98 OX THE CLASSIFICATION OF THE THEItlODONTIA. 



and Deinocephalia have been independently acquired by thes« 

 two groups. 



The study of Thecodont structure and evolution, which is the 

 content of this paper, thus leaves us still without any satisfactory 

 classification of that group, but in its establishment of a series of 

 evolutionary trends, which persist throughout the history of the 

 Anomodontia, has I hope laid a solid foundation on which a 

 natural arrangement may in future be built up when increased 

 knowledge allows an examination of the " adaptive radiation " of 

 the order to be entered on. Meanwhile, the forms whose skulls 

 are well known fill in with considerable completeness the great 

 morphological gap which exists between the Lower Permian 

 Pelycosauria and the Lower and Middle Triassic Oynodonts, and 

 in this way enable us to understand the material steps in the 

 evolution of almost all the structures of a Cynodont skull from 

 those in so primitive a reptile as Seymouria or, indeed, in the 

 still more primitive Embolomerous amphibia. Dealing as' it does 

 with many diverse faces of the subject with which it is concerned, 

 this paper does not lend itself to summarisation, but the point of 

 widest interest brought out in it is undoubtedly the demonstration 

 of the occurrence of the same evolutionary trends in so many 

 allied branches in the Pelycosaurs, Gorgonopsids, Therocephalia, 

 and Deinocephalia ; and the fact that the changes brought about 

 in accordance Avith these trends often serve an adaptive purpose 

 and appear to depend on mere mechanical necessities. 



My thanks are due to the Percy Sladen Trustees for defraying 

 part of the expenses of my visits to South Africa and Texaf 

 To Dr. A. Smith Woodward and Dr. 0. W. Andrews I am 

 indebted for many facilities and much kindness in the British 

 Museum, and to the Department of Industrial and Scientific 

 Research I owe the stipend which has enabled me to carry out 

 this work. 



