134 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.58. 



The free border of the carapace is sharp on the tenth and eleventh 

 scute areas, but farther forward it rounds off and thickens. Evi- 

 dently there was no keel connecting the hinder free border with that 

 in front. At the tenth marginal area the shell is 13 mm. thick. In 

 the region described the free border of the shell is not rolled up, but 

 nearly flat ; as is also the underside of the same region. In another 

 fragment the edge along the ninth marginal is slightly turned up- 

 ward. In T. Whitney i the ninth marginal scute on both sides is pro- 

 longed upward considerably between the third and the fourth costal 

 scutes ; in E. bulverda this marginal rises but little between the costal 

 scutes mentioned. 



There is present a left xiphiplastral bone which appears to be- 

 long to this genus (pi. 10, fig. 1). Indeed, it fits against the part of 

 the carapace just described as if it belonged there originally. While, 

 however, the bones of the carapace are solidlj^ coossified, the xiph- 

 iplastral was connected with its fellow and the hypoplastral in 

 front by suture. This bone resembles in many ways the same bone of 

 Terrapene mitipex.^ It is however, rather more pointed behind. In 

 T. antipex the flattened upper surface of the xiphiplastral, that which 

 in life was covered by horn, terminates mesially abruptly ; but in the 

 xiphiplastral here described this surface slopes off gradually into the 

 rest of the bone. In T. antipex there was found to be a sharp ridge, 

 a keel, connecting above the bridge the rear free border with that in 

 front. As shown by the type fragment of carapace and another 

 supposed to belong to the same species, there is no trace of such a 

 keel. There is present also a right humerus somewhat larger than 

 that of a good-sized T. Carolina. It may or may not have belonged 

 to T. bulverda. In 1908 ^ the writer referred to T. marnocMi a fine 

 carapace which had been found on San Diego Creek, probably near 

 San Diego, Duval County, Texas. This identification was pro- 

 visional and somewhat arbitrary. It may yet be shown that the 

 carapace belongs to a distinct species and that the carapace here 

 described as T. bulverda is that of T. maimochii; or T. marnochii 

 may have nc claim to either carapace. Future discoveries alone can 

 remove these doubts. 



Family TESTUDINIDAE. 



GOPHERUS ATASCOSAE? (Hay). 



Testtido atascosae Hay, Foss. Turtles, N. A., p. 467, figs. 627, 628. 

 In the collection from the cave at Bulverde there is a single femur 

 (Cat. No. 9222), which is referred with doubt to this species. 



' Eighth Ann. Rep. Geol. Surv. Fla., 1916, p. 58, pi. 5, fig. 1. 

 ' Fossil Turtles North America, p. 360, pi. 58, figs. 1, 2. 



