Page images

Gazella lydekkeri to G. capricornis and G. paotehensis.
Tragoportax islami to T. leskewitschi of Sebastopol.
Proamphibos (?) lachrymans to Parabos (?) macedoniae.
Giraffa punjabiensis to G. attica.

Hydaspitherium grande to Helladotherium duvernoyi.
Indarctos salmontanus to Indarctos cf. atticus.

Plesiogulo brachygnathus to P. brachygnathus of China.
Sivaonyx bathygnathus to S. hessicus.

Enhydriodon falconeri to E. (?) latipes of Pikermi.
Ictitherium sivalense to I. wongi of China.
Lycyaena macrostoma to L. chaeretis.

Crocuta mordax to C. eximia and C. variabilis.

Crocuta gigantea var. latro to C. gigantea of China.
Paramachaerodus pilgrimi to P. orientalis.

Machairodus or Epimachairodus sp. to Machairodus and Epi-
machairodus sp. of Europe and China.

Hystrix sivalensis to Hystrix primigenius.

Matthew and Colbert consider that the Tatrot fauna, which contains the early Bubaline, Proamphibos, closely allied to Parabos of Montpellier, but lacks not only more advanced Bubalines but also Leptobos and primitive Taurina is not Astian, as the Bovine fauna. would indicate, but the equivalent of the Villafranchian deposits of the Val d'Arno, Perrier, Senèze and Nihowan (China).

Finally, pursuing the same sequence of correlation which they observe throughout, they regard the Pinjor stage with Elephas (Archidiscodon) planifrons, the early Bubaline Hemibos, primitive forms of Bos and Bison, species of Crocuta which are closely allied to Villafranchian species, and the primitive goat, Sivacapra, as contemporaneous not with the Elephas planifrons beds of the Villafranchian, but with the Elephas antiquus beds of the Sicilian or Cromerian.

We have to infer from Matthew's and Colbert's correlation that the demonstrated European correspondence of the Indian faunas is merely homotaxial, and must be due either to migration from Europe to India or from some region midway between Europe and India which furnished migrants to Europe a stage earlier in each case than it did to India.

Matthew and Colbert refer to the general process as I have outlined it as exemplifying a series of "relict faunas". The term “relict fauna" is inapplicable. A relict fauna is one which being

originally much more widely distributed than it became later, has become confined from some cause or other to a restricted portion of its original area of distribution. The classic example of such a relict fauna is that of Australia, which provides us at the present day with the much modified descendants of a fauna which in the Cretaceous or earlier was of world wide distribution. The living fauna of Madagascar, consisting mainly of Lemuroidea with peculiar types of Carnivora, Rodentia, and Insectivora is a relict of the Eocene. Africa at the present day contains a relict fauna in the persistence of forms like the Tragulidae, the loxodont type of Elephant, Rhinoceros, the Giraffidae, most of the antelopes, Hippopotamus, the Tubulidentata (Orycteropus), the anthropoid apes, and many others of which some existed in Europe as lately as the Pleistocene, such as Elephas Hippopotamus and Rhinoceros, while the Tragulidæ, Orycteropus, the anthropoid apes and the majority of the antelopes died out in Europe as in most other parts of the world as early as the Pontian. There are many instances scattered throughout the world of faunas which at various periods contained relict forms, in some cases few, in others many in number. Thus in India itself Hyaenaelurus in the Kamlial stage is a relict of the Upper Oligocene, which persisted in India after it had died out elsewhere. In the Chinji, Hyænodonts, Proælurinæ and the Palaeochoerus type of Suid are obviously relicts. Even to-day, Tragulids, ibbons, elephants, and others less obviously so are relicts. In fact modern India shares with Africa but in a much smaller degree the possession of a fauna which in the Pleistocene or earlier was distributed over Europe and Asia.

It will be realized that such relict faunas and relict types are in a wholly different category from that in which we must place the successive fossil faunas of India which according to Matthew and Colbert represent those of one stage earlier in Europe, Central Asia and elsewhere. The constituents of each of these faunas are not in any sense survivals from the fauna of the preceding stage, but products of a contemporaneous migration, the forms allied to which perished in Europe or Central Asia simultaneously with the arrival of their congeners in India. Moreover, the migrants themselves did not flourish in India, as did the relict faunas of Australia, Madagascar and the African continent, but died out within the period of their migration.

Even if we suppose, which is indeed more likely, that the centre of distribution lay, not in Europe or Central Asia, but in a region

still unknown, situated between India and Europe, we are still faced with the curious circumstance that the fauna, while reaching Europe and Central Asia earlier than it reached India, died out immediately in its adopted country, even assuming that it may have continued in the country of its origin.

Geological history provides us with instances of the retarded migration of certain forms or even of an entire fauna from the country of its origin. This may be explained by the removal of physical barriers which had previously existed or by a change of climate or by both causes. The best instance of this in the case of terrestrial faunas which can be adduced is that of the Upper Sarmatian and Pontian mammals of Europe, which are so markedly different from those of the preceding stages embraced in the Miocene that it is only explainable on the assumption of a more or less sudden migration which had not been possible hitherto. When once the migration routes had been opened, however, contemporaneous migration occurred and fresh invaders entered Europe, in most cases almost simultaneously with their appearance in adjacent regions without the retardation which had occurred during the Miocene.

I can, however, find no parallel, at any rate in the case of land faunas, to what Matthew and Colbert suppose to have taken place in India. We might perhaps imagine that the Pontian fauna was hindered from some cause or other from entering India contemporaneously in the Chinji period, but after the removal of the barriers one would expect that fauna which entered India in the Dhok Pathan period [i.e., in the Middle Pliocene (Astian), ex hypothesi Matthew and Colbert] to have been accompanied at any rate by some members of the contemporaneous Astian fauna of Europe. Let us see how far this expectation is realized.

Generic identities exist. Hippopotamus, in both regions an invading form, is extremely rare in the Dhok Pathan. Though it has not been found at Rossillon or Montpellier, yet it occurs in the Casino lignites which are believed to be only slightly younger than Pikermi and in the beds of the Wadi Natrun in northern Africa which are considered to be of Astian age. Its value as evidence of a similar age for the beds of Dhok Pathan and the Astian is somewhat discounted by our ignorance of its origin or adaptive centre of radiation. If it is related to the Anthracotheres, then it might be imagined that the centre in question lay so near to India, where Anthracotheres are more abundant than elsewhere, that it would

not be surprising that it should have reached that country earlier than Europe. The early arrival of the Suidæ into India may be explained in a similar way. Touching the possibility of an identity between the Astian Potamochoerus provincialis and the Dhok Pathan species which have been named Propotamochoerus, the more primitive character of the latter is shown by the less scrofic structure of the male lower canine and the simpler construction of the main cusp of P. The Dhok Pathan Gazella lydekkeri is undoubtedly more primitive than the Astian Gazella borbonica. The primitive buffaloes, Parabos cordieri and P. boodon of Montpellier are certainly allied to forms from the Dhok Pathan, which have been attributed to the same genus as Proamphibos lachrymans of the Tatrot. The two genera are not however identical, the European form being more primitive as regards the teeth. Moreover a presumably ancestral form exists in the Pontian of Europe as Parabos (?) macedoniae, while the Dhok Pathan stage has yielded another species referred to Proamphibos, P. (?) hasticornis, which has certainly more primitive skull characters than Parabos. Amongst the Carnivora the species Megantereon (?) præcox from the Nagri (or Dhok Pathan) stage of India, supposing it to be generically the same as Megantereon megantereon of Roussillon, is at any at smaller and more primitive than the latter. Agriotherium palaeindicum which presumably occurs in the Dhok Pathan stage side by side with two species of Indarctos is undoubtedly more primitive than Agriotherium insigne of Montpellier. Vishnuictis salmontanus is not closely related to Viverra pepratxi of Roussillon. Hipparion crassum of Roussillon is a more advanced form than either of the Dhok Pathan species of Hipparion in the atrophy of the protostylid in the lower premolars and of the median accessory cusp in the lower milk molars, in the broader median metapodials and the less prominent lateral metapodials.

On the other hand the Ursinæ represented by Ursus (Helarctos) arvernensis, the Canidae represented by Vulpes donnezani, hyænas of the type of H. striata represented by H. arvernensis at Roussillon, and rodents of a modern type such as Mus donnezani do not appear in India until the Pinjor stage or later, although their absence from the Tatrot fauna may be only apparent and due to the poverty of the latter.

On the whole contemporaneous migration of the Astian types of Europe into the Dhok Pathan deposits of India, if it took place

at all, was much less abundant than one would expect, judging by the extremely close correspondence of many species from Pikermi, Samos and the Hipparion beds of China with those of the Dhok Pathan.

To consider the earlier horizon of Chinji, regarded by Matthew and Colbert as Pontian. Although its fauna is essentially of a Vindobonian type, yet according to Colbert it must have largely migrated to India in the Pontian accompanied by the contemporary form Hipparion. My comment on this theory is the same as above. It would be reasonable to expect that other contemporary forms should have entered India either from Central Asia or Europe in addition to Hipparion. At first sight the Chinji fauna, containing as it does large quantities of antelopes, as well as Giraffidæ, Suidæ, Hyænidæ of the type of Crocuta eximia and Ictitherium seems to offer a prospect of fulfilling this expectation. Gazella, it is true, occurs in the Chinji, but the vast bulk of the Bovidæ consists of a group of antelopes which bear little affinity to those of Pikermi and Samos, but are related rather to Boselaphus. The sole representatives of this subfamily in Europe apart from Tragocerus are Protragocerus chantrei, "Tragocerus leskewitschi, " Tragocerus" latifrons and Miotragocerus monacensis. The first of these is from the Tortonian beds of La Grive Saint Alban, the second from the Sarmatian of Sebastopol, and the two last from the Sarmatian of the Vienna basin. The first is only known by fragmentary remains but is related both to Strepsiportax and Tragoportax, the second is probably a Tragoportax, the third is probably a Strepsiportax but more advanced than either of the Chinji species; the last is allied to the Chinji Sivaceros but is much more advanced in the direction of Tragocerus. The large Hippotraginæ of Pikermi and Samos such as Palaeoryx, Protoryx, Microtragus and equally those of China which include in addition to these the genera Sinotragus and Prosinotragus are absent from the Chinjis. The numerous Pontian Tragelaphinæ of Europe are represented by but a single genus, Sivoreas, which is smaller and more primitive than any Pontian form. The ubiquitous genus Tragocerus does not reach India until the Dhok Fathan, although its probable ancestor Sivaceros appears in the Chinji.

[ocr errors]

Matthew and Colbert (Colbert, 1935, pp. 25, 26) consider that the Giraffidae afford a strong proof of the. Pontian age of the Chinjis. In my opinion, however, too little is known of the origin

« PreviousContinue »