Page images

trappean fish-material in the collection of the Geological Survey of India which had not previously been studied by any specialist.

For the arrangement of the material I have followed Jordan's (1923) system of classification. The evidence furnished by these fish-remains in regard to the age of the beds and the ecological conditions under which the fish lived during that period are discussed at the end of the paper.

It may be indicated at the very outset that the study of fish scales is not sufficiently advanced to enable a specific determination. to be made of the various types represented in the material; their broad relationships are, however, quite clear and in some cases it has even been possible to refer them to genera or to very closely allied forms. The great value of the study of fish-scales for purposes of identification and classification is being more and more realised as the science of lepidology makes progress, for, like all other structures of a fish organisation, scales also vary and exhibit different patterns of ornamentation in various groups of fishes. The palæontological value of fish-scales appears to me on a par with that of fossil seeds or teeth. In fact, in most cases it should be more valuable than that of teeth or small fragments of bone.

In determining the fossil material I had to examine the scales of a large number of living forms, and where a similarity with the fossil scales was found, both the present-day and the fossil scales are figured side by side to bring out the full significance and stability of their characters.

Mr. Crookshank has kindly given me the following note concerning the Intertrappean beds at Deothan and Kheri :

These beds vary in lithology from place to place but the chief types at Deothan are black carbonaceous and buff ferruginous shales. Both these varieties are fossiliferous and contain numerous irregular masses of volcanic ash. At Kheri the fossils are found in calcareous shales.

“In both areas the Intertrappean bed is believed to overlie the basal flow of the Deccan trap.

“Other localities in the same region where fossiliferous Intertrappeans were noted are: Khuramba (22° 22' : 77° 29'), south by east of Patlai (22° 23' : 77° 33'), and east of Napupura (22° 21' : 77° 37'). None of these areas has been carefully searched. It is quite likely that a thorough search of them would bring to light further fossil remains of importance.

As regards Deothan and Kheri the outcrops were carefully searched, but no attempt was made to quarry the fossiliferous shales as it was found that the unweathered material was not fissile and the fossils in it were, therefore, very fragmentary. It is probable that good fossils might be obtained from the fresh material

[ocr errors]

if it were heated and later quenched, as this would make it split along the planes where the fossils occur.

“As well as the fish remains in these beds there are plant and insect remains, Cypridae, and Gastropods. These fossils have not so far proved to be of much interest."

I am very grateful to Mr. Crookshank for the opportunity of examining this interesting material and for his valuable criticism and helpful suggestions.

The whole of the material dealt with in this paper is preserved in the collection of the Geological Survey of India.


Superorder TELEOSTEI.



Family CLUPEIDAE. (Plate 17, fig. 3; plate 18, figs. 8, 9 and 10; text-figs. 1 and 2.)

Material.-Specimens with transverse radii widely interrupted in the middle, Nos. K32/154, K32/156, K32/159, K32/162, K40/291, K40/292, K40/293, K40/294 (from Deothan); specimens with transverse radii complete, Nos. K32/162, K54/293 (from Deothan).



6. TEXT-FIG. 1.—Two types of fossil Clupeoid scales. a. A scale (K32/159) with transverse radii widely interruptod. x 18}; 6. A scale

(G 152) with transverse radii complete. X ca 11.

There are a large number of small, fragmentary Clupeoid scales in the collection, but there is not a single complete specimen.

Two types of scales can be readily distinguished; (i) those in which the transverse radii are widely interrupted in the middle (Plate 17, fig. 3; text-fig. la) and (ii) those in which the transverse radii are complete (Plate 18, figs. 8, 9 and 10; text-fig. 1b).

In both cases the extremely fine circuli are more or less transverse and meet the margins at right angles or somewhat obliquely. Both types of scales are found in the genus Clupea Linn. and the nature of the transverse radii is often utilised in distinguishing species of these commercially important fishes, most of which superficially appear to be very much alike. Weber and de Beaufort (1913, pp. 76, 81) have figured similar scales (text-fig. 2), while Cockerell (1913, p. 123) has given a good general description of both kinds.

[blocks in formation]

TEXT-FIG. 2.-Scales of the two living species of Clupea Linn. (After Weber and de

Beaufort). a. Clupea (Harengula) fimbriata (C. V.). X6; 6. Clupea (Harengula) moluccensis

Blkr. x 6.

Remarks.—The Clupeidae comprise a great variety of marine forms and are found practically all over the world. A number of species, such as the American Shad and the Indian Hilsa, enter the mouths of large rivers at certain seasons and give rise to immense fisheries. Though some representatives of the family have been recorded from the Lower Cretaceous, Clupea is, according to Zittel (1932, p. 155), “not certainly known below the Upper Eocene of Monte Bolca, near Verona”. In view of the occurrence of the scales of Clupea both in the Infra- and Inter-trappeán beds, however, this view may have to be revised.


Family OSTEOGLOSSIDAE." (Plate 17, figs. 7 and 8; text-figs. 3 and 4.) Material.-Specimen No. K29/631 (two pieces), and two more pieces (from Deothan).

There are altogether four fragments of scales which I refer to the family Osteoglossidae. The scales of this family are large and thick and are composed of mosaic-like pieces. The largest piece in Mr. Crookshank's collection (pl. 17, fig. 7) is about 16 mm. in its greatest length and as this fragment represents more than half of the scale, it may safely be presumed that the diameter of the scale would be about 16 mm. The small mosaic-like pieces are rhomboidal ; they are approximately 3 to 4 mm. in length and 11 to


TEXT-FIG. 3.-A fragment of an Osteoglossid scale (K29/631), genus Musperia Sanders.


1 Jordan (1923) notes that the Osteoglossidae, the Arapaimidae and the Clupisudidae should perhaps be merged into one family.

2 mm. in width. Some of the pieces at one end are considerably larger. The longer axis of each piece points towards the centre and the periphery. Each piece is marked with very fine circuli which are moniliform throughout. In places where the rough, greyish skin has peeled off, the bony portions exhibit a more or less vermiform sculpture (text-fig. 3).

Remarks.-Of all the bony fishes, the Osteoglossidae are unique in regard to their present-day geographical distribution,

The living members of the family are found in South America [Osteoglossum (Vandelli) Cuvier and Arapaima Müller], Africa [Heterotis (Ehrenberg) Müller), Siam (Scleropages Günther), Indo-Australian Archipelago (Scleropages) and Australia (Scleropages). These are among the largest freshwater fishes known; the South American Arapaima reaches about 15 feet in length. The four living genera have only five species, of which two belong to Scleropages. S. formosus (Müll. & Schl.) is found in Sumatra, Banka and Borneo and has only recently been recorded from Siam (Smith 1931, p. 177); while the other species—S. leichhardti Günther-is mainly known from Queensland, but on the basis of a photo it has also been recorded from New Guinea (Weber & de Beaufort, 1913, p. 14). remarkable discontinuous distribution of the family is more or less parallel to the distribution of the living Dipnoan fishes, and shows the great antiquity of the Osteoglossidae.


[graphic][subsumed][merged small]

TEXT-FIG. 4.–Fossil Osteoglossid scales from Sumatra. After Sanders.

a. Scale of Scleropages sp.; b. Scale of Musperia radiata Sanders. The geological history of the family Osteoglossidae takes it back to the Eocene period (Zittel, 1932, p. 153).

Phareodus Leidy 1 Jordan (1919, p. 202) points out that Clupisudis Swainson should replace Heterotis Ehrenberg.

« PreviousContinue »