346 



DR. R. BROOM ON SOME NEW 



they are not ancliylosed. Eacli has an ascending thin plate and 

 a short descending ph^te, and from each bone a transverse plate 

 passes outwards which approaches and doubtless further back 

 meets the )3alatine. The pa.la.tines as seen hei'e in section are 

 slender curved bones which below meet the maxillaries, and on 

 passing inwards and then upwai-ds again approach and possibly 

 meet the upper borders of the maxillaries. 



The prevomers extend backwards a considerable distance 

 behind the plane of the last molars and meet the anterior ends of 

 the pterygoids. The palatines form the greater jiart of the bony 

 roof of the mouth, passing about as far back as the posterior end 



Text-fienre 3. 





Palatal surface of skull of Ictidosuclius longiceps Broom. 



of the maxilla. There is behind the palatine a large suborbital 

 vacuity as in Scylacosaiu-us, and, as in that genus, this vacuity is 

 bounded externally by the ectopterygoid. The ectopterygoid is 

 moderately slender, but in front of the pterygoid pi'ocess which 

 lies along the inner side of the mandibles it has a well-developed 

 descending process, which adds much support to the pterygoid 

 process. 



The pterygoid is fairly similar to that already described and 

 figured by me in Scf/Iacosaurns, and possibly some of the apparent 

 differences may be due to certain features being preserved in the 

 pi-esent specimen which were lost in the other. As in most 

 primitive i-eptiles, there is an anterior process which, passing 



