SOUTH AFRICAN THEROCEPHALIAN REPTILES. 



353 



the canine, where the maxilla passes up to form the vault of the 

 palate, the palatine lies closely against the maxilla except where 

 there is a large foramen, presumably for a nerve. Haughton 

 figures two foramina in a corresponding structure in Akido- 

 gnathus. Near the njiddle of the palatine is an oblique ridge, 

 which probably supported a soft palate and carried the internal 

 nares back to the anterior pterygoid region. The inner side of 

 the palatine articulates with the prevomer and the pterygoid. 

 The greater part of the posterior border forms the margin of the 

 suborbital vacuity. Further out is a moderately large articulation 

 with the ectopterygoid. 



The ectopterygoid is an irregularly shaped bone, not unlike a 

 bird's quadrate. Anteriorly it articulates with the palatine; 

 extf^rnally with the maxilla and probahly with the jugaL Be- 

 tween the maxilla and the ectopterygoid is a small foramen. 

 Posteriorly tliere is a large flat articulation between the ecto- 

 pterygoid and the pterygoid. In front of the pterygoid process 

 there is an even larger ectopterygoid process. AH the outer side 



Text-figure 8. 



Palate of MoscJ/orJiinus kitchhigi Broom. 



of the suborbital vacuity is formed by the ectopterygoid and 

 much of the posteiior. 



Only the anterior portions of the pterygoids are preserved. 

 From the broad deep lateral portions which articulate with the 

 ectopterygoids they pass forward only a comparatively short 

 distance to meet the pre vomers and the palatines. In this reoion 

 each pterygoid sends down a thin vertical plate, which, uniting 

 with its neighbour, forms a median keel. 



In general structure the palate agrees more closely with that in 

 Akidognathus, though the proportions of the bones are very 

 difFerent. In Akidognathus the pre vomers have, as in Aloscho- 

 rhinus, a suture between them, and yet are anchylosed in front. 

 Moschorhimts differs in having no interpterygoid vacuity so far 

 as can be seen in the specimen. If one be"^ present, it must be 

 much further back than in AkidognathiLS. 



