On the Classification of the Reptilia. 



269 



Belated also to these Sauropsidan reptiles would appear to be the Aetosauria 

 or Pseudosuchia (v. Huene, 16), the Barasuchia (McGregor, 22), the Simcedo- 

 sauria (Lernoine, 20 ; Brown, 10), and the Bhynchosauria (Huxley, 18), since 

 they all have upper and lower temporal foramina and a hook-shaped fifth 

 metatarsal. On the other hand, the Procolophonia, believed by Broom to 

 have Bhynchocephalian affinities, and placed by Osborn in the Diapsida (26), 

 are now known to have neither temporal fossae nor a modified metatarsal. 

 For the present they may be placed in our Brotosaurian group, and are 

 probably related to the Baraiasauria and Cotylosauria as maintained by 

 Boulenger (7) and Seeley (30). 



Pelycosauria. — Founded by Cope in 1878 for certain North American fossil 

 reptiles, such as Dimetrodon and Clepsydrops, supposed to be allied to the 

 Bhynchocephalia, the systematic position of this order has been a subject 

 of much controversy. Cope soon changed his mind and, struck by the 

 resemblance of the Pelycosauria to the South African Theromorpha, believed 

 them to have mammalian affinities. In 1897, however, Baur and Case (4) 

 thought they had proved the existence of upper and lower temporal fossae, 

 and later the Belycosauria were included in the Diapsida by Osborn (26). 

 Subsequently Case in his valuable revision of the Belycosauria (11) still 

 insisted on their close relationship with the Bhynchocephalia, Proganosauria, 

 and Protorosauria. Now the Proganosauria (Mesosaurus and Palaeohatteria) 

 we have already seen are probably not Diapsidan, and Broom has recently 

 brought forward strong evidence that the skull of the Pelycosauria is really 

 built on the Synapsidan plan (9). Cope's later opinion is thus borne out, and 

 is certainly more in agreement with the fact that the foot has five normal 

 metatarsals, as described and figured by Case (11) and Osborn (27), see fig. 3,C. 



Arceoscelis. — Under the name Arseoscelis Williston has recently described 

 (35) an interesting Permian reptile of lizard-like shape. It is provided with 

 single upper lateral temporal foramen and a broad arch below. These and 

 other characters lead him to suppose that the Lacertilian skull has been 

 evolved from some such form by the narrowing of the arch and the loosening 

 of the quadrate. But if the figure given of the hind foot (36, fig. 5K), and the 

 restoration (36, fig. 7) correctly represent the fifth metatarsal as an elongated 

 straight bone, there can be little doubt that Araeoscelis is not closely related 

 to any modern reptilian order. Bather would it seem to belong to the 

 Theropsida ; unless perchance it is an early representative of the Sauropsida 

 before the metatarsal had become modified. 



Plewrosaurus. — Watson, who has recently reclescribed the skull of 

 Pleurosaurus goldfussi, V. Meyer (33a), supports Boulenger's contention (5) 

 that it is not Bhynchocephalian, restores it with one temporal foramen and 



z 2 



