CLASSIFICATION OF THE THERIODONTIA. 91 



larger temporal fossae, visible from above, in the outward bowing 

 of the zygoma, in the vertical occiput, and in the smaller 

 lachrymal — all changes which follow the ordinary trends of 

 Theriodont development. 



The neural cranium, as shown in von Meyer's figures and in 

 text-fig. 27, is Pelycosaur-like in general build, and especially in 

 the complete absence of that forward growth of the supraoccipital 

 and pro-otic above the notch for the Vth nerve which occurs in 

 Theriodonts, and in the occurrence of a special notch for a vein 

 above the incisura prooticis. 



This occiput is, however, specialised in the development of a 

 mass of bone below the basioccipital condyle, which is presumably 

 associated with a very vertically placed fenestra ovalis. The 

 development of this plate is the explanation of the extreme 

 depth of the pituitary fossa. 



The palate of Rhopalodon recalls that of Dimetrodon in its 

 massive flanges on the pterygoids and in the row of teeth which 

 decks them. The very large internal nares also recall certain 

 advanced Pelycosaurs. 



In certain ways the skull of Rhopalodon resembles that of the 

 more primitive Gorgonopsids ; the dorsal surface of the skull, for 

 example, is very like that of Gorgognathus. The side view 

 differs, however, in the great depth of the snout and in the 

 relatively powerful molar dentition. 



Although Seeley's figures are not very readily interpreted in 

 that region, it seems that the quadrate of Rhopalodon is large 

 and well exposed from behind, and that its outer edge lies on the 

 outer surface of the skull exactly as in the South African 

 Deinocephalia, and not at all as in the Gorgonopsids, although it 

 represents a state from which that in the latter group could 

 readily be derived. The brain-case of Rhopalodon differs from 

 that of any Gorgonopsid in the vertical plate below the condyle 

 and in the non-extension of the supraoccipital, etc., forward. 



It is unfortunate that no part of the palate behind the flange 

 is preserved, but judging from the front of the fragment of basi- 

 sphenoid preserved in the occiput, and the general structure, there 

 can have been no narrow bar separating the subtemporal fossae 

 as in Gorgonopsids, but the conditions must have been more as in 

 the South African Deinocephalian Mormosaurus. There is, in 

 fact, no doubt that Rhopalodon is not a primitive Gorgonopsid, 

 but is a primitive Deinocephalian, with the members of which 

 group it agrees in all the characters in which it differs from the 

 Theriodonts. Its general resemblance to primitive Theriodonts 

 suggests, however, that we are very near the point of separation 

 of these two orders. 



The present seems a suitable opportunity for adding to the 

 description which I gave (Proc. Zool. Soc. 1914, p. 770, etc.) of a 

 skull referred to Titanosuchus. I recently found a block fitting 

 on to the fragment of maxilla of that specimen which contains 

 the anterior end of the maxilla, parts of the premaxillaa, septo- 



