149 - 



para ted, indeed, by a relatively wider tract than in 

 BUnosaums. The orbits are large, distant and inclined 

 to the side. As in Rhinosaurus the intervening frontal 

 Fig. 2. platform equals in breadth the orbital dia- 

 meter. The internarial tract is convex (bom- 

 be), but difïers from that in Rhinosaurns, 

 in its greater relative breadth, Avhich equals 

 the interorbital tract. The parietals are lar- 

 ge, elongate, and show at the fore part of 

 the sagittal suture, in the same relative po- 

 sition as in Rhinosaurus, the circular im- 

 pression ov' foramen Horneanum , to which 

 Fischer de Waldheim specially calls atten- 

 tion (p. 5). In the non-extension of the ar 

 ticular part of the lower jaw (fig. 2, 29) 

 backward beyond the joint afforded by the 

 lympanic element of the temporal bone (ib. 

 28) — a character which is especially pointed out as diffe- 

 rentiating Rhinosaurus from Enaliosauria and Crocodi- 

 lia — my South African fossil agrees: and I append a magni- 

 fied view of the joint showing the precise relations of the 

 articular element, fig. 3, 29, to the tympanic, ib. 28. 

 The temporal fossae are completely roofed over by bo- 

 ne-plates (fig. 1, t, t) extending from the parietals, 7, 

 to the zygomatic arches, 27, as in Rhinosaurus: this 

 structure I have not seen in any Jurassic Saurian, save 

 Rhinosaurus^ if this should prove to have been derived 

 from the formation specified. But the most marked cha- 

 racter of correspondence between the Reptilian skulls 

 compared is the layer of finely granulate 'ganoine' cove- 

 ring the exterior surface of the skull-bonues, answering 

 to that described as (d'epiderme plutôt chagrinée ou 

 granulée qu'écailleuse» (loc. cit. p. 5): and which, al- 



Petrophryne 

 granulaf.a. 



