OSTEOLOGY OF AMERICAN PERMIAN VERTEBRATES 387 



projection. Broom compares the pelvis of Ophiodeirus, or his 

 so-called Bolosaurus, with that of Poecilospondylus , but restores 

 the latter wrongly. All the Poliosauridae, as indeed all of the 

 Pelycosauria in the narrower sense, have a very characteristic 

 pubis, standing horizontally far in advance of the true pelvic brim. 

 I have seen the type specimen of Poecilospondylus and know that 

 it is intimately allied to the genus I have described as Varano- 

 saurus^ but which is not congeneric with the genotype. Indeed, 

 there is no more characteristic bone in the true Pelycosauria than 

 the pubis, an element that I have relied upon, perhaps unduly, as 

 a determinant of the relationships between the Sphenacodontidae 

 and Poliosauridae. 



Numerous juvenile or incomplete /gwo^a (Fig. 5, A-D) are 

 represented among the free material found in the clay, but there is 

 only one complete femur of an adult, with parts of three others. 

 It is the longest bone in the skeleton, measuring about 2^ inches 

 in length. In front view the bone is nearly straight, with a long, 

 cylindrical shaft and only moderately expanded extremities. In 

 side view the bone is shaped much like the italic letter /, with the 

 concavity above, the convexity below, on the front side. The 

 head is stout and rounded, with its articular surface almost wholly 

 at the extremity. The trochanter, a rather thin plate, springs 

 inward and backward nearly at right angles, a little below the head, 

 and continues as a thin edge, winding about the shaft till it reaches 

 the middle, a little above the middle of the bone, and then con- 

 tinues as a linea aspera to the lower third, where it becomes obso- 

 lete. The tibial condyle juts inward, forming a right angle. The 

 tibial surface is narrow, but connects broadly with the fibular 

 surface. The two surfaces together make an angle of nearly 45 

 degrees with the long axis of the bone. 



Tihia and fibula. — The tibia and fibula (Fig. 5, E-I), Hke the 

 radius and ulna, are remarkably slender bones, fully as long as the 

 femur. No free specimen of either bone is preserved entire, 

 though one of each, in all probability those belonging with the femur 



' The recent descriptions by Broili and Watson of the genotype of Varanosaiirus, 

 V. acutirostris, disclose, as I suspected, generic differences from the species V. brevi- 

 rostris Williston. I therefore propose the generic name Varanoops for the latter. 



