40 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA. 
14 and 17 being from the middle of the cheek-series of the left side, and the one in 
fig. 1 6 the last upper true molar of the same side ; two are from bed Gc, while the 
third is from Gd. These teeth agree with those of existing species in the excessive 
length of the antero-internal pillar (e), and are inferior in size to E. namadicus of 
the Narbada pleistocene. 1 The specimen represented in fig. 16 appears scarcely 
larger than the corresponding opposite tooth of E. asinus represented in fig. 15 ; 
but this is accounted for by its much earlier stage of wear, its antero-posterior 
diameter at the horizontal plane corresponding to that of the latter specimen being 
0-98 inch. 
Lower molar. — The third right lower true molar from bed Cb in the Cathedral 
represented in fig. 13 exhibits the difference in size between the corresponding tooth 
of E. asinus (fig. 11). 
Metatarsal. — A right third metatarsal (No. F. 259) from bed Gf in the Cathedral, 
which has lost a portion of its distal extremity, agrees in relative size with the fore- 
going teeth ; its extreme length being 9*3 inches. This specimen is much smaller 
than the average of the metatarsals of the fossil races of E. caballus (and therefore 
than those of E. namadicus), and apparently indicates a species of the size of E. 
onager. 
Affinities. — The foregoing specimens are insufficient for specific distinction, and 
all that can be said about them is that they indicate a species superior in size to E. 
asinus , which is certainly distinct both from the larger E. namadicus of the earlier 
pleistocene, and E. sivalensis 2 of the pliocene of India. This species was about equal 
in dimensions to the existing Indian E. onager and some of the south African 
species ; 3 and, judging from the marked Ethiopian facies of a considerable portion 
of the Karnul fauna, and the absence at the present day of any existing wild Equus 
in southern India, it is not improbable that its affinities may be with the latter. 
Rhinoceros karnuliensis, nobis . 4 
History. — The remains on which this species is founded were provisionally 
identified by Mr. Foote 5 with B. sondaicus, but their distinctness was shown by the 
present writer in the paper cited above. The remains comprise numerous detached 
upper and lower cheek-teeth, many of which are imperfect, the greater portion of 
the left ramus of the mandible, a fragment of a right ramus with one true molar, 
the greater part of a humerus, three imperfect cervical vertebrae, and the distal half 
of a metapodial. The mandible and the more perfect teeth, to which comparisons 
will be mainly confined, are figured in pi. X. 
Mandible. — It will be convenient to commence the description with the left 
mandibular ramus, of which two views are given on a scale of one half in pi. X. 
1 Vide supra, vol. II. pi. XTV. fig. 3 — on the assumption that some of the specimens are premolars. 
2 Distinguished hy the antero-posterior shortness of the antero-internal pillar of the upper cheek-teeth. 
3 In the upper molars of E. zebra figured by Riitimeyer in the “Pferde der Quaternar-Epoche” (‘ Abh. shweiz. pal. Ges.’ 
vol. II.) pis. I. and II. fig. 7 (1877), the antero-internal pillar is more elongated antero-posteriorly. 
4 ‘ Rec. Geol. Surv. Ind.’ vol. XIX. p. 120 (1886). 5 Ibid. vol. XVIII. p: 232 (R. javanicuej . 
