FOSSIL TURTLES OF NORTH AMERICA. 



The first neural is hexagonal with the long sides parallel. The 

 second, third, fourth and fifth neurals are coffin-shaped; the 

 sixth neural is pentagonal and pointed behind. The table here- 

 with gives the dimensions of the neurals. 



The nuchal is firmly joined by its whole posterior border to 

 the first costal plates and to the preneural. Its antero-posterior 

 length is 38 mm.; its width from side to side is close to 154 mm. 

 Its whole upper surface, with the exception of the beveled edge, 

 is covered with the ridges and pits of the ornamentation. It is 

 therefore in form, in extent of articulation, and ornamentation 

 extremely different from the nuchal of T. puercensis. 



The eighth pair of costals together form about 125 mm. of the hinder margin; fore and 

 aft in the middle line they measure 38 mm. 



The sculpture (plate 93, figs. 2, 3) consists of a coarse network of ridges which inosculate 



somewhat irregularly and inclose deep pits of varying sizes. 

 The sculpture is coarsest on the distal third of the costal 

 plates, where there are about 5 pits in a line 25 mm. long; 

 and here there is a tendency toward an arrangement of the 

 stronger ridges across the costal plates. Nearer the midline 

 of the shell there will be found about 6 or 7 pits in this distance. 

 Near the free margins of the costal plates there is a tendency 

 for the ridges to break up into tubercles having elongated or 

 circular bases. It might be difficult in some cases to distin- 

 guish fragments of the shell of?", puercensis bearing its coarser 

 sculpture, from fragments of the present species bearing the 

 finer ornamentation. If they are marginal pieces the species 

 may be determined usually from the fact that the margin of 

 P. puercensis is beveled off less abruptly and with a plane or 

 convex surface, while that of P. sagatus has the bevel abrupt 

 and often slightly concave. 



This species resembles in many ways those figured by 

 Professor Cope on plate xxvi of his Extinct Vertebrata 

 obtained in New Mexico (Wheeler Survey West looth Merid- 

 ian, vol. iv). A . sagatus, however, differs from T . leptomitus 

 and T. carwsus in not having the sculptured layer of bone 

 overhanging the supporting layers at the outer ends of the 



FlG. 652. Aspideretes sagatus. 

 Carapace of type. X i. 



costal plates; likewise, in not having the ridges of the sculpture arranged in lines parallel with 

 the length of the shell. 



Aspideretes? nassau sp. nov. 



Text-fig. 653. 



The only known specimen of this species was collected by Dr. Marcus S. Farr, of Princeton 

 University, Princeton, New Jersey, in Fort Union deposits, at Duffy's ranch, 1 8 miles from 

 Melville, Sweet Grass County, Montana. The specimen belongs to the university named. 

 Only the carapace (fig. 653) is represented. 



The nuchal bone, all of both first costals except a small part of the one of the right side, 

 and the first and second neurals are missing. It is assumed that this species, like those from 

 the approximately equivalent Puerco deposits, possest a preneural bone and belonged, there- 

 fore, to the genus Aspideretes. There were certainly 8 pairs of costals and 7 neurals. 



The individual here described was evidently an aged one, since the disk extended out to 

 the ends of the ribs. The species was a small one. The length can not be exactly determined, 

 but was approximately 195 mm. The width, in a straight line, is 200 mm. The rear of the 

 carapace is truncated, more exactly, slightly concave in outline. The upper surface is convex 

 from front to back, more strongly so from side to 'side. The free borders are cut off nearly 

 perpendicularly to the upper surface and are about 9 mm. thick. Toward their proximal ends 

 the costals are reduced in thickness to 5 mm. 



