627 



cannot be placed in any of our previously known orders. It shows 

 affinities with the Cotylosaurs and the Pelycosaurs, but also some 

 striking resemblances to the Lacertilia. For some years I have held 

 the opinion, also shared by Huene and Williston, that the lizards are 

 quite as primitive as the true Rhynchocephalians. In Yaranosaurus 

 we have a type which though neither a lizard nor the ancestor of the 

 lizard yet has sufficient resemblance in skull, girdles and limbs to 

 suggest that it has some affinity with the ancestral lizard types. When 

 WiLUSTON has more fully studied the skull we shall probably have a 

 nearer idea of the affinities of Yaranosaurus, but wherever it be 

 ultimately placed its shoulder girdle seems to show us how the single 

 ventral element of most reptiles arise from the ancestral pair. 



There is a large expanded scapula with a very broad lower end 

 which articulates with a single large ventral element. If we compare 

 this shoulder girdle with that of Ophiacodon, of which though not yet 

 described Prof. Williston has kindly sent me a sketch, we see a 

 scapula very similar to that of Yaranosaurus and articulating ven- 

 trally with a large anterior element and a small posterior one. There 

 is here no doubt that the small posterior element is the true coracoid 

 and the large anterior the precoracoid. In Yaranosaurus the large 

 precoracoid remains but the small coracoid is lost. Williston is quite 

 satisfied that there is no trace of a posterior ossification though there 

 may have been a cartilaginous coracoid. The humerus unquestionably 

 articulates with the scapula and the precoracoid. 



In Seymouria Williston shows that a very similar condition to 

 that in Yaranosaurus is present, the posterior element or true coracoid 

 being absent or at least unossified. In most Cotylosaurs, however, 

 they appear to be a distinct coracoid and precoracoid though in some 

 owing to the fusion of the bones it is difficult to be sure of the ele- 

 ments. In the Temnospondylous amphibians Williston believes there 

 is only a single ventral element corresponding to the anterior one of 

 the pair, but this is not always the case. In our large S. African 

 Amphibian allied to Eryops recently described by v. Hoepen under the 

 name Myriodon senekalensis but which probably belongs to the genus 

 Rhinesuchus previously described by me, there are two ventral ele- 

 ments — a large precoracoid with a foramen and a small coracoid. 



The conclusion to which we seem driven is that the anastral 

 forms had two ventral coracoidal elements. The two are retained in 

 many of the later types and can be foUowed through Cotylosaurs into 



40* 



