28 
MEUOIJRS OF TUB QTJFEN ^LAFV :^LUSETJM. 
type, as illustrated by Atlaniochelys morionii, Agassiz, although the absence 
of the radial process prohibits adequate comparison. AYieland also notes 
resemblance between Atlantoclielijs and Desmatochelijs loivii, Williston.^ 
Radius and Ulna, — About 32 centimetres of the proximal ends only are 
preserved, and the eetal surfaces of these are incomplete. The proportions of 
these bones are considerably in excess of the relations borne to C. niydas by the 
other fragments; from this one may infer very powerful fore-limbs and flippers. 
The robustness of the front flippers in Toxoclielys and Archelon has been 
demonstrated by Wieland,^ who, after comparing the Cretaceous remains with 
those of present-day carnivorous and algapliagous turtles, says; '' There is in 
the comparison just made the very strongest suggestion that Protostegaj more 
distinctly than any marine turtle thus far known, hunted prey, which swam 
actively, and, bearing in mind other features, was perhaps even powerful.^’ 
AVhen compared with corresponding portions of a large skeleton of Chelone 
mydas, representing an individual about 5 feet 9 inches, our fossil bones, except 
where noted, are a little more than twice the size. Following out these pro- 
portions, a turtle nearly 12 feet in length is suggested. Some specimens of the 
genus Archelon, described by Wieland, were a])out 13 feet in length. 
Plasfral Plate (PI. XII and XIII). — The evolution of the dermal armours 
of Ohelonians is one of the most perplexing yet fascinating problems to be 
elucidated by palaeontologists, and the researches of such writers as Wieland 
and Hay in America are here of great value. The plastral plate is therefore the 
most interesting bone in the collection, and it is a matter for deep regret that the 
portion preserved is so incomplete. Of this fragment the greatest Icngtli is 51 
centimetres and the greatest breadth 22 centimetres. The thickness is exceedingly 
variable, the maximum — 33 millimetres — being in the centre on the left of the 
longer axis. From below this and to the right the plate shelves away and 
terminates in dactyloid processes. The broken ends of two dactyloid processes 
arc exposed on the lower side of the plate to the right. (See PI. XIT.) It has 
been possible only partially to remove the matrix at this part, but the course of the 
processes is now apparent. The loAvest digit is broken across 60 mm. from its 
base, but the upper, running apparently parallel to it at a distance of but 10 mm., 
extends for 130 mm. to its exjmsed fracture. During fossilisation this digit has 
evidently been broke]i, and 1he distal portion is partly superimposed not far 
from the base. The course of these processes is almost parallel to the long axis 
of the plate. On the ental side immediately above these broken digits is a furrow 
almost deep enough to appi'ar through the plate, and this points to another 
dactyloid process more robust than the ones seen in situ. Within the margin of 
the anterior edge of the plate on the right are two other furrows, marking the 
bases of still stronger dactyloid ])rocesses, and the adjoining anterior border is 
chamfered away to a thin edge. (PI. XIII.) 
*Willjston, ITniv. Kansas Geol, Sur., vol. iv., Pal., 1898. 
® Wieland, Mem. Carnegie Museum, vol. ii., Xo. 7, 1906, p, 289. 
