CLASSIFICATION OF THE THERIODONTIA. 97 



When the Cynodontia are reached we find a still further 

 decrease in the relative importance of the incisors, together with 

 an increase in the number of cheek-teeth in Gynosiichus to 8, all 

 still much smaller than the incisors, in NytUosaurus to 8, 

 all larger than the incisors, Galescmrus i 10, Cynognathus 9, 

 Diademoclon ? 12, Trirachodon 9. This sudden increase in the 

 number of cheek-teeth seems to be associated with the develop- 

 ment of a bony secondary palate, which allows of prolonged 

 mastication being carried on without obstruction of the naso- 

 pharyngeal passages. 



It is important to note that the anterior five "molars" of 

 Clynognathits crateronotus are sharply distinguished from those 

 which succeed them, and that in Diodemodon, when the four 

 anterior cheek-teeth are similarly distinguished by structure, 

 these teeth and these alone amongst the cheek-teeth give indica- 

 tions of replacement. 



In fact, the evidence brought forward by Broom (Bull. Amer. 

 Mus. Nat. Hist. vol. xxxii. p. 465), although it is not quite so 

 conclusive as one could wish, does tend to show that these 

 Cynodonts had a thoroughly mammalian dentition with premolars 

 replacing milk-predecessors and molars never replaced. 



Comparison with Gorgonopsids suggests that the premolars are 

 the original cheek-teeth, the molars representing a new backward 

 growth of the dental lamina, in which the teeth all belong to a 

 single generation, and from their origin have never had either 

 predecessors or successors. 



I hope to return to a consideration of the whole problem of 

 dental succession in early reptiles shortly. 



The Gorgonopsids being characterized throughout their history 

 by the possession of a very short series of molar teeth, we have to 

 consider the problem presented by the fact that whilst some 

 Therocephalia, e.g., Scymnosaurus and Hycenosuchns, I'esemble 

 them in this feature, othei'S, Alopecodon etc. amoiigst large forms 

 and Icticephcdus and Scaloposaurus amongst the small, retain 

 a large sei'ies of molars, without, so far as known, having a 

 secondary palate, and possess in addition small incisors. 



It seems reasonable to regard these latter animals as unmodified, 

 the dentition being derived directly fi'om their ancestors, whilst 

 Hycenostichus represents a. parallel reaction to that of the Gor- 

 gonopsids to similar feeding habits. All detailed discussion of 

 the dentitions of Theriodontia are rendered nugatory by our 

 complete absence of knowledge of the postcranial skeleton, for 

 only by a study of the whole structure is it possible seriously to 

 consider the habits and adaptations of an animal, and teeth react 

 perhaps more quickly than any other structures to external 

 impressions. 



In any case the occurrence of these Therocephalians with a 

 long tooth-row and small incisors, and of a similar structuie in 

 CUorhizodon, shows that the heavy incisors of Gorgonopsids 

 Proc. Zool. Soc— 1921 , No. VII. 7^ 



