BIG BADLANDS OF SOUTH DAKOTA. 487 



on the under side, and its greater thinness at the point where the 

 temporal process is given off (i8 mm. in comparison with Troxell's 

 figure of 30 mm.). There is the same peculiar wing-hke shape and 



Fig. 18. Megachcerus sygomaticus Troxell. No. 11 156. Side view of skull 

 and jaws, about thirteen one-hundfedths natural size. The orbits are de- 

 formed by vertical crushing. Drawn from photograph and the specimen. 



backward sweep to these great structures as in the type of the 

 genus, correspondence in their width, and in the shape and relation 

 of the posterior jugal process to the glenoid cavity, into the forma- 

 tion of which it enters. In the shape of the anterior process of the 

 temporal and the shape and position of the infraorbital foramen 

 there is also close agreement. In the Princeton specimen the fron- 

 tals have a shallow posterior depression lodging two small foramina 

 which the engraver has failed to show (Fig. 19). 



The characters of the upper dentition agree almost completely. 

 In our specimen pi_ was small and directed forward, but whether 

 single- or double-rooted can not be ascertained from the empty 

 alveolus. An extensive injury, sustained during life, obliterated all 

 trace of teeth and their alveoli on the right side from canine to p2, 

 inclusive. On the left side p2 is missing, but was preceded by a 

 diastema of 52 mm. and followed by a space of 11 mm. or more. 

 P3 is pyramidal, but not noticeably " angular," rounded on the outer 

 side, wider posteriorly than in front and broadly concave at the base 

 internally. The ridging and pitting of the crown mentioned by 



