494 DR. E. BROOM ON THE STRUCTURE [-^P^^- 19; 



Though the known Theriodont genera difler greatl}- in their 

 dentition, it will be observed that in the structure of the lower 

 jaw and quadrate region they agree with each other very closely, 

 and form by themselves a very distinct order. Much confusion 

 has been caused in the past by there having been placed with the 

 Theiiodonts a large number of other forms which are not at all 

 nearly related to them, I refer to such for^s as ^^hn'osaurus, 

 and probably the majority of Owen's types with a similar dentition. 

 While very little was known of these forms except that they had 

 the dentition specialised as incisors, canines, and molars, it was but 

 natural to place them with Galesau7'us ; but since the structure of 

 the skull of these " primitive Theriodonts " has become known, it 

 has been necessary to place them in a distinct order, which I 

 have named Therocephalia (4). The Therocephalians differ from 

 the Theriodonts in having, besides a large number of other 

 distinctive features, a palate of the Rhynchocephalian type and a 

 single occipital condyle. In the lower jaw the dentary, though 

 possessed of a large coronoid process, is relatively much smaller than 

 in the Theriodonts (see text-fig. 100, p. 496), while the angular 

 is of large size, and the surangular and articular moderately well- 

 developed. The Theriodonts are probably direct descendants of 

 the Therocephalians, but the gap between the groups is very 

 considerable. 



In Dicynodon and Udenodon the lower jaw is essentially similar 

 to that in the Therocephalians, though the coronoid process is 

 quite rudimentary. As the Dicynodonts are probably, like the 

 Theriodonts, also descended from the Therocephalians, the loss of 

 the coronoid process is probably, as in the Monotremes, connected 

 with the loss of the incisor teeth. 



The only other group to Avhich the Theriodonts ai'e closely 

 related is the Mammalia. In the present paper I shall avoid any 

 lengthy discussion of the question of the oiigin of mammals, but 

 I wish to point out how nearly related the Theriodont jaw is to that 

 of the mammal. 



The mammal differs from the reptile in having the jaw formed 

 entirely of one bone — the dentary, and in the dentary articulating 

 with the squamosal with apparently no quadrate. What has been 

 the fate of the quadrate is a question which has received a numbei* 

 of very different answers. Owen i-egai-ded the tympanic bone as 

 the mammalian equivalent of the quadrate, and this view has the 

 support of Gadow (5) among others. Huxley and many others 

 have argued in favour of one or othei- of the auditory ossicles 

 being the mammalian quadrate, and the view that the incus is 

 the quadrate appears to be the one chiefly supported by compara- 

 tive anatomists at the present day. This latter view, though 

 having the strong support of Kingsley (6) and Gaupp (7), has 

 recently been veiy severely criticised by Gadow (8). Gadow not 

 only shows that the mammalian auditory ossicles are together 

 homologous with the columella and extra-columella of the Sauro- 

 psida, but has shown how impossible it would be for an animal to 



