CLASSIFICATION OF THE THERTODONTIA. 85 



tlie florsMl surface of tlie nose, in liotli foi'ms the extreme lowness 

 of the face in early Gorgonopsids is lo.st, the maxilla being deep 

 and the anterior part of the skull in general high. 



The incisor-teeth in Scylacosaarits are small and the premaxilla 

 shallow below the nostril. 



There is no step between the canine and the incisors, the teeth 

 forming a curved row like that of the latest Gorgonopsids and the 

 Cjnognathids. 



A preparietal appears to be lacking in Therocephalia, other- 

 wise the interorbital region does not difier gi'eatly from that of 

 Gorgonopsids. 



In the palate the Gorgonopsids are all more primitive than the 

 two Therocephalians in not possessing a suborbital fossa. They 

 are, however, all far more advanced in their possession of the 

 vaulted palate, which leads so directly to the development of 

 a secondary palate, and in the complete suppression of an inter- 

 pterygoid fossa. 



The median pai't of the palate of Scymnosauriis even projects 

 slightly above the general level, that of Scylacosaurus is essen- 

 tially flat. 



Scymnosaurus shows an advance on Scylacosmirus in the median 

 vomer Avhich appears on the palate. 



Both Therocephalians agree with one another in cei'tain special 

 features, such as the extent to which the ectopterygoid con- 

 tributes to the pterygoid flange (in which they differ from the 

 Gorgonopsids); and in their general appearance, in the structure 

 of the occiput, etc., they in no way recall the Oynognathids, as do 

 all the Gorgonopsids dealt with in the preceding parts of this 

 paper. 



/Scy7nnosaurtts and Scylacosanrtis are Tapinocephahis-7.owe 

 forms, and there is no evidence of any animals with similar 

 structure in the succeeding Endothiodon- and Cistecephalns- 

 zones. 



No certain Therocephalian is known in the JEndothiodon-zone 

 (as I understand it), unless Broom's Ictidognat/ms is of that age. 



In the Cistecephcdus- zone Scaloposaums, i-epresented only by 

 the type-skull from Stylkrantz, is the only satisfactoiily jDreserved 

 form. It has been well described and figured by Owen and Broom, 

 whose accounts should be referred to. The little skull difiers 

 exceedingly from Scymnoscmriis, the temporal fossae are short, 

 there are no pi-onounced sagittal and lambdoid crests, the squamo- 

 sals are not expanded, and the j^ostorbital apparently does not 

 reach the jugal behind the orbit. [This may be onlj- on account 

 of weathering of the surface.] 



On the palate there is evident a wide interpterygoid vacuity, 

 agreeing with tlie much narrower opening in Scymnosaurus ; there 

 are large suborbital vacuities as in that form. Nothing is shown 

 of the anterior part of the palate, nor are the details of the basis 

 cranii well displayed. 



Thus the extant material of Therocephalia sheds no light on 



