io/o 



CLASS REPTILIA. 



never develop plates to form a floor to the nasal passage, so that 

 the posterior nares also open directly into the mouth by horizontal 

 apertures (fig. 991). 



This order ranges in time probably from the Permian and cer- 

 tainly from the Trias to the Upper Chalk, and we are enabled to 

 trace the gradual evolution of the specialised marine forms from 

 those less widely separated from a normal type. All these reptiles 

 appear to have been carnivorous. 



Family MesosauriD;E — The genus Mesosaurus, originally de- 

 scribed from the Karoo system of Griqualand in South Africa, 

 which is probably of lower Mesozoic age, but subsequently found 

 in beds of uncertain age in Brazil, and described under the name 

 of Stereo sternum, includes small reptiles regarded by Dr Baur as 



Fig. 989. — Longitudinal section 

 of a Sauropterygian humerus ; 

 from the Kimeridge Clay ; one- 

 sixth natural size, a, Proximal, 

 b, Distal epiphysis ; c, Shaft. 



Fig. 990. — Ventral aspect of the 

 left pectoral limb of Mesosaurus 

 tenuidens ; from the Karoo system 

 of Griqualand. ent.f, Entepicon- 

 dylar foramen of humerus; r, Ra- 

 dius J u, Ulna. 



constituting a distinct order — the Proganosauria, — but which 

 appear to be so closely related to the Nothosauridce, that there can 

 be little, if any, hesitation in including them in the same order. 

 The Brazilian form was originally referred with some hesitation by 

 Professor Cope to the Amphibia. One of the most peculiar feat- 

 ures of this genus, in which it differs from all other groups except 

 the Amphibia, the extinct Palceo hatter ia^ and perhaps the Chelonia 

 is the separation of the fourth and fifth tarsalia, so that each meta- 

 tarsal articulates with a distinct tarsale. The centra of the vertebrae 



See Rhynchocephalia, infra. 



