Page images

These small Percoid scales are similar to those of the Nandidae described above. The nuclear area is relatively smaller. The basal




TEXT-FIG. 12.-A fossil and two present-day scales of the family Pristolepidao. a. Scale of Pristolepis marginatus Jerdon. x8; b. Fossil scale (K29/629). ×171 c. Scale of Badis badis (Ham.). X14.

radii are long and converge in the nucleus. The number of radii varies considerably but they are never less than 8 or more than 15. As a rule, all the radii are complete.

Remarks. There appears to be a general similarity between the fossil scales and those of the genera Pristolepis Jerdon and Badis Bleeker. This similarity is more marked between the fossil scales and those of Pristolepis. From a consideration of the present-day distribution of the two genera, it appears probable that Pristolepis is more ancient than Badis. The former is found in South India, Burma, Siam, Cochin China and the Sunda Islands, whereas the latter is confined to India and Burma.

So far as I am aware no fossil remains belonging to this family have been described so far.



In a short preliminary article I (1938) have already briefly expressed my views on the probable age of the Deccan trap as evidenced by the fossil fish-remains, but below is given a detailed analysis of the main points at issue.

From a stratigraphical point of view the above results may be summarised as follows:

[merged small][ocr errors][merged small]

Family Serranidae: Genus? Serranus Cuv..
Family Nandidae: Genus Nandus C. V.

Upper Eocene to Recent.

Tertiary to Recent.

Recent; no fossils previously recorded.

Family Pristolepidae: Genus Pristolepis Recent; no fossils previously reJerdon. corded.

[ocr errors]

Lower Eocene to Recent.

Recent; no fossils previously re


Of the seven families to which Mr. Crookshank's material has been assigned, the first two belong to the Isospondyli, and therefore represent a very primitive stock of the teleostean fishes. Though the Clupeidae go back to the Lower Cretaceous, there is no doubt that the fossil fish scales referred to the family belong to the genus Clupea, which, according to Zittel, is not known from below the Upper Eocene. For the determination of the age of the beds, the evidence furnished by the Osteoglossidae is very conclusive, for the fossil remains of this remarkable family are not known from below the Eocene. The Cyprinidae, of which there is only one scale, also go back to the Eocene. The Serranidae are known in the fossil state mainly from the Tertiary formations. So it would seem that these beds cannot possibly be older than the Lower Eocene.

The present-day geographical distribution of the Polyacanthidae, the Nandidae and the Pristolepidae, of which no fossil remains had hitherto been described, throws some light on the upper limit of these beds. The Polyacanthidae and the Nandidae are represented by allied forms in Tropical Africa; in the case of the former the genus Anabas Cuvier (family Anabantidae) is common to Africa and south-eastern Asia. Of the two families, the Nandidae would seem to be older, as some allied forms are found in South America

as well. The family Pristolepidae is confined to south-eastern Asia. These facts of distribution suggest that the Labyrinthid fishes, such as the Anabantidae, Polyacanthidae, etc., and the Nandidae wandered over to Africa when there was still a land connection between India and Africa. The relative localised distribution of the Pristolepidae, on the other hand, suggests that these fishes were probably evolved after the severance of the connection between the two adjacent land masses. The conclusion which may, therefore, be drawn from the fossil records of these families is that the Inter-trappean beds at Deothan and Kheri were probably laid down when the land-bridge between India and Africa had already disappeared.

The only other fossil fish associated with the Deccan traps were reported from the Lameta beds at Dongargaon by A. S. Woodward (1908) who described one new species of each of the following genera: Eoserranus A. S. Woodw., Lepidosteus Lacépède and Pycnodus Ag. From these records he concluded that:

[ocr errors]

Although the Lameta fish-remains are not very satisfactory from the zoological standpoint, they are quite adequate for geological purposes. No true Percoid has hitherto been recognised in a typically Cretaceous formation in any part of the world; the oldest member of the group, as already remarked, being Prolates from the Montian of France. Eoserranus must therefore be regarded as a Tertiary type of fish. Lepidosteus is also typically Tertiary, differing essentially from the numerous Secondary ganoids to which it is related, by its highly specialised vertebrae. It ranges from the Lower Eocene to the Lower Miocene in Europe, and from the Eocene to the present day in North America; but has not previously been found in Asia. While Eoserranus and Lepidosteus can hardly be older than the beginning of the Tertiary period, the Pycnodont, according to the European standard, cannot be later than the close of the Eocene. The age of the Lameta fish-fauna is therefore fixed to be between the Danian Cretaceous and the Upper Eocene.'

[ocr errors]

It is worthy of special remark that Mr. Crookshank did not find in the Inter-trappean beds at Deothan and Kheri any representative of the fauna described by Woodward. Whereas Woodward described two Ganoids and one Teleostean fish, in Mr. Crookshank's material there are Teleostean fishes only. I have now found remains of Clupeoid scales in the material collected by Hislop and other earlier workers from the various Infra and Inter-trappean beds of the Central Provinces. The Ganoids are a very ancient race, but from the beginning of the Cretaceous period their dominant position became more and more displaced by the Teleosteans. By the beginning of the Tertiary period an almost complete change

over of the faunas had taken place and the Ganoids were then represented only by a few forms. It may, however, be noted that Lepidosteus and Amia continued to flourish for a considerable period, as their fossil remains are recorded for the last time from the Lower Miocene of Europe. They are among the few living genera of the Ganoid fishes. On a priori grounds it is clear that Woodward was dealing with a much more primitive fauna than that investigated by the present writer. Woodward's material was collected by Hislop (1861, p. 196), and it seems worth while to enquire into the history of this collection. Hislop made the following observations with regard to the remains of fishes:

[ocr errors]

The remains of fishes at Tákli and Pahádsingha consist chiefly of detached scales, some being Ganoidan and other Cycloidan. In the subtrappean yellow limestone of Dongargaum, sixteen miles E. S. E. of Chikni, the impression of a fish was found about 6 inches long and 18 inch in the broadest part, which must have been covered with cycloid scales, of a pattern that Sir P. G. Egerton had never seen. The same locality yielded the head of a fish about 9 inches long, with a produced muzzle, armed with sharp sauroid teeth, and rows of smaller ones. A fragment of bone, with 21 of apparently the very same smaller teeth, was dug out from the intertrappean of Tákli. According to Sir Philip, the ichthyolite most nearly allied to this is the Sphyrænodus of the London clay. In the Dongargaum limestone there was also embedded a portion of a Ganoid fish, Lepidotus or Lepidosteus (?) which is 6 inches broad, and when perfect was probably 2 feet long. This seems to be the species which has left so many of its scales in the intertrappean of Takli and Pahádshingha. A separate vertebra of a fish from the subtrappean red clay of Phisdura, 8 miles E. of Chikni, measures 7 inch in diameter."

In the above passage Hislop refers to three types of beds and enumerates the fossil fish-remains collected from each, but Woodward apparently dealt with only the fish-remains obtained from the Lameta beds at Dongargaon. If, however, the Dongargaon material collected by Hislop is listed with the material from the same beds examined by Woodward there appear to be some inconsistencies. The following table illustrates my point:

[blocks in formation]
[blocks in formation]


K1/316. Ganoid scales.
K1/317. Cycloid scales. Inter-trappean1

K1/318. Cycloid scales.
K1/319. Fish roe.


It would thus appear that Woodward examined more specimens from Dongargaon than are noted to have been collected by Hislop in the passage quoted above. Presumably at some later date Hislop may have made further collections of fish-remains from the Dongargaon beds.

From the sub-trappean red clay of Phisdura, Hislop obtained a vertebra, about 7 inch in diameter, which does not appear to have received the attention of any paleontologist so far. The specimen is not present in the collection of the Geological Survey of India, so I am unable to say anything about its systematic position.

[ocr errors]


From the Inter-trappean of Tákli and Pahársingha Hislop obtained detached Ganoid and cycloid types of scales. Further, fragment of bone with 21 small teeth, was also obtained from the Inter-trappean beds at Tákli; unfortunately this material has not yet been studied by any specialist.

Besides some of the Dongargaon specimens studied by Woodward and now preserved in the collection of the Geological Survey of India (Nos. 9360-68; vide Woodward 1908, pl. I, fig. 1-6, 9 b-d, 10), there are four other specimens (No. K1/316-319) in the collection of that department which are entered in the register as a donation from Rev. S. Hislop. The following further particulars are also noted against each

[ocr errors]

Skull and a group of head-bones with scales and vertebrae.

One skull.

[ocr errors]

Pahársingha, Nagpur, C. P.
Dongargaon, Nagpur, C. P.
Pahársingha, Nagpur, C. P.

Tákli, Nagpur, C. P.

1 Mr. Crookshank informs me that this should be infra-trappean. He wrote to me as follows:

"As regards the age of the Dongargaon beds we are satisfied that the Inter-trappean in our register should read infra-trappean. All accounts agree that there is only one fossiliferous bed and that infra-trappean. The map shows the same. The disagreement is as to whether this bed should be regarded as infra-trappean of Inter-trappean age or as Lameta."

« PreviousContinue »