382 S. W. WILLISTON 



are shown in Fig. 4, /. They are quite Hke the usual distal caudals 

 of other reptiles. Although the entire number of caudal vertebrae 

 cannot be determined, even approximately, the structure of the 

 numerous isolated vertebrae preserved indicates decisively a long 

 and slender tail. A median caudal of a young animal, shown in 

 Fig. 4, /, is of interest because it is the only vertebra in the collec- 

 tion which shows any indication of the suture between arch and 

 centrum. 



PECTORAL GIRDLE AND EXTREMITY 



Of the pectoral girdle nothing has been detected of the inter- 

 clavicle in any specimen, and of the clavicle only one specimen has 

 been discovered, that of a young animal not more than half adult 

 size, as indicated by the associated long bones in the matrix. This 

 bone is shown in Fig. 4, Z, as it lies on the matrix in one plane. 

 Its concave border is a little thickened, the convex thin; its scapular 

 extremity is a little thickened, as shown by the section at the broken 

 end. The bone has an unusual shape for a Permian reptile, as 

 might be expected from the unusual shape of the scapula. 



The coraco-scapula (Fig. 3, C, Z?) is of remarkable structure, so 

 unlike anything I am acquainted with among reptiles that I was 

 puzzled at first to interpret it. One complete specimen was found 

 in an isolated nodule, overlapped on its coracoid border by a frag- 

 ment of the opposite side. In order that the complete structure 

 may be seen, this fragment has been sacrificed. The specimen 

 seems to be quite perfect and undistorted, save that the upper 

 part was smoothly broken and turned over the lower, as shown in 

 Fig. 3, ^, from behind. In Fig. 3, C, I have shown the whole 

 bone as in one plane, though doubtless the upper part was curved 

 strongly inward in life, perhaps more than I have indicated in the 

 dotted line. The postglenoid facet is smooth, rounded, and gently 

 concave; its face looks directly outward from the plane of the 

 coracoid. The preglenoid facet of other Permian reptiles is divided 

 into two very distinct processes, separated by a considerable space, 

 in which there is a smooth, rounded border. There can be no 

 possibility of error in this, since two other fragments, one of which 

 is illustrated in Fig. 3, E, show the same peculiarities. Nor can 



