902 DR, R, BROOM ON THE 



the prefrontal and forms part of the orbital inai-gin. Exteri^ially 

 it gives a large articulation to the jugal and a small articulation 

 to the squamosal. Internally it overlaps the parietal for only a 

 very short distance. 



The jugal is relatively considerably larger than in Nythosauriis. 

 Anteriorly it meets the maxilla and lachrymal, but extends further 

 forvt^ard than in either Nythosatbrus or Bauria. Immediately below 

 the oi'bit there is a small but very distinct tubercle. The 

 ascending process of the jugal is unusually broad and forms a 

 large articulation with the postorbital, the two together forming 

 a very strong postorbital arch. The posterior portion of the jugal 

 extends to the articular region. 



The parietal is a, narrow bone which forms a low median crest. 

 There is a small pineal foramen. 



The squamosal is the largest bone in the skull, with the excep- 

 tion of the dentary. The inner and posterior portion forms 

 nearly the whole of the back wall of the temporal fossa, and lias 

 a large articulation with the parietal. This back portion of the 

 squamosal is veiy thin and is closely united with the flat upper 

 expansion of the opisthotic. The squamosal forms the outer wall of 

 the lateral occipital foramen. Inferiorly it meets the exoccipital 

 and on passing outwards supports the small quadrate. The 

 zygomatic portion is much larger than in JSFythoscmrus, articulating 

 with the whole of the upper side of the posterior limb of the 

 jugal and meeting the postorbital. There is a deep groove along 

 the middle of tlie back part of the zygomatic portion, which 

 curves downwards and inwards, and, most probably, the pos- 

 terior part was for the suj)j)ort of the external auditoiy canal. 



The occiput is best known from the specimen of Cynognaihis 

 herryi in the S. African Museum, which though imperfect shows 

 the sutures very distinctly. The large occiput figui'ed by Seeley* 

 (p. 130) and doubtfully referred by him to G. herryi, is, in my 

 opinion, considerably too large. It also differs from the known 

 occiput of G. herryi in the shape of the foramen magnum, the 

 slope of the exoccipitals, and the moulding of the interparietal 

 region. It is pretty clearly not the occiput of G. crater onottis, 

 while it is much too large to be that of G. platyceps. As it thus 

 seems to belong to a new species, it may appropriately be named 

 Gynognathus seeleyi. 



The occiput proper (PI. XLYI. fig. 2) is made up of seven 

 bones, or, if we include the squamosals, nine. About a quarter 

 of the occiput proper is formed by a median bone, which is 

 apparently the interparietal. I have not seen any specimen which 

 enables me clearly to differentiate this bone from the parietal in 

 front, but, judging by analogy and by the direction of the fibres 

 of the bone, the probability seems much in favour of its being a 

 distinct interparietal. 



On either side of the interparietal is articulated a large bone, 

 which is apparently the opisthotic. Its occipital portion is com- 

 paratively thin and to a considerable extent covered in front by 



* Phil. Trans. Vol. clxxxvi. B. (1895). 



