1088 CLASS REPTILIA. 



most complete among existing forms in Chelone and Dermochelys, 

 where the parietal joins the squamosal. In such cases, in addition 

 to the inferior temporal arcade formed by the quadrate, quadrato- 

 jugal, and jugal, as in fig. 1007, there is also a superior arcade con- 

 stituted by the squamosal and postfrontal. Whereas, however, such 

 arcades in the Crocodilia (fig. 1089) are separated by the infratem- 

 poral fossa, in the Chelonia they are in immediate contact. 



The feet may either have all the digits free, or enclosed in a com- 

 mon integument to form paddles. 



The humerus of existing Chelonians is a very peculiar bone char- 

 acterised by its extremely prominent globular head ; but in certain 

 Jurassic forms {e.g., Acichelyidce) this head was much smaller, so that 

 the bone departs less from a normal type. On either side of this head 

 there is a projecting ridge, of which the radial, or preaxial, one (generally 

 termed the lateral process) corresponds with the deltoid crest of the Cro- 

 codilian humerus ; while the ulnar or postaxial (mesial) process repre- 

 sents the inner tuberosity of the same. In all Pleurodiran Testudinata 

 the radial process is comparatively small, and the ulnar process placed 

 in the same transverse line as the distal surface of the bone. In many 

 Cryptodira, however, and more especially in the land Tortoises, the 

 radial process forms a thin plate extending towards the ventral aspect, 

 and the ulnar process becomes twisted round to the same aspect ; thus 

 causing the pit between the two processes to form a narrow funnel-shaped 

 channel instead of being very broad and open. At the same time the 

 shaft becomes extremely curved. In the marine Turtles, where the shaft 

 is nearly straight, the radial process tends to become aborted, and to 

 attain a position more or less below the head. In the Athecata, which 

 likewise have a nearly straight humerus, the radial process, while de- 

 scending on the shaft, tends to an excessive development. 



In time this order dates from the Upper Trias ; and it had 

 attained great development in the Upper Jurassic, from which date 

 it appears to have gone on increasing till the later Tertiary. 



Considerable diversity of views obtains as to the classification 

 of the Chelonia, but according to the system now followed in the 

 British Museum it may be divided into the two suborders Athecata 

 and Testudinata. 



Suborder i. Athecata. — This group contains those forms which 

 have been usually regarded as showing the nearest approximation to 

 other Reptiles, and therefore representing the most generalised type 

 of the order. Dr Baur, however, takes the opposite view, and re- 

 gards them as the most specialised group, which has tended to more 

 or less completely lose the carapace. Before, however, a decisive 

 opinion can be given on this question it must be determined whether 

 the absence of a bony connection in this group between the parietals 

 and pterygoids is to be regarded as an acquired or as an original 

 feature. It may be observed that Dr Baur regards the group as 

 closely allied to the Chelonidce, but if the undermentioned Triassic 



