130 Annals of the South African Miisemn. 



the last dorso-lumbar and the first caudal may to some extent aid 

 the true sacral vertebrae in the support of the ilium. The first sacral 

 vertebra in P. serridens corresponds to the second of the four sacral 

 vertebrge described by Seeley in P. haini. 



The upper caudal vertebrae have unfortunately lost the spines, so 

 that the heights are uncertain. So far as preserved, they do not 

 differ greatly from the one figured by Seeley (I. plate 18), and agree 

 pretty closely Wxih. those figured and described in P. haini (II. 1892). 

 Intercentra probably occur as far as the fifth caudal, while about 

 here chevrons begin to be developed. Long chevrons are developed 

 to those caudal vertebrae which I believe to be the 32nd to the 40th, 

 and even the little caudal vertebra from near the point of the tail, 

 and which may be the 47th, has a well-developed chevron. 



The Shoulder Girdle. 



The shoulder girdle is in almost perfect preservation, but it so 

 closely resembles that of P. haini already figured and described by 

 Seeley that a detailed description is unnecessary, while the shape of 

 the different elements will be much more readily understood from 

 the figures given than from any verbal description. Seeley, in his 

 restoration, places the scapula in an almost horizontal position with 

 the precoracoid above the coracoid, and the interclavicle directed 

 mainly downwards. By fitting the head of the humerus accurately 

 in the glenoid cavity it can readily be demonstrated that the position 

 in which Seeley places the shoulder girdle is quite erroneous. In 

 the restoration given in Plate XVI. the humerus is accurately so 

 fitted, and it will at once be seen that the scapula must have been 

 in an almost upright position. 



The coracoid in the skeleton of P. haini has apparently lost nearly 

 the whole of its lower half. In the figure I have given of the 

 shoulder girdle the margins of the precoracoid and coracoid are 

 drawn from the bones of the right side, which are well preserved, 

 and if the coracoid in the skeleton of P. haini be completed, as in 

 the present specimen, the very great resemblance between the two 

 will be manifest. Seeley (II. p. 355) says "the coracoid appears to 

 have formed a wide flattened base upon which the weight of the 

 shoulder girdle was supported." Surely the weight of the shoulder 

 girdle rested on the humerus, and the great strength of the lower 

 part of the scapula and the large glenoid articulation are just such 

 as one would expect in such a heavy-bodied animal. 



