TERTIARY FISHES. 
13—253 
♦ 
indicates a slightly larger individual. The fossil cranium agrees almost exactly 
with that of the existing species, but the sculpture on the median line is perhaps 
rather coarser, and the tubercles on the lateral portion are somewhat more regularly 
conical. 
Afinities . — In the absence of the more characteristic anterior portion of the 
cranium it is almost impossible to say whether the fossil is specifically distinct from 
the existing African species, although on distributional grounds there is a con- 
siderable probability that such is the case. Two of the Indian species - A. thalassinus 
and A. gagoroides — are of considerably smaller size, but both present a median 
supraoccipital ridge. The largest of the twelve Oriental species of this group 
mentioned by Dr. Day^ is A. sagor, which attains a length of at least three feet, 
but has not the prominent median ridge on the supraoccipital. It seems therefore 
very probable that the fossil is specifically distinct from any existing Indian species, 
but all that can be safely said of it at present is that it indicates the existence of a 
siluroid in the Siwaliks fully equal in size to the largest individuals of the African 
Arms latiscutatus, and doubtless closely allied to that species and the two smaller 
Indian species mentioned above. 
II. Teeth on the palate in tivo generally zvidely separate patches] 'when vomerine 
teeth are present they are widely separated into two small groups., which may he con- 
tmuous with the palatine teeth. 
A. Teeth on the palate granular, or obtusely conical.^ 
Arius (?), sp. h. 
[From the Siwalilcs). 
History. — The first specimen noticed below has been described by Dr. Griinther,’^ 
who considered that it might possibly belong to Azdus. 
Palatme. —The specimen represented in plate XXXVII. fig. 7 was obtained 
from the Siwaliks of the Punjab by Mr. Theobald, and is described as follows by 
Dr. Gunther. “ One side (the inner) is more straight than the other (outer), which 
especially towards the front is curvilinear. One surface is entirely covered with small 
granular (molar) teeth, somewhat irregular in size and shape, the largest being about 
the size of a join’s head. This dental surface is strongly convex in its transverse as 
well as longitudinal axis, and, especially the part in which the fossil is widest, forms a 
conspicuous process. There is scarcely any doubt that this bone is the right palatine 
of a large siluroid.” 
An almost precisely similar, although less perfect, palatine of the same side is 
represented in fig. 8 of the same plate, which was collected by Mr. F. Fedden from 
the lower Siwaliks of the Laki Hills, Sind. There is every probability that this 
specimen is specifically identical with the pioceding. 
1 “ Fishes of India,” pp. 458-463 (1878). 
2 There are two species belonging to section I (p. 252) in which the teeth are granular, hut as they are both American 
it is improbable that the fossil can be related to them. 
3 ‘ Rec. Geol. Surv. Ind.” vol. XIV. p. 240 (1881). 
D 
