COLOSSOCHELYS ATLAS. 3C1 



parallel. The humerus of the sea-turtles is so peculiar as to 

 have no analogy with the fossil except in the olecranon-like 

 production of the deltoid crest. 



It has more resemblance to the Emys, or the box-tortoises, 

 than to any of the others. The two crests are approximated 

 in both — as also in Testudo, and the smaller crest in both 

 reaches half way up, and well across the articulating head. 

 With this resemblance, however, it differs sufficiently. The 

 deltoid crest is greatly more developed than in Emys ; it is 

 more pointed upwards, and projects more; and runs down with 

 a convex outline where it falls off in Emys. This convexity 

 and pointed form of crest is not seen in any of the existing 

 Testudinata. The articulating head is very globular, the 

 greatest diameter being 5*1 inches, and across between the 

 roots of the crests, 4-5. Calculated on the proportions 

 of a small Emys, the dimensions would give a buckler of 

 12 feet in length. We possess similar fragments of the right 

 humerus and a femur ; but the crest and articulating head 

 are broken off, and have no character for description. 



Of the bucklers we have a great variety of fragments ; some 

 of them indicating animals of prodigious size, but it is difficult 

 to make anything intelligible out of them, as the ribs and 

 plates of the plastron have become completely synostosed, 

 and afford no trace of suture. But some of them are suffi- 

 ciently distinct to show that the animal to which they 

 belonged differed remarkably from all known Testudinata. 

 One of these, which we consider to be the caudal extremity of 

 the plastron, has a wide and deep re-entering angle between 

 two projecting horns. The only thing approaching this is 

 the slight emargination of the plastron in Emys. 



Some other pieces are very remarkable. They rise out of 

 the inner surface of the shield and project beyond its margin 

 with a rugged and gradually thinning convex edge, like the 

 ribs of Trionyx. The dimensions are : — 



Inches 



Width of the nb-like extremity 8-2 and 8-6 



Greatest thickness ■ . , . 5'5 and 4-2 



Least thickness where hollowed out 0'8 



Supposing them to be ribs, which, however, is by no means 

 certain, what prodigious animals must they not have come 

 from ? But when we are certain of such large tortoises as 

 those indicated by the humerus described, it were easy to 

 imagine others with humeri as large as those of the largest 

 mastodons. The surface of the shield-plates in aU the great 

 fragments is rough and loosely cancellated, like the interior 

 structure of the articulating heads of bones. There is none of 

 the shagreening of Trionyx, nor of the smooth surface of Emys. 



To what description of Testudinata do the humerus and 



