114 Bulletin American Museum of Natural History. [Vol. XXVII, 



case." And yet Kingsley is so impressed with certain comparisons between 

 the Mammalia and the Amphibia in respect to the relations of the quadrate 

 and its supposed homologue the mammalian incus, of the anterior abdominal 

 vein, of the mesenterial structures, of the amnion and trophoblast and of the 

 hair, that he concludes that all the osteological resemblances between the 

 mammals and the mammal-like reptiles of the Permian and Trias are the 

 results of parallel development; and that the mammals have descended 

 from forms "with amphibian affinities but more primitive in structure than 

 any known stegocephalan." He therefore implies (p. 255) that the com- 

 mon ancestors of the Amphibia, Reptilia and Mammalia must be sought 

 in the Devonian. The purpose of the present article is to show that so far 

 as the osteological evidence indicates it is not necessary to have recourse to 

 unknown Devonian amphibians for the ancestry of the Mammalia; but that 

 all recent evidence tends to support the view developed by Owen (1876), 

 Seeley (1888-1895), Osborn (1898-1903) and Broom (1901-1908), namely, 

 that the ancestry of the mammals is to be sought within, or very near to, 

 the Triassic reptilian order of Cynodontia ("Theriodontia" in part). 



I. The Synapsid Reptiles. 



The subject may conveniently be approached by a cursory review of the 

 reptilian orders constituting the subclass Si/napsida of Osborn (1903). 

 These are typically distinguished by their prevailing resemblances to the 

 Mammalia, especially in the possession of a single temporal arch in the skull, 

 and they contrast with the Diapsida, or primitively two-arched reptiles 

 which center around the Rhynchocephalian type. The composition of the 

 Synapsida, as defined by Osborn (1903, pp. 455, 456) is as follows: 



The Diadectids, Chelonia, etc. 



First we may set aside certain highly specialized groups which are more 

 or less widely removed from the ancestors of the mammals. Among these 

 are: (1) The Permian Diadectidse, the original type of the order Cotylosauria 

 Cope (see Williston, 1908). These distinctly foreshadow the Chelonia in 

 many significant characters, as observed by Cope, Case (1905) and Hay 

 (1908, p. 29). (2) The Chelonia themselves. (3) The Placodontia, which 

 Jaekel (1907) has recently shown to include forms resembling the Chelonia 

 in the general configuration of the body, but allied in other directions to the 

 Nothosaurs and Plesiosaiu-s. (4) The Xothosaur-Plesiosaur group, which 

 is considered by some authorities (Williston, 1907) to be derived from very 



