CLASSIFICATION OF THE THEllIODONTIA. 55 



There can be little doubt, especiallj^ when the conditions in 

 Arctops and Gorgonoios are considered, that the inner pair of pro- 

 cesses which support the internarial bar are the anterior ends 

 of the pterygoids. It remains to be shown by other material 

 whether the median internarial bone and the median vomer in 

 the back of the palate are parts of the same bone or are, as is 

 more probable, separated. 



The strange way in which the median internarial bar rises as 

 a thin but very deep septum from the much sunk posterior nares, 

 nearly to the general level of the palate, seems to be only 

 explicable if its ventral border supported the middle of a small 

 soft secondary palate stretched between the maxillfe and the 

 palatines. 



I have already described the mode of ai'ticulation of the 

 squamosal with the brain-case and with the fused quadrate and 

 quadrato-jugal in Proc. Zool. Soc. 1914, p. 1034, fig. 6. 



The squamosal above the level of the post-temporal fossa bows 

 out backward, so as largely to increase the size of the dorsal 

 opening of the temporal fossa. It thus makes the occiput very 

 deeply cupped, the interparietal region being narrowed and the 

 outer part of the tabular running nearly antero-posteriorly. 



At the extreme postero-lateral corner of the skull, the 

 .squamosal turns sharply into a process passing forward and 

 inwaixl in the zygomatic arch. This process is clasped by other 

 bones both admesially and externally. One of these bones is 

 the jugal. The other conceivably also jugal, but much more 

 probably postorbital. A gap about 2 cm. long in both sides of 

 the specimen prevents a definite decision on this point. 



The squamosal at the corner is made of a very peculiar, 

 extremely dense, though finely cancellous bone. This structure 

 is found in this region in all Theriodonts I have examined. 



Leptotrachelus eupachygnathus, gen. et sp. nov. 



Type : a skull and lower jaw, described in error as Scymno- 

 gnathus whaitsi by the writer (Ann. & Mag. Nat. Hist. ser. 8, 

 vol. x. p. 578, fig. 3, and Proc. Zool. Soc. 1914, pp. 1027, 1032, 

 figs. 3, 4, & 5). 



The material is a largely disarticulated skull varying in 

 preservation, with one complete and one partially disarticulated 

 ramus of the lower jaw. The skull is represented by the brain- 

 case, interorbital region, left nasal, lachrymal, prefrontal, jugal, 

 and squamosal in natural articulation, the right jugal, lachrymal, 

 and prefrontal in natural articulation, but separated from the 

 skull, an isolated maxilla, and quadrate and quadrato-jugal. 



The mode of articulation of the quadrate with the squamosal is 

 clear, and with the perfect lower jaw gives the length of the skull 

 and the position of the maxilla. The large a,rticulated part of 

 the skull gives practically all the dorsal and the posterior 

 part of the lateral sui^face directly. The occiput is essentially 



