TRIONYCHID^. 497 



it differs in having a nuchal with greater antero-postenor width, the bone not so thick, and 

 with the free anterior border beveled off, instead of being dipt off at a right angle with the 

 upper surface. 



Aspideretes? vagans (Cope). 

 Plate 96, fig. 3. 



Trionyx vagans, Cope, Ann. Report U. S. Geol. and Geog. Surv. Colorado, 1873 (1874), p. 453; Bull. 

 U. S. Geol. and Geog. Suiv. Terrs., 1, No. 2, 1874, p. 29; Vert. Cret. Form. West, 1875, pp. 96, 260, 

 plate vi, fig. 13. Hay, Bibliog. and Cat. Foss. Vert. N. A., 1902, p. 454. Hatcher, Bull. U. S. 

 Geol. Surv. No. 257, 1905, p. 73. 



The type of this species is a fragment of a costal plate, which was collected by Professor 

 Cope in 1873, from what have usually been regarded as Laramie deposits in Colorado. From 

 Whitman Cross (Monogr. U. S. Geol. Surv., xxvn, 1896, p. 244) we learn that the more exact 

 locality is on Bijou Creek, 40 miles east of Denver. Mr. Cross believes that the deposits belong 

 to the Arapahoe beds, of the Post-Laramie. 



The type fragment has a length of 50 mm., a width of 37 mm., a thickness at the sutural 

 borders of 5 mm., and thru the rib, of 6 mm. It was figured by Cope in 1875, as cited. 

 Another figure of it is presented on plate 96, fig. 3. This specimen is now in the American 

 Museum of Natural Historv and has the number 1847. The otner fragment figured by Cope at 

 the same time has been lost. Still other fragments were mentioned in the original description, 

 but were not described and thev have now disappeared. 



It is to be remarkt of Cope's figure of the type specimen that it brings out the ridges into 

 too great contrast with the pits. The latter are really very shallow, and they rise gradually 

 from the bottom to the summits of the surrounding walls. For this reason Hatcher's remarks 

 (op. cit. p. 74) regarding Lambe's Judith River form are not pertinent. 



Various specimens from Montana, Wyoming, and Canada have been referred to this 

 species; but after careful comparison, the writer has come to the conclusion that it is impossible 

 to identify any of them with Cope's type. It seems best, therefore, to describe and name 

 specimens from other localities without regard to it, until better materials have been collected 

 in the tvpe localitv. 



Whether or not the fragmentary materials which Cope has referred to this species from 

 near the mouth of the Big Horn River, in Montana, from the Judith River beds, from North 

 Dakota, and from the Milk River region of British America, belong to it is extremely doubt- 

 ful. Some of them probably belong to the species here described as A. splendidus Hay, A. 

 beecheri Hay, and A. coalescens Cope. 



It is to be noted that in the Vertebrata of the Cretaceous Formations of the W 7 est, Cope 

 states that there are 4 or 5 pits in 19 mm. This is evidently a typographical error; tor in the 

 original description there are said to be 4 or 5 pits in 10 mm., which is correct. 



Aspideretes sagatus sp. nov. 

 Plate 93, figs. 1-3; text-fig. 652. 



This species is represented in the collection of the American Museum of Natural History 

 by a complete carapace. It was collected in the year 1892, in the Puerco deposits of New 

 Mexico. The catalog number is 1 20 1. 



This carapace (plate 93, fig. 1; text-fig. 652) is approximately circular in form, having a 

 length of 375 mm. and an extreme breadth of 312 mm., exclusive of the projection of the rib- 

 ends beyond the margin of the shell. The front margin of the nuchal plate, in the midline, is 

 slightly concave. Behind, the shell is narrowly truncate. Transversely the shell was moder- 

 ately convex, less so longitudinally. All around the margin the carapace is beveled off rather 

 abruptly, and concavely, so as to produce a broad shallow groove. Nowhere does the sculp- 

 tured layer overhang the free portions of the ribs or the deeper layers of bone. The bone has 

 a thickness of about 9 mm. It is somewhat thicker in the middle of the width of the costal 

 plate, but not much. The costal plates also grow slightly thicker toward their free margins. 



There are 2 plates between the proximal ends of the first pair of costal plates, a preneural 

 and a neural. The preneural is a trapezoid, whose broadest side, 50 mm., articulates with the 

 nuchal, while its shortest side, 22 mm., joins the first neural. The median length is 29 mm. 



32 



