Observations on the Structure of Mesosaurus. Ill 



so. The pubis and ischium in the Plesiosaurs, though greatly flat- 

 tened out, are probably but modifications of the radiating type seen 

 in Sphenodon. Had the land ancestor of the Plesiosaurs had a 

 plate-like pubis and ischium it would have been retained in the 

 aquatic descendants, where there is clearly a need for the plate-like 

 form, but as the pubis and ischium are only secondarily flattened it 

 is evident that the land ancestor of the Plesiosaurs had a pelvis 

 somewhat resembling the Sphenodon type. 



The Plesiosaurian skull is not likely to have been derived from a 

 type like that of Sphenodon, though the temporal vacuity is probably 

 homologous with the upper vacuity of the Sphenodon skull. It 

 seems more probable that the Plesiosaurs are sprung from a land 

 ancestor somewhat resembling Sphenodon in general characters, but 

 with only a single supra-temporal fossa. In Permian times there 

 must have been a very large number of land types of which we know 

 little except their specialised Triassic descendants. The Phyto- 

 saurians, the Ichthyosaurs, the Plesiosaurs, the Proterosaurs, the 

 Pelycosaurs, and the Proganosaurs, all point to there having been a 

 large group of land reptiles resembling in many w T ays the Rhyncho- 

 cephalians, but much more primitive in many characters. For this 

 large group Osborn's name of Diaptosauria seems the most suit- 

 able. The more primitive members of it had the skull roofed as in 

 the Cotylosaurs, and retained the precoracoid and the plate-like 

 pubis and ischium. Procolophon is evidently a descendant of these 

 earlier types, and so probably is Mesosaurus. The higher members 

 of the group, owing to their lizard-like habits, lost the precoracoid, 

 and had the pelvis modified into a tri-radiating type. First an 

 upper, then a lower temporal fossa seems to have been formed, 

 though possibly in some types the order may have been reversed. 

 From some of these higher types with only the supra-temporal fossa 

 in all probability sprang the Plesiosaurs. 



It would thus seem probable that Mesosaurus and the Plesiosaurs 

 are descended from different and rather dissimilar land animals, 

 which, as being members of the same large Super-order, the Diapto- 

 sauria, had, however, a good many characters in common — espe- 

 cially the Diapsidan digital formula and well-developed abdominal 

 ribs. A few other characters which the groups have in common 

 appear to be due to similar specialisations to suit like habits. 



Boulenger in his forthcoming paper on Telerpeton, of which he 

 has kindly sent me a proof, objects to Osborn's names for the sub- 

 divisions of the Eeptilia — Diapsida and Synapsida — owing mainly 

 to Osborn's inclusion in the latter of the Plesiosaurs and Chelonians. 

 While I am inclined to think that Boulenger is right in removing 



