Page images

of different ages. In the Sarmatian deposits of France, Germany, and the Vienna Basin, the Equidae are represented by Anchitherium, whereas near Sebastopol beds of the same age contain the much more advanced genus Hipparion. But for the fortunate circumstance that the latter are intercalated between marine strata of undoubted Sarmatian age, there would have been no hesitation in assigning the Sebastopol fauna to the Pontian. It is certain that the Sarmatian sea which lay between the Crimea and the Carpathians prevented Hipparion from proceeding farther to the West, but once the way was open migration was resumed and proceeded very rapidly in Pontian times.

Throughout the Miocene and Pliocene, India or an immediately adjoining area a great centre of Proboscidean evolution and several groups, including Choerolophodon, the stegodonts, and true elephants, appear to have arisen there. These three and others eventually became migrant, they are mentioned in this place because recognisable teeth of the first two are found in India at a much earlier date than in any other localities, and this stresses the need for caution when dealing with the first appearance of the true elephants. Similar caution is needed when the Bovinae are considered.



II. The correlation of the Fatehjang, Dhok Pathan and Pinjor.

In considering the correlation of the Siwalik stages with those of Europe, one would wish to extend ones comparisons to Burma, but this is not practicable because it is not yet possible to correlate the deposits of India and Burma by means of the Proboscidea. Indeed, so far as this particular group is concerned, the fossil Proboscidea of the pre-Pleistocene deposits of Burma appear to have closer affinities with those of Borneo and Java than with those of India, a fact which suggests that during part of the Tertiary, probably the Upper Miocene and the whole of the Pliocene, Burma supported a terrestrial fauna belonging to a different biological province from the Indian fauna.

In India, the Proboscidea were invaders during the Fatehjang stage, but in the later stages they were indigenous, and except in a few well-defined instances they ceased to be directly comparable with those of Europe. No matter which of the European genera is considered, all the Proboscidea found in that continent are immigrants. Never-the-less, three of the Indian stages, namely the


Fatehjang, the Dhok Pathan, and the Pinjor are marked by the immigration or emigration of Proboscidea ; if their European equivalents can be established the choice of possible equivalents of the remaining stages will drastically be curtailed.

Fatehjang Stage.The earliest known Proboscidea found in India are those of the Gaj series of Baluchistan. Pilgrim (1912) pointed out that in its morphology Hemimastodon is intermediate between Palaeomastodon and the trilophodonts of Sansan and Simorre. Forster-Cooper (1922) showed that the same genus, to which he referred as Bunolophodon, differs from the animals of Sansan and Simorre in that it has the last two premolars and the first two molars all in wear at the same time. These observations indicate that the Fatehjang stage is earlier than the Lower Vindobonian of Europe.

Direct comparison between the earliest mastodonts of Europe, namely those from the Orleans sands, and the specimens from Baluchistan is not easy, because the former have not adequately been described. Mayet (1908) mentioned very important specimen from Beaugency and now in the Natural History Museum at Orleans. He did not publish a figure of the specimen, but he says that it enabled him to establish the dental formula in the upper and lower jaws as Il: P2 : M3. This statement suggests that in this specimen the last two premolars and first two molars are in wear, and, one would presume, the fully formed third molar is still in the crypt. The isolated third molars found in the Orleans sand are slightly more advanced than those of Hemimastodon, but the difference is not so great, and these teeth together with the skull mentioned by Mayet indicate that, on the evidence of the Proboscidea alone, the Fatehjang stage should be compared with the Calcaire de Beauce, and that in any event it may safely be regarded as the equivalent of the Lower Burdigalian.

Dhok Pathan Stage.— The Dhok Pathan stage is distinguished by a profusion of specimens of Choerolophodon. This genus is indigenous to India ; its remains are rare in strata older than the Dhok Pathan, although characteristic teeth are found in the Chinji stage (e.g., tooth regd. Geol. Surv. Ind. No. A502), but when that stage is reached they are found so frequently, and in such variety, as to indicate that the genus had then achieved a dominant position. This same genus is found in Asia Minor and in Europe, but it is restricted to the Pontian deposits of Maragha, Samos, and Pikermi. These extra-Indian animals were described by Gaudry as Mastodon pentelici : they are curious for the circumstance that whereas skulls of immature animals are fairly common, remains of adults, even isolated teeth, appear to be rare in collections.

The sudden appearance of Choerolophodon in the European deposits proves it to have been an immigrant. The only problem is to which of the Indian species does Choer. pentelici correspond ? Of this there can be no reasonable doubt: although they are not identical, the specimens from Maragha and Samos are at exactly the same evolutionary stage as the species described by Pilgrim as Mastodon corrugatus and by Osborn as Synconolophus dhokpathanensis. If the Pontian were older than the Dhok Pathan one would expect the European species to be more primitive, for it is not credible that a migrant should be in advance of its contemporaries at or near the centre of evolution. The history of Choerolophodon, then, supports the view that the Dhok Pathan is the equivalent of the European Pontian.

There is an additional parallel between India and Europe when the European Tetralophodon longirostris (Kaup) is considered. This species is very closely allied to an Indian species which may, for the moment, be called Tetr. perimensis F. & C. The proper name of the latter is at present uncertain, but it has always been quoted under the name M. perimensis. It is well known that Tetr. longirostris is restricted to the Pontian deposits in Europe (cf. Schlesinger, 1922) but it is not always realised that this s ecies is a migrant, and that the supposed descent from Tri. angustidens is illusory. There are certain features of Tetr. longirostris which never occur in the supposedly transitional teeth. In India the position is quite otherwise. There are three-ridged teeth known from the Chinji stage which apart from the number of ridges and the height of the crown possess the essential features of Tetr. perimensis, and one has every reason to look upon them as the immediate ancestors of that species. Even in the Kamlial stage broken th have been found which seem to represent a still earlier form.

Tetr. perimensis itself is never common in the Salt Range and the Siwalik Hills, but it has frequently been found in the deposits of the Dhok Pathan stage on Perim Island in the Gulf of Cambay; indeed the majority of the specimens from Perim Island belong to this species. Once again we have to deal with a zoological type, in this case a species, which appears to have evolved quietly in India and which first becomes abundant in the Dhok Pathan stage.

[ocr errors]

In the Pontian of Europe there suddenly appears a so-called species which, at least in some individuals, is hardly to be separated from the Indian species of the Dhok Pathan stage.

Although the matter is complicated by the fact that as at present understood, both the Indian and the European “species appear to be collective, it is significant that there is in each country a group of specimens which is doubtfully separable, and then only with difficulty, from the corresponding group of teeth in the other country. Moreover, the evidence that these teeth are indigenous to India, that they flourished during Dhok Pathan times, and that they suddenly appear during the European Pontian, is such a precise parallel to the history of Choerolophodon that they afford the strongest possible support for the view that on the basis of the fossil Proboscidea the European Pontian and the Indian Dhok Pathan stage are exact equivalents.

Pinjor Stage. The fauna of the Pinjor stage has always attracted attention because it so closely resembles that of the Villafranchian of the Val d'Arno. The mastodonts found in the two deposits, M. arvernensis Croizet and Jobert and M. sivalensis Cautley, are so closely alike in the characters of their skulls and teeth that there is no justification for keeping them apart as separate genera. Indeed, it is not impossible that in course of time evidence will accumulate to show that they are geographical races of one species.

In the Chinji stage there was a group of mastodonts which had alternating cusps of

of the same

structure as those of M. sivalensis and M. arvernensis but which were otherwise in a more primitive evolutionary condition appropriate to the lower geological horizon. These teeth appear in each of the higher horizons, but do not become really abundant until the Pinjor stage is reached. How frequently they may occur in the Tatrot stage is not known for that stage, like the Nagri, has yielded very few remains of Proboscidea hitherto. This is an obstacle to be overcome before one may say precisely when the migration from India to Europe took place, but actually it is not of major importance in the present instance, and for the following reason.

There is abundant proof (Schlesinger, 1922.) that M. arvernensis invaded Europe in post-Pontian times (e.g., Levantine of Hungary); that it persisted until the Red Crag is proved by the presence in the British Museum of molars which retain their roots (regd. M7956). If it be shown at a later date that the stock first became abundant in the Tatrot stage, then, by the same reasoning as that adopted hitherto, the migration to Europe may be considered as having taken place at that time. In such case, the Tatrot would be the equivalent of the Levantine and the Pinjor of the Villafranchian.

On the information generally available, and on much more which is in preparation for publication, the fossil Proboscidea indicate that the most probable correlation for the Indian and European deposits of the Middle and Upper Tertiary is

[blocks in formation]

Within this framework the other stages must be fitted, but the group I have been considering does not afford the information necessary for the attempt to be made here.


ARAMBOURG, C. AND PIVE- Note préliminaire sur un Ruminant du
TEAU J. 1929 (1).

Pliocène inférieur du Roussillon. C. R.
S. Soc. Géol. France. No. 10, pp. 144-146,

1929 (2).

Les Vertébrés du Pontien de Salonique.

Ann. Paléont. Paris, XVIII, pp. 59-139,

pl. i-xii. BORISSIAK, A. 1914. Mammifères fossiles de Sebastopol. I, Mem.

Com. Géol. St. Petersburg, N. S. Livr.

87, pp. xii, 154, pl. i-x. 1915 . Mammifères fossiles de Sebastopol. II,

Mem. Com. Géol. N. S. Livr. 137,

pp. 1-47, pl. i-iii. COLBERT, E. H. 1935 (1). Distributional and phylogenetic studies on

Indian fossil Mammals. I. American
Museum collecting localities in Northern
India. Arme. Mus. Nov. No. 796,
pp. 1-20, 12 text figs.

« PreviousContinue »