MES0SAT7RIA OF SOFTH AFRICA. 603 



The most remarkable of these is Neusticosaurus. When describing 

 that type ^ I inferred from the shoulder-girdle that it should be 

 affiliated to the Nothosauria. Now, however, I believe that too 

 little importance was then given to the mode of attachment of the 

 dorsal ribs. This character entirely separates Neusticosaurus from 

 the Nothosauria, and approximates it to the Mesosauria. In all 

 Nothosaurians the dorsal ribs are carried upon exceptionally stout 

 transverse processes, which only differ from those of Plesiosaurus in 

 being more massive and deeper ; while in the Mesosauria there is 

 no trace of this relation. In Mesosaur us, Stereosternum, and Neustico- 

 saurus there is the same barrel-like contraction of the ends of 

 each dorsal centrum, though this is but a. family character. The 

 articular faces of the centrum are conically cupped in the African 

 and Brazilian types, but in Neusticosaurus this surface is flat, as in 

 associated genera, although the neural arch is not anchylosed to the 

 centrum. Indeed, the European Neusticosauridse fall into a family 

 which has the tail short and the neck long, and shows points of 

 affinity to the Nothosauria ; while the African Mesosauria have the 

 neck short and the tail long, and show points of affinity to the 

 Anomodontia. The Nothosauria are nearer in affinity to the Ano- 

 modontia than is consistent with their inclusion in the order 

 Sauropterygia. The chief difficulty in recognizing this relation has 

 been in the apparent differences of the shoulder-girdle. 



In the shoulder-girdle referred to Nothosaurus mirahilis (Miinster) 

 figured by Von Meyer, a small notch is seen in the coracoid, between 

 that bone and the scapula. Internal to the notch the coracoid 

 develops a strong process which terminates abruptly forward in a 

 transverse but oblique line. The notch is in the position of the 

 foramen which in certain Anomodont types occurs at the junction 

 of the precoracoid with the coracoid and scapula. Since an open 

 angle stretches forward between the latter two bones, it is possible 

 that it was occupied by a cartilaginous precoracoid during life, in 

 contact with the internal border of the scapula and the anterior border 

 of the coracoid, which have the aspect of being articular surfaces. 

 Such a condition would make an approach towards the Anomodont 

 type. But in Neusticosaurus there is no trace of the coracoid notch 

 or foramen, and there are no internal articular surfaces in the 

 scapular arch, like those seen in Nothosaurus. 



The Lariosauridae, according to Deecke,^ have the vertebral cha- 

 racters of the Nothosauria. And although there is similarity in 

 form in the humerus to that of the Mesosauria, there is no such ent- 

 epicondylar foramen as characterizes the bone in the genera so 

 grouped. It seems to be transitional between the Nothosauria and 

 the Sauropterygia, and to have no near affinity with the Mesosauria, 

 in so far as detailed comparison can be made. But the development 

 of a sacrum of many vertebrse in Lariosaurus, as figured by Zittei,^ 



^ Quart. Journ. Geol. Soc. vol. xxxviii. (1582) p. 350. 



^ Zeitschr. d. Deutsch. geol. Gesellsch. vol. xxxviii. (1886) p. 170. 



^ ' Handbuch cler Palaeontologie,' Bd. iii. p. 485. 



Q. J. G. S. No. 192. 2 t 



