COLOSSOCHELYS ATLAS. 379 



of wticli a large fi-agment is attached to the ribs, and gives 

 the form of the inguinal sinus for the emission of the leg, 

 exactly as in the Testudo Indica. The engroined apices 

 indicate the age to be young, and afford an excellent element 

 of measurement, as the sterno-costal pieces bear a direct 

 ratio to the development of the ribs to which they are 

 attached. Another crucial proof of the size is derived from a 

 comparison of corresponding measurements in Golossochelys 

 and Testudo Indica. (See antea, pp. 364 and 374.) 



The correspondence shown in the symphysized portion of 

 the sterno-costal pieces is borne out by other pieces all round 

 the rim, viz., the great curvature anteriorly, and the abrupt 

 reflexion as in the Land Tortoises, and also the great thicken- 

 ing of the rim, and the farrows of the epidermal scutes. 

 The concave or spoon-shaped form of the supra-caudal plate 

 is also exactly alike. 



The rib pieces are less common among the remains dis- 

 covered. They show a total disappearance of sutures. One 

 large piece, comprising the second vertebral plate and several 

 vertebrae of the back, with the keel, indicate that the back 

 was flattened out and had little convexity. Other pieces 

 indicate the same character in the sacral vertebrae. In 

 short, the whole gives the form of a gigantic Tortoise, with a 

 flattened back, compressed sides, vertical in front, with an 

 abrupt marginal reflexion. 



There is a peculiarity in the bodies of the vertebrae being 

 detached, requiring to be artificially held together in the 

 recent skeleton. None are available for desciijption, but the 

 vertebral pieces to which they are attached, the sacral 

 ones, are among the specimens. This is also indicated in 

 the keel. 



As to the relation in size of the plastron to the carapace in 

 the smaller sj)ecies, for safety the j)lastron is elongated so as 

 to close up the openings before and behind ; this goes on 

 increasing nearly in the ratio of the reflexion of the limb. 

 Taking the Testudo Indica as the standard, we find the 

 plastron to the carapace being nearly in the ratio of 3 to 4, 

 according to which the carapace of Golossochelys would be 



112 feet 6 inches long, and the curvature 15-9 inches, the size 

 intended to be given in the figure. 

 We shall now see how the vertebrae agree. The bones of 

 the skeleton of the Testudinata, viewed as a whole, undergo 

 more singular raodifications than we observe in any other of 

 the higher reptilia, and more than is observed throughout 

 the whole order of birds or mammalia. The number of 

 vertebrae, excepting of the tail, is pretty constant in all ; and 



