xv. 4 SYNAPTOSAURS 399 



plates. All of these forms are best considered as aberrant crypto- 

 dirans. 



We can trace the history of the Chelonia back rather satisfactorily 

 to the Triassic, but unfortunately there is little to show how they 

 evolved to that stage from some Carboniferous cotylosaurian ances- 

 tor. * Eunotosaurus from the Permian of South Africa had a small 

 number of vertebrae, with very broad, expanded ribs. This perhaps 

 suggests some affinity with Chelonia, though in the latter the ribs 

 themselves are not expanded. There is therefore little to tell us how, 

 when, or why one of the early reptilian populations shortened its 

 bodies and covered them with armour for protection against the 

 hazards of the land they had recently invaded. 



4. Subclass *Synaptosauria 



Order *Protorosauria 



All the Synaptosauria characteristically possessed a single temporal 

 fossa in the upper or parapsid position. The earliest forms were small 

 terrestrial lizard-like creatures such as *Araeoscelis from the lower 

 Permian (Fig. 228). A few more specialized forms, including the 

 remarkable Triassic * T any stropheus with, a long neck and short body, 

 are known, but the protorosaurs seem never to have been an impor- 

 tant element of the early reptilian fauna. Their relationships are not 

 well understood but it is possible that they gave rise to the sauro- 

 pterygians. The theory that the Squamata were derived from proto- 

 rosaurs by the emargination of the lower temporal region is now held 

 to be unlikely. 



Order * Saw o pterygia 



This was a very successful line of marine reptiles, extending from 

 the Trias to the end of the Cretaceous. The earlier nothosaurs, such as 

 *Lariosaurus (Fig. 228) from marine Triassic deposits, were small 

 (3 ft long) and had a long neck, and limbs partly converted into 

 paddles. The upper temporal fossa was enlarged and the nostrils lay 

 rather far back, as in many water reptiles. 



All of these features were further developed in the plesiosaurs of 

 the Jurassic and Cretaceous, such as *Miiraenosaurns (Fig. 228). In 

 some the neck became very long, presumably for catching fish; 76 cer- 

 vical vertebrae have been recorded. In others the neck was shorter 

 and the skull longer. The limbs were developed into huge paddles, 

 the ventral portions of the girdles being large for the attachment of 

 muscle masses inserted on the flattened humerus and femur. The 



