270 0. P. Hay — Fossil Turtles in the Marsh Collection. 



broad hexagonal suprapygal, the homologue of which is partly 

 shown in Leidy's figure of the type, and is fully shown by a 

 specimen in the American Museum of Natural History. In 

 all three examples mentioned, this suprapygal is crossed, near 

 its anterior end by a sulcus bounding the supracaudal scute in 

 front. In Leidy's type, in front of this suprapygal is another 

 which widens backward. A similar bone is found in the 

 American Museum specimen and in the one here described ; 

 but, in both of the latter, the bone is somewhat longer, and is 

 crossed by a transverse sulcus. In Leidy's type, the sulcus 

 crosses on the neural immediately in front, — the eighth. 



The last pair of costals in the Yale specimen has all the 

 characteristics of the eighth pair in Leidy's type* and in the 

 American Museum specimen. They come in contact with the 

 two suprapygals, are crossed at their anterior ends by a trans- 

 verse sulcus, and are occupied in their length by the lateral 

 sulci of the hinder most vertebral scute. In the specimen here 

 described these costals do not indeed come in contact with the 

 hinder most neural ; but it is no unusual thing for the last pair 

 of costals to be pushed out of contact with this neural, or the 

 real eighth neural may be suppressed. 



It is concluded, therefore, that the intercalated costal plates 

 of the present specimen are the eighth and the ninth from the 

 front of the carapace. 



Since there is only one extra neural present, it appears to be 

 impossible to determine whether this is the eighth or the ninth 

 from the front. It may be, however, as already suggested, 

 that the true eighth has been suppressed, and that the eighth 

 and the ninth, counting from the front, are both intercalated. 



It is likewise uncertain which is the intercalated peripheral, 

 the tenth or the eleventh of this specimen ; but it is more likely 

 the tenth, since it is in contact with both the intercalated 

 costals. 



As to the vertebral scutes, the first, second, and sixth seem 

 clearly to correspond with the first, second, and fifth, respec- 

 tively, of a normal emyd. ' It appears to be quite as certain 

 that the anterior portion of the fourth corresponds with the 

 same portion of the fourth of any other emyd ; and again 

 that the posterior portion of the fifth is homologous with 

 the posterior portion of the fourth vertebral of ordinary 

 turtles. Does not this evidence lead to the conclusion that 

 no new scute has been intercalated, but, rather, that the 

 area occupied by the fourtli scute in a normal turtle, having 

 been greatly enlarged, has become divided by a transverse 

 sulcus ? In the same way the extra costal scute on each side, 

 as well as the extra marginal, may be explained. 



