MESOSAFEIA OF SOUTH AFRICA. 



601 



large coracoid, and a transverse element anterior to it which he 

 regards as either clavicle or interclavicle.^ I regard the bone as 

 either the episcapula if it is blended vs^ith the scapula, or clavicle if 

 it is a distinct element (as appears more probable). 



If the transverse expansion seen in the Paris type of Mesosaurus 

 is the same bone, its form is imperfect, but it is in the same 

 position as the lateral crescentic bone of Stereosternum. There is 

 nothing in the Cape Town Mesosaurus which corresponds in form 

 with these bones in Stej'eosternum, And the shoulder-girdle in the 

 two types seems to be unlike, because the coracoids in the Brazilian 

 genus met (as shown by the thickened margin) in the median line, 

 while in Mesosaurus there seems to have been a squamous overlap 

 as in Monotreme Mammals, and as the coracoid cartilages overlap 

 in Triton and Salamandra. This condition, so far as I am aware, is 

 not otherwise suggested by remains of fossil reptiles. There is also 

 a possible resemblance to Salamanders in the fact that the scapula 

 and coracoid are not separable, though the Cape Town Mesosaurus 

 appears to indicate a suture. 



Pig. 5. — Restored outline of slioulder-girdle of Mesosaurus. 



c, clavicles ; sc, scapula ; cr, coracoid. 



The British Museum specimen of Stereosternum (R. 536) has the 

 humerus 2*6 centimetres long. In E,. 537 it is 3*6 centimetres 

 long. It is very like the humerus of Mesosaurus. In all these 

 types the bone would be separable only by generic characters from 

 the humerus of the Edentate Mammal Megalonyx, if the epiphyses 

 in that genus were removed. The ulna is Ig centimetre long, 

 and the radius a little shorter. The carpus is badly preserved. 

 Prof. Cope states that it includes a radiale, a large intermedium, 

 a small ulnare, a large centrale, and four distal carpals. The distal 



^ Op. et loc. supra cit. 



