110 Transactions of the South African Philosophical Society. 



than to the Ehynchocephalians. This latter opinion has had the 

 support of Smith-Woodward, Osborn, and Williston, and until re- 

 cently I favoured the same view. In the Plesiosaurian skull 

 there is, as in the Anomodonts, a single large temporal fossa, and 

 the superficial resemblance of the two skulls is thus great. When 

 we study the structure of the fossae, however, we find reasons for 

 doubting if the fossae are really homologous. One of the peculiari- 

 ties of the temporal fossa in the Anomodonts is that the inner wall 

 is formed by the post-orbital and squamosal bones. Occasionally 

 these bones are not fully developed and the parietal forms a part of 

 the wall, but it is manifest that the primitive fossa has been between 

 the post-orbital, squamosal, and jugal bones. In the Plesiosaurs the 

 fossa is apparently a fossa between the post-orbital, parietal, and 

 squamosal bones. It thus seems probable that the two fossae, 

 though apparently so similar, are really distinct in origin. 



The structure of the pectoral arch in the Plesiosaurs has given 

 rise to much dispute. There is now, however, little difference of 

 opinion that the arch is composed of a large coracoid, a short scapula 

 with a peculiar anterior and inner process, and a short clavicle and 

 an interclavicle. In different genera the development of the various 

 parts is very different, but in no known genus is there any trace of 

 a distinct precoracoid bone. For the peculiar function of the Plesio- 

 saurian fore limb a precoracoid would have been useful, and had it 

 been present in the ancestor it would have been retained. The pre- 

 coracoid bone is typically present in the Cotylosaurs and in the early 

 Diaptosaurs. It is retained in the Anomodonts and Theriodonts 

 because it is useful for the attachment of the muscles in those 

 animals in which the body has. been well supported off the ground 

 by the fore limbs. In the lizard-like animals where the body rested 

 mainly on the ground the need for a precoracoid disappeared, and at 

 a very early stage in the lizard-like group the ossified precoracoid is 

 no longer met with. Thus we find even in the very primitive 

 Palceohatteria no ossified precoracoid, and it never again reappears 

 in any of the typical reptilian groups. The need for it reappears as 

 in the Plesiosaurs, but some other arrangement has to be effected. 

 As the precoracoid would not have disappeared had the need for it 

 remained, we may safely conclude that the Plesiosaurs are descended 

 from a land animal of lizard-like habits, and which had lost its 

 precoracoid. 



If we examine the pelvic arch of the Plesiosaurs we find the 

 evidence pointing in the same direction. The pubis and ischium of 

 the Plesiosaurs have been described as plate-like, but they are only 

 secondarily plate-like, while the bones in Mesosaurus are primarily 



