SKULLS OF Cir>rODONT REPTILES. 895 



across a considerable number of new Cynoclonts, mostly collected 

 by Mr. A. Brown, and also to add a good many facts to our 

 knowledge of the anatomical structure. Probably the most 

 important of the recent finds has been the discovery by myself 

 of the nearly pei'fect skull which I have called Bauria cynops. 



Of all extinct reptilian groups there is probably no one of 

 greater interest than the Cynodontia, Many years ago Owen 

 recognised the I'emarkiible mammalian characters in the Permian 

 and Triassic South Afi'ican reptiles, and though the Cynodonts 

 were so little known, he ventured to suggest tliat certain of the 

 Anomodonts were fairly closely allied, and perhaps ancestral, to 

 the Monotremes. Cope held much the same view. When the 

 very much more mammal-like Cynodonts were described by 

 Seeley, many recognised in this higher group the looked for Sauro- 

 Mainmalia. Osborn has been the chief advocate of this opinion. 

 Seeley himself, though at first inclining to it, afterwards came to 

 the conclusion that the Mammals were in no way nearly related 

 to the Cynodonts, but sprang from some unknown ancestor that 

 lived in Devonian or Silurian times. 



If the Cynodonts are not nearly related to Mammals, the group 

 is still of great interest as showing a marvellous parallelism with 

 the Mammals in skull, teeth, girdles, limbs, and digits ; but if, 

 as all recent Avork seems to indicate still more clearly, the mam- 

 malian ancestor was probably a Cynodont, the group becomes 

 vested with an interest altogether unique, and everything bearing 

 on it becomes worthy of the most careful study. I have fortu- 

 nately been able to examine every known skull, and in the present 

 paper I give the results of my researches. As the paper is 

 morphological rather than systematic, I propose to give a detailed 

 account of the skulls of only the principal Cynodont types, and 

 to consider more fully those points which seem to have a special 

 bearing on the question of mammalian descent. 



Bauria. 

 (PI. XLYI. figs. 6, 7, 8, and text-figs. 168, 169.) 



Though Bauria ci/nops occurs in the same horizon as Cyno- 

 gnathus, it is the most primitive Cynodont at present known, and 

 may be regarded as the type of a distinct family which may be 

 called the Bauridse. 



As I have just recently, at considerable length, described the 

 only known skull of Bauria cynops, it will be unnecessary here 

 to do more than supplement that description in a few details 

 and to consider its relationships to the other known Cynodonts, 

 the Therocephalians, and the Mammals. 



Further development and examination of the skull has revealed 

 one or two points not previously noted. Under the nostril and 

 forming not only its floor but covering a considerable part of 

 the premaxillaiy is a large septomaxillary bone. The lachrymal 

 and prefrontal bones cannot in the specimen be clearly sepaiated 



