168 DR. R. BROOM ON 



superficial resemblance. As, however, it has a different dental 



formula and must be placed in a distinct genus, I have much 



pleasure in proposing for it the name Cyniscodon lydekkeri. 



Gynosuchus suppostus, with which this new form has been 



confused, is known only by the imperfect type skull. It is a 



most remarkable form, having cuspecl molars and a secondary 



palate like the typical Cynodonts, but in other respects differing 



from all known Cynodonts and resembling more the Gorgono- 



4 17 

 psians. The dental formula is probably i q, c ?, m =, and to 



whatever suborder a more complete skull may show it to belong, 

 it must be placed in a distinct family — the Cynosuchidse. 



In Cyniscodon lydekkeri the dentary is considerably smaller 

 than in Gynosuchus suppostus. It has the deep symphysis 

 characteristic of the Gorgonopsians. In the specimen as pre- 

 served are the roots of three incisors which are probably i\ i 3 , 

 and i\ Together they measure 5 mm. The whole four probably 

 measured 7 mm. The canine measures at its base 6 mm. x 

 3 "2 mm. Behind the canine is a diastema of 7 mm., followed by 

 four molars which together measure 9 mm. They are small and 

 rounded, and about equal in size. 



Cyniscodon in the general structure of the jaw resembles most 

 the small Gorgonopsians such as -sElurosaurus, but differs from 

 them in having a loose symphysis, and in being smaller than any 

 of the known Gorgonopsians. 



The specimens described by Lydekker have the Register 

 number 49404, but as this is now seen to include two different 

 animals the number 49404 will be retained for the type of 

 Cyniscodon lydekkeri, and the small Dicynodon skeleton will be 

 numbered 49404 a. 



Cerdognathus greyi, gen. et sp. n. (Text-fig. 5.) 



The type of this new genus and species is an imperfect lower 

 jaw obtained by Sir G. Grey at Klippoort, in the Cradock district. 

 The left dentary is nearly complete, and there is a fragment 

 of the surangular and of the angular. The front portion of the 

 dentary is represented only by an imperfect impression. 



The contour of the jaw is unlike that of any previously 

 described form. There were probably four incisors, of which 

 there are preserved only the obscure impressions of two. There is 

 a single moderately large canine and five molars. The molars 

 are small and uncusped. The whole dentary is unusually straight, 

 there being no great deepening of the symphysis, and the 

 coronoid process does not rise greatly from the line of the 

 horizontal ramus. From the symphysis to the point where the 

 dentary meets the upper border of the surangular, the measure- 

 ment is probably about 105 mm., and the depth at the last molar 

 about 15 mm. From the upper side of the surangular to the 

 notch in front of the descending wing of the angular the 

 measurement is 23 mm. 



