598 PROF. H. G. 8EELET ON THE 



being margiaed by a longitudinal ridge, chiefly developed above the 

 posterior zygapophyses. In the anterior half of the vertebrae these 

 ridges converge forward above the pre-zygapophyses. The lateral 

 contour of the neural arch is concave in length-. The pre-zygapo- 

 physial facets are horizontal, transversely ovate facets, above which 

 the neural arch rises abruptly.^ 



The entire length of the fragment of this skeleton preserved is 

 12^ centimetres. I have no doubt that it indicates a new species, 

 but there is no character available for its definition except the 

 unsatisfactory one of relatively stout ribs. 



§ 5. The Relations op Mesosatjetjs with Steeeosternum. 



In 1886 Prof. Cope published his well-known description of 

 Stereostemum tumidum.^ He recognized the close general resem- 

 blance of that type to Mesosaurus ; but since the Brazilian fossil 

 which he figures represents the hinder half of the body, and the Paris 

 specimen only shows the anterior half, no very close comparison 

 could be made. The author observes, " As the dorsal vertebrae [of 

 Mesosaurus] are obscured by matrix, the only point in which actual 

 comparison can be made is the ribs. These are quite identical in 

 the two types . . . The humerus is almost identical, and the carpus 

 is nearly what one would expect to find in the Brazilian form." 

 Shortly afterwards the British Museum acquired two skeletons of 

 this animal from Sao Paulo, one of which is better preserved than 

 that figured by Cope, in showing the entire dorsal region with the 

 fore and hind limbs, as well as some indications of the shoulder- 

 girdle, and important pelvic characters. Dr. Henry "Woodward, 

 P.H.S., had a cast taken from the better of the two slabs, and I 

 studied the remains and found that comparison of the Kimberley 

 and Brazilian specimens suggested the conclusion that Stereoster- 

 num was generically indistinguishable from Mesosaurus^ unless the 

 clavicular arch should be found to separate them. Mr. Lydekker, 

 in his Catalogue of the Fossil Beptilia in the British Museum, Part ii. 

 p. 302, referred the specimens to Mesosaurus^ remarking that " there 

 appear to be no characters by which Stereostemum can be specifi- 

 cally distinguished from the type-species \_Mesosaurus tenuidens]^ 

 There are important regions of the skeleton which have not been 

 compared, such as the pelvis, tbe shoulder-girdle, and the skull. 

 The coracoid figured by Cope amply justifies specific separation, 

 and makes generic distinction not improbable, though Mr. Lydekker 

 would refer the coracoid to some other type of animal ; but it 

 has enough in common with the coracoid of Mesosaurus as now 



^ There are imperfectly preserved doubtful indications of transverse processes. 

 In the Cape Town fossil these possible plates (if they are not division-planes in 

 the matrix) are hidden beneath the ribs, but appear to be triangular and di- 

 rected outward and backward for nearly half the length of the ribs. A similar 

 appearance is seen in the Grahamstown specimen, where the transverse process 

 appears to be broken. It is more slender than in the other example. These 

 indications are imperfectly displayed, and better specimens must be obtained 

 to show how far the indications may be relied upon. 



2 Proc. Am. Phil. Soc. vol. xxiii. p. 7. 



