Page images

at Roussillon and at Perrier (Roccaneyra); the Ctenodactylida as Ruscinomys ; Trogontherium as a variety minus of the wide spread Villafranchian T. cuvieri. Castor as the species C. praefiber, which does not differ except in size from the later C. fiber ; Agriotherium replaces Indarctos of the Pontian and is represented by A. insigne of Montpellier and A. sp. Flower recorded by Newton from the Red Crag. Although Hipparion does not generally occur in the Villafranchian, yet it has been recorded by Stehlin from Perrier (Roccaneyra). Besides the Ctenodactyl Ruscinomys and the Macacine primate Dolichopithecus, the only genera which do not persist into the Villafranchian are Parabos and Viverra, of which the last European representative occurs at Roussillon as V. pepratri.

Numerous species of Gazella and Crocuta occur in the Villafranchian which have not been found at Roussillon, as well as Mastodon (Zygolophodon) borsoni. The absence of all of these from the Astian fama of Europe especially of the last may be accidental. An ancestral form of Mastodon (Zygolophoclon) borsoni occurs in the Pontian as M. (Zygolophodon) turicensis.

The differences which the fauna of the Villafranchian displays from that of Roussillon are mainly due to what are invading forms, which does not necessarily imply any great difference in age. This part of the fauna includes Equus, Elephas planifrons and E. meridionalis, Pannonictis, Sus strozzii, numerous Caprine lineages such as Deperetia, Procamptocerus, Megalovis, Nemorhædus, the spiral-horned Gazelline, Gazellospira torticornis, and Leptobos.

When we observe that the comparable forms are either specifically the same or differ very little from one another, we are forced to conclude that little development took place between the two levels. Very little of the Roussillon and Montpellier fauna became extinct between that stage and that of the Val d'Arno, Perrier and Senèze, as compared with the multitude of families and genera which appear for the last time in the Pontian. I can see no grounds for regarding the two faunas as very different in age, while on the contrary a great zoological gap is suggested between Roussillon and the Pontian. Two littoral or marine stages, the Plaisancian and the Astian, intervene between the Villafranchian and the stage which corresponds to the mammaliferous lignites of Casino. The fauna of the Casino lignites is perhaps slightly younger than that of the Pontian. Depéret, therefore, most reasonably considered that the Roussillon fauna corresponded with the Astian, while the Plaisancian filled the faunal gap between the Pontian and Roussillon faunas. There is no mammal fauna known in Europe which can be said to correspond with certainty to the Plaisancian. Besides the Casino fauna, Depéret suggested that the Baroth Kopecz lignites of Hungary and those of Alcoy in Spain might belong to this period. Depéret's conclusions as to the general faunal succession have never been seriously disputed, and Lewis has advanced no arguments for overthrowing the view generally held.

To return to India, it will be convenient if as a preliminary to a discussion of the main correlation I bring forward certain arguments based on the fuller examination of the Tatrot fauna, of which Lewis, De Terra and Teilhard were unaware when they expressed the above views. I hope these will show at any rate the utility, if not the absolute necessity, of dividing the Upper Siwaliks below the Boulder Conglomerate into two faunal stages. Originally (Pilgrim, 1913, pp. 276, 321-323) I based my separation of the Tatrot from the Pinjor on two main considerations : 1. The apparent absence of Equus and Elephas from the

Tatrot. 2. The assumption that certain species which abounded at

Tatrot but had not been rediscovered in the Siwalik Hills since the time of Falconer, belonged to a deposit which was older than that to which the bulk of the Siwalik Hills fossils were obtained. Examples of these are Hipparion, Hippohyus, and Mastodon (Pentalophodon) sivalensis. This assumption was based on the fact that in more than one place on the plains side of the Siwalik Hills there exist deposits which are not only lower than the Pinjor beds but also differ in their lithology. The best known of these is that in the Kalawala Rau from which Falconer (1868, pp. 33-36) obtained numerous fossils ; I have myself found Hipparion in the same spot. Regarding Falconer's collection it is unknown whether it still exists or has become mixed up with those of the Pinjor stage. The beds lie east of the area which Colbert has reproduced in his sketch map of the neighbourhood of Moginand (Colbert, text figure 25,

Reference has already been made (Pilgrim, 1919, p. 193 ; 1932, p. 3) to the possibility of their being a representative of the Tatrot stage.

p. 49).

The discovery of Equus at Tatrot, and of a primitive species of Elephas (Archidiscodon) cf. planifrons by Lewis disposes of the first of these arguments.

The second though not actually proved to be wrong is still invested with some doubt, because with the exception of Merycopotamus, none of the suspected species has yet been rediscovered in the Siwalik Hills, while Osborn has erected for the Tatrot Mastodon a new species, Pentalophodon falconeri, distinct from that of the Siwalik Hills. There are, however, other important differences in the known faunas of the Tatrot and the Pinjor, which seem amply to support my contention that these two stages are faunistically distinct. These were for the most part unknown to Lewis, as he had only my provisional and incomplete list of species (Pilgrim, 1913, p. 322) before him, the corrections and additions to it being scattered through publications dealing with special groups, or in the case of the Bovidae in an unpublished memoir (Pal. Ind. N. S. XXVI, 1939). Even now I regret that I have not had recent access to the original collections made by Vinayak Rao and myself on the Tatrot plateau. They are in any case not rich in species, though intensive collection at Tatrot, Kotal Kund and Bhaun would probably add considerably to the number of these.

I should first like to comment on the occurrence of Hipparion at Tatrot, which both De Terra and Teilhard as well as Lewis have stated to occur as rolled fragments derived from the Dhok Pathan, and regard as important. The presence of rolled Hipparion at Tatrot does not in the least affect the certainty that Hipparion must have lived at the Tatrot period, since numerous specimens of it from the Siwalik Hills exist in the British Museum (Lydekker, 1885, pp. 59-64), which even if they did not come from the Pinjor stage (a supposition which is unlikely since the character of the matrix agrees with that of the bulk of the fossils from that stage), yet cannot belong to an earlier horizon than that of Tatrot, since Dhok Pathan beds are not exposed in that area. Consequently the occurrence of rolled fragments of lipparion in the basal conglomerate of Tatrot, however interesting, is without significance in connection with the question of the correlation of the Tatrot fauna.

Following the principles which have determined the separation of the European faunas of Roussillon from that of the Val d'Arno, Perrier and Senèze, I will now state some of the reasons which, in my opinion, justify the faunal separation of the Tatrot and Pinjor stages. Quite a fair proportion both of species as well as genera pass up from the Dhok Pathan into the Tatrot.

The list given below would probably be much greater if the Tatrot fauna were richer. Unfortunately, too, we have no assurance that some of the specimens obtained at Hasnot by the

the early geologists did not come out of Tatrot beds. In those days the accumulation of fossils in the basal beds of the Tatrot stage must have been much greater than is the case to-day, and it may be taken as certain not only that they ran more risk of being washed down on to Dhok Pathan outcrops, but also villagers who brought fossils to the geologists cannot have failed to visit the Tatrot cliffs, since they lie little more than two miles from Hasnot. In such cases the locality label attached to the specimen would be misleading, and our only clue to the horizon is the character of the matrix which is wholly different in the case of the two deposits.

Where the matrix has been removed, this indication is lacking. Even, however, if we disregard the possibility that the Tatrot fauna is richer in such specimens than the actual list of species would lead us to imagine, yet there remain some which we definitely know to have been collected from the Upper Siwaliks of Tatrot, Kotal Kund or Bhaun which cannot be separated specifically from those of the underlying Dhok Pathan stage, but which do not persist into the Pinjor stage, except in the case of Merycopotamus. These are :

Merycopotamus cf. dissimilis.
Dicoryphochoerus tilanoides
Dicoryphochoerus vagus
Hippohyus lydelikeri
Hippohyus grandis
Dorcutherium majus
(?) Antilope (?) planicornis
Selenoportux lydekkeri
Pachyportar latidens type variety

Sivaonye bathygnathus The following Dhok Pathan genera occur also in the Tatrot stage, represented by species which may not be different from those in the Dhok Pathan: Hippopotamus; Sivacharus ; Proamphibos ; Hipparion ; Stegodon ; Stegolophodon (if the locality Lehri from which S. stegodontoides was obtained is Tatrot); Agriotherium (if the species A. pulvindicum is Dhok Pathan). Reduncinae which occur in the Dhok Pathan stage may belong to the same genera as those from Tatrot. They agree in size and in degree of development.

Considering the poverty of the Tatrot Palmit, it seems useless to assume that various Dhok l'athan genera have become extinet, because they have not been found at Tatrot, but it seems likely that most of the forms corresponding to those which became extinct in Europe before Roussillon did not persist into the Tatrot stage. We have, however, to take into account that many of these were Pontian invaders into Europe, which may not have been the case in India. Such are forms like the Giraffidae, the Chalicotheriidae, Enhydriodon and the Catarrine monkeys which we know to have existed in India up to the Pinjor or later.

Genera in the Tatrot which are the same as those of the Pinjor, though the species are generally different, are : Mastodon (Pentalophodon) possibly identical with the species M. (Pentalophodon) falconeri ; Stegodon, as the species S. bombifrons ; Elephas (Archidiscodon) cf. planifrons ; Equus sp. ; Hipparion sp. ; Sus, as the species S. peregrinus ; Potamochoerus, as the species P. palaeindicus ; Sivatherium ; Sivatragus, as the species S. brevicornis ; Siradenota (?) as the species S. (?) sepulta ; Hydaspicobus as the species H. auritus ; Indoredunca (?) as the species 1. (?) theobaldi.

The only Tatrot genus, so far as we are aware, which does not persist into the Pinjor is the early Bubaline Proumphibos, which is ancestral to Hemibos and possibly to Bubalus. Of the various antelopine species from Tatrot which have been referred to Pinjor genera, all of them are represented by species which are undoubtedly more primitive than the Pinjor species. The same applies to Potamochoerus palaeindicus and Sus peregrinus.

The occurrence of Equus and Elephas (Archidiscodon) cf. planifrons as pewcomers in the Tatrot, does not seem to me to militate against assigning it to the Astian stage, in spite of the fact that they do not occur in Europe previous to the Villafranchian, seeing that they are invaders into both regions, unless India happens to be the country where Elephas originated, in which case its earlier appearance in India is even less surprising.

In every other respect, so far as the meagre fauna of Tatrot enables us to judge, the differences between it and that of the Pinjor are quite comparable with and of the same quality as those between the faunas of Roussillon and the Villafranchian.

« PreviousContinue »