks in such large traces, according to Joly’s determinations, that if the same radium content persisted to the centre of the earth, the constant conduction of heat from the earth’s interior, which can be verified by measurement of the increase of temperature with depth in mines, would be more than compensated.

Whether we must therefore assume, as Strutt believes we should, that radium only occurs in the uppermost crust of the earth, is still uncertain, but it is evident in any case that a noteworthy contraction of the earth from loss of heat by radiation cannot be seriously maintained. We know, indeed, no way of avoiding the conclusion that the heat-content of the earth is on the increase.

But even if, in spite of all this, such a contraction did take place, we should be forced to the fatal assumption, made, for example, by Heim, that the shrinkage of a complete great circle takes effect only at one single point upon it. This involves the impossible assumption of a transference of pressure within the earth’s crust over an arc of 180 degrees. Numerous authors, as Ampferer, Reyer, Rudzki, Andrée, among others, have objected to this and have demanded that the contraction must affect the whole surface of the earth in a manner similar to that of the puckering shown by the drying apple. Recently Koszmat, in particular, has repeatedly emphasized the fact that an explanation of mountain building must take into account immense tangential crustal movements, which do not fit in with the notion of a simple contraction theory. Thus doubt after doubt is experienced, so that the standpoint of geology has for some time been summarized in the statement: “The contraction theory has long ceased to be completely accepted, and mean no theory has been found that will replace it and is capable of explaining all the facts.”

But in my opinion it is chiefly in regard to another problem that the contraction theory is forced to declare its complete bankruptcy, namely, the problem of the ocean basins and the continental blocks. A. Heim had already thrown a little light on this problem in the statement, “that until accurate observations on the continental oscillations of the past are made, … and until we have more complete measurements of the amounts of the average contraction of the majority of mountain ranges, scarcely any essential and certain progress in our knowledge of the causal connection between mountains and continents and the form of the latter in relation to one another can be expected.” The problem becomes increasingly more urgent as the soundings of the seas of the earth become more numerous and the contrast between the extensive smooth surfaces of the oceans and the similar level surfaces of the continents, about 5 km. higher, becomes more sharply marked. E. Kayser wrote in 1918:—

“Contrasted with the volume of these stone colossi (the continental blocks), all the elevations of the mainlands appear to be small and trifling. Even such high ranges as the Himalayas are only ripples, insignificant in altitude, on the surface of those pedestals. This fact already makes the old view, according to which the mountain chains form the constructive framework of the continents, appear nowadays as untenable. … The converse rather must be assumed, that the continents are the older and the determining factors, the mountains being only of subsidiary and more recent formation.” But how can these colossi of stone be explained by the contraction theory? It says that detached portions of the crust are left behind at the time of a general foundering and remain as steps or horsts under the action of the pressure due to arching. But no account is taken of the enormous areas which are affected thereby. This whole conception, already refuted theoretically by Hergesell, of a stationary and everywhere effective arching pressure in the outermost layers of the crust, is in absolute contradiction to the more recent and increasingly well-proven doctrine of isostasy or the flotation of the crust of the earth on a plastic lower shell.

The contraction theory, with its conception, due to Lyell, of a limitless alternation of the emergence of the deep sea floor above water and the submergence of the continents to the deep sea floor, is also in contradiction to the doctrine of the permanence of the oceans and continents. As will be shown later, we cannot completely admit this doctrine, but its arguments directed against the contraction theory are quite valid. From the standpoint of the universally admitted theory of isostasy, it appears to be physically impossible that a whole continent should sink such a large amount as 5 km. On the other hand, the marine deposits in the present continents show that these—with insignificant exceptions—were never deep sea areas, but were covered by the shallow seas of the continental shelves. In this way the contraction theory is refuted in a striking manner by the greatest features of the face of the earth.

The theory of the displacement of continents avoids all these difficulties. It permits the assumption of the requisite horizontal contraction in the folded mountain chains; in fact, it makes this possible for the first time. For if the earth as a whole does not diminish to the same amount as its crust is compressed, then every compression of the crust corresponds to a rift at another place, and we thus arrive of necessity at the idea that the outermost skin of rocks no longer clothes the whole world. In addition, the continental masses, with their contrast to the extensive oceanic areas, certainly do not lend themselves to any other explanations. The theory of the displacement of continents thus replaces the contraction theory, which must be completely rejected.

Further, a separate explanation is also needed with reference to the doctrine of the sunken bridging-continents and the already mentioned contrary doctrine of the permanence of the oceans. The relation of the displacement theory to both these doctrines is different from that to the contraction theory. To anticipate the results. The arguments which are led out into battle by both these doctrines are correct, and so therefore are their mutual refutations. Each is based on only that portion of the facts necessary for a favourable judgment and receives its refutation directly the other portion is introduced. The displacement theory will fit the entire facts, and therefore prepares the way for a reconciliation of these hostile doctrines in a manner which satisfies all reasonable demands of both parties. In order to do this we must go further into the subject.

The advocates of the bridging-continents rely on the nowadays well-established fact that the close affinities of fauna and flora of widely separated continents require extensive land connections in the past. The ever-increasing stream of separate discoveries allows the picture of these connections to grow under our eyes, and to-day a very far-reaching agreement already prevails among the various specialists upon the most important of the land-bridges, although individuals cannot always “see the wood for the trees.” We refer in this connection to the summary given in of the views of twenty specialists on the several bridges, whether favourable or hostile. It is regarded as certain that there was a land connection, sometimes broken, between North America and Europe, which finally broke in the Glacial Period; a similar one between Africa and South America, which vanished in the Cretaceous Period; a third, the “Lemurian” bridge between Madagascar and India, which broke down at the beginning of the Tertiary; and finally a “Gondwana” bridge from Africa through Madagascar and India to Australia, which split up in the earliest Jurassic. Formerly also a land connection must have prevailed between South America and Australia, but the view that this was formed by a bridging-continent in the Southern Pacific is advocated by only a few workers. Most of them assume that this connection lay via Antarctica, since this lies on the shortest connection between both continents; moreover, the affinities are confined to such elements as can bear the cold.

Naturally a large number of bridges which to-day are represented by shallow seas are also assumed. The adherents to the doctrine of the bridging-continents have up to the present made no differentiation between the bridges over the oceans and those over the continental shelves. It must be especially emphasized that the displacement theory only deals with the question of land connection over the present-day deep sea areas for which new ideas are developed, whilst for the shelf-bridges, as the Bering Straits between North America and Siberia, the earlier view of submergence and re-emergence of dry land remains absolutely undisputed.

The adherents to the doctrine of the submerged continental bridges have thus a very strong argument: the former existence of broad land connections between continents which are widely separated at the present day can scarcely be doubted on account of the similarity of the fossil faunas and floras and the affinity of the existing ones. That these land-bridges were formed by continents which later sunk to the depths and form the present floor of the ocean, was assumed, on the grounds of the contraction theory, as self-evident and not requiring further proof, since the possibility of another solution by horizontal displacement was certainly not considered. As Ubisch emphasizes, the displacement theory will fit these requirements just as well as the present-day assumption of sunken intervening continents, indeed even better, because the close affinity of their faunas and floras would still appear rather mysterious on account of the present great distance from each other of these existing continents, even if an exchange of forms was at an earlier date possible over an intervening continent.

The arguments of the opposing doctrine of the permanence of continents and oceans rest not on biological but on geophysical grounds, and are directed essentially not against the former existence of land connections, but only against that of bridging-continents. The first argument consists in the facts already touched on, that on the continents deep-sea deposits do not occur to any considerable extent, so that the continental blocks as such are undoubtedly “permanent.” Strata taken earlier to be abyssal deposits have been found to be shallow-water sediments—as, for example, Cayeux demonstrated for the Chalk. For a very small number, such as the practically non-calcareous ‘radiolarites’ of the Alps and certain red clays which are reminiscent of the red abyssal clays, a great depth of origin is still assumed mainly because the sea-water acts as a solvent on the calcareous material only at great depths. The interpretation of these discoveries is, to be sure, still disputed, and it is believed that many are deposited at a depth of 1 to 2 km., which is still to be reckoned as belonging to a continental slope. But even if we follow Koszmat and Andrée, who, for example, assume a depth of deposition of the Alpine radiolarite of 4 to 5 km., the space occupied by such oceanic deposits, compared with the extent of the continents, is still so small that the principle of the general permanence of the continental blocks is not at all damaged by it. The present continental blocks, with but trifling exceptions, have never in the history of the earth formed the floor of the ocean, but were, as at present, continental platforms. The conception of Lyell of a repeated submergence and emergence is thus limited, in that it only concerns the alternating shallow floodings of the permanent continental platforms.

But this gives rise to a great difficulty in the construction of bridging-continents where now oceans, exist. If their re-elevation were not compensated by an equivalent re-submergence in another place, the remaining much diminished ocean basins would not have nearly enough room for the total amount of water of the oceans. The re-elevation of those former continental bridges would raise the surface of the water so much that all the continents, old as well as new, with the exception of lofty mountains, would be completely flooded. In other words, the assumption of continental bridges does not lead to the desired end, which consists of a land connection between extended continents. In order to surmount this difficulty, which was emphasized by Willis and Penck, we must make the assumption, not otherwise warranted, and therefore improbable, that the water on the earth had increased proportionately at the same time as the continental bridges sank. But this hypothesis has not up to the present been seriously considered and championed. With the more probable assumption that the amount of water has remained practically unaltered, the fact that in all geological periods extensive portions of the continental blocks have remained dry land, forces us to the conclusion that the total oceanic area has been essentially constant. This would mean that if the position of the continents is unchanged, hitherto considered as obvious, the several ocean basins have also been permanent features on the face of the earth.

Further, the advocates of the permanence doctrine base it on the geophysical facts of isostasy or the equilibrium in the earth’s crust. According to this theory, the relatively light crust of the earth swims in a somewhat heavier magmatic lower shell. Just as a piece of wood is immersed more deeply into water by loading it, so this uppermost rocky crust sinks deeper into the heavy magma, according to the law of Archimedes, at the place where it is burdened. This happens, for example, with an ice-cap, in which case the shore-lines formed during the depression will be elevated after the melting of the ice. Thus the isobase maps of de Geer, founded on raised beaches, show a depression of the central portion of Scandinavia of at least 250 m. for the last glaciation, an amount which becomes gradually less as we recede from the area, but which assumes still higher values during the Great Ice Age. De Geer has detected the same phenomena in the case of the glaciated area of North America. Rudzki has shown that on the assumption of isostasy plausible values for the thickness of the inland ice-sheet can be calculated, namely, 930 m. for Scandinavia and 1670 m. for North America, where the depression amounted to 500 m. Since the magma of the underlying shell is not so mobile as water, but is extremely viscous, all such isostatic movements of compensation must lag greatly. The strand-lines were formed chiefly after the melting of the ice, but before the elevation. Indeed, Scandinavia is still rising, as levelling shows, at the rate of about 1 m. in 100 years. Also deposition of sediments, as was first recognized by Osmond Fisher, results in the depression of the block. Each loading leads to a somewhat delayed sinking of the platform, and thus the new surface lies again at practically the old elevation. Sediments many kilometres thick may originate in this manner, which have, however, all been formed in shallow water.