36 



University of California Publications. [Geology 



the median portion of the ulna is drawn in from both sides, 

 showing a true constriction. In Delphinosaurus this constric- 

 tion appears somewhat stronger than in this specimen, but the 

 ulna is much shorter and broader, and the first two segments of 

 the limb are, in general, more specialized. 



On both the proximal and distal ends of the ulna the thick, 

 cartilage covered portion of the margin extends far back toward 

 the middle of the posterior side. The thickened margins come 

 so near to each other that the posterior notch is made quite nar- 

 row. At the distal end this part reaches around quite to the 

 middle of the shaft, reminding one of the surface against which 

 the pisiform rests in Mixosaurus. At the proximal end the sur- 

 face makes a sharp turn, forming a face which looks as if it 

 might have supported a small supernumerary bone. 



With another specimen of the same type as that described 

 above there is an element appearing to represent either a fibula 

 or an ulna showing a very primitive form. The shaft is nar- 

 rower and the articular surfaces of the two ends are not swung 

 back as far is in the specimen figured here. If this is really an 

 ulna, it represents a stage more primitive than that described 

 above. 



Unfortunately, we have no very definite evidence concerning 

 the character of the phalanges in this species. With both this 

 specimen and the one from which the simpler form of ulna was 

 obtained, there are several rounded ossicles resembling carpals. 

 It may be that they are really reduced phalanges which rested 

 in cartilaginous pads as in Baptanodon. If this be true, the gen- 

 eral primitiveness of the limb would be reduced by some degrees ; 

 even in that case, however, the fact that the propodial and epi- 

 podial segments of the limb had failed to adapt themselves per- 

 fectly, even when the manus had become specialized, would tend 

 only to show more distinctly the influence of an ancestry in which 

 locomotion by crawling or running was more common than by 

 swimming. 



As far as we know it, the type of limb found in this specimen 

 is more primitive than any heretofore described. The humerus 

 is hardly more slender than that of certain other Triassic forms, 

 but is among the most slender known, and is not more abbrevi- 



