314 Annals of the South African Museum. 



the maxillae to the bones behind cannot be made out, and it is 

 uncertain whether there is a lachrymal, a pre-frontal, or a nasal. 

 Probably the bone forming the anterior part of the raised rim of the 

 roof of the orbit was the pre-frontal, while that forming the pos- 

 terior part was the post-frontal, but no sutures are visible. Neither 

 can sutures be seen between the frontals and the surrounding bones. 

 The pineal foramen seems to have been just opposite the anterior 

 border of the temporal fossae and probably marks the anterior limit 

 of the parietal bones. These formed a high, narrow, sagittal crest 

 between the temporal fossae, but in this specimen its summit is 

 broken away. Posteriorly these bones widen out and project back- 

 wards behind the level of the foramen magnum ; how far the lateral 

 expansions of the parietals are overlapped by the upper limbs of the 

 squamosals cannot be determined. The ventral process of the 

 squamosal (sq.) can be seen to be closely adherent to the posterior 

 and outer face of the large quadrate (q.) extending far down its side. 

 The zygomatic process seems to have been wide, but the relations 

 of its anterior end to other bones cannot be seen. The quadrate (q.) 

 is relatively very large : its anterior face is concave from side to side, 

 while its posterior face is convex in the same direction. The 

 articular surface for the mandible is imperfect, but must have 

 been very wide, as is shown also by the width of the articular 

 surface of the mandible. The outer face of the quadrate, as already 

 mentioned, was overlapped by the ventral process of the squamosal, 

 while the inner side unites with the quadrate ramus of the pterygoid 

 and probably also with the outer end of the paroccipital process 

 of the opisthotic, though possibly this only supports the quadrate 

 through the medium of the pterygoid. 



The mandible (PI. XVIII., Figs. 2, 3) is only imperfectly preserved, 

 the symphysial region being completely wanting. It seems to 

 have been very strongly built for the size of the skull. The 

 articulation for the quadrate is very broad ; it consists of two con- 

 cave surfaces, the inner being the larger, separated by a slight 

 convexity : both in front and behind it is bordered by a strongly 

 raised rim, so that the articulation must have been a very strong 

 one. On the outer side of the jaw, just beneath the articular sur- 

 face, there is a deep and roughened pit or groove, apparently for the 

 attachment of ligament. The post-articular process is short and 

 broad ; its inner border is rounded, while the outer forms a sharp edge. 

 Its upper portion is formed by the hinder part of the united sur- 

 angular and articular bones (PI. XVIII., Figs. 2 and 3, s. ang. and art.), 

 the lower portion is constituted by the angular (Fig. 3, ang.). In 



