264 Transactions of the South. African Philosophical Society. 



portion of the squamosal. The inner part of the bone is well 

 developed, and probably articulates with the ex-occipital. It also 

 gives articulation to the bone which I regard as the tympanic. 

 The articular surface of the quadrate looks forwards and slightly 

 downwards. It appears to me to be convex from above downwards 

 and concave from within outwards. 



The occipital bones are not very satisfactorily preserved, and no 

 sutures can be detected. Though in general the occiput resembles 

 that of Dicynodon there are one or two important differences. The 

 condyle is strikingly different from that of any reptile hitherto 

 described. It is of large size and passes back about 25 mm. from 

 the foramen magnum. Instead of being convex, as in almost all 

 reptiles, it is concave in the centre. At the sides it is fairly flat, 

 while the lower third is distinctly convex. Though no sutures can 

 be made out, it is probable that the lower convex part is formed by 

 the basi-occipital bone and that it articulates with the large lower 

 piece of the atlas. The side parts which appear to articulate with 

 the arch of the atlas are probably formed by the ex-occipitals. 



The foramen magnum is of small size and about twice as high as 

 broad. The supra-occipital bone, or at least that portion of the 

 occiput above the foramen magnum, is directed almost vertically 

 upwards. 



The lateral part of the ex-occipital is well developed and is fairly 

 similar to that of Dicynodon. Instead, however, of being directed 

 mainly outwards, it passes in Endothiodon almost as much back- 

 wards as outwards, so that the posterior point of the ex-occipital is 

 as much behind the condyle as the condyle is behind the foramen 

 magnum. There is apparently a pretty large foramen between the 

 ex-occipital and what is possibly the supra-occipital. 



The bone which lies below the ex-occipital, and which I regard 

 as the tympanic, is considerably stouter than in Dicynodon and 

 Oudenodon. Its relations, however, appear to be exactly similar 

 in that it lies between the quadrate and the par-occipital process. 



The basi-occipital bone so far as preserved agrees very closely with 

 that in Dicynodon. 



There is a small part of the basi-sphenoid preserved which shows 

 that it sent a plate-like process down in front of each par-occipital 

 process exactly as in Dicynodon. 



The mandible has lost the posterior halves of both rami, but the 

 anterior part of the jaw is well preserved. The fused dentaries 

 and the united splenials are present in almost perfect condition, 

 and there are also seen portions of the angulars and part of the left 

 sur-angular. A small portion of the left articular is attached to the 



