180 



Having such a form as this, the stapes recalls essentially that 

 of the Rhachitomous Stegocephalians and differs greatly from the 

 CotylosaurUke Captorhinus and, indeed, from all other primitive 

 Reptiles — Anomodontia sensu lato, Pelycosauria, Ichthyosauria — 

 in which the stapes has the form of a massive rod abutting by its 

 distal end upon the quadrate and usually connected also with the 

 end of the paroccipital bar. (In later Reptiles, as far as can be 

 judged from existing forms, this connection of the stapes and 

 quadrate is more or less distinctly repeated in the ontogeny.) 



Text-fig. 1. Kotlassia, middle presacral vertebrae. 



Text-fig. 2. Karpinskiosaurus, 17th and 18th presacral vertebrae, lateral view. 



Text-fig. 3. Same, dorsal view. 



These features of the stapes are of prime importance for deter- 

 mining phylogenetic relations. The connection of the stapes 

 with the quadrate, which we find in the primitive Reptiles, is 

 evidently a reminiscence of the ancient hyomandibulo-quadrate 

 junction characteristic of the fishlike ancestors of the Tetrapods. 

 In the Rhachitomi this connection does not any longer exist. 

 This character alone affords sufficient ground for removing the 

 Rhachitomi from the ancestry of Reptiles and Amniotes gener- 

 ally, the Rhachitomi standing in this respect farther from the 

 fishlike ancestors than the primitive Reptiles do. The same ad- 

 vanced condition of the stapes we find in the Seymourians. 



Hence, the Seymourians are not only to be removed from the 

 order Cotylosauria, but they cannot be regarded at all as an- 

 cestors of the Reptiles. The condition of the stapes character- 

 istic of the primitive Reptiles cannot be derived from that of the 

 Seymourians. 



