1902.] Matthew, A Horned J^odent fro fn Colorado Miocene. 295 



madillo resemblances may be explained as all associated with 

 the development of digging claws, chiefly on the fore foot, and 

 walking on one side 

 of the foot in conse- 

 quence. The Casto- 

 romorph characters, 

 on the other hand, 

 appear to be such as 

 would indicate real 

 relationship, al- 



though not close re- 

 lationship to any 

 living form. From 

 the phalanx just de- 

 scribed we have rea- 

 son to infer that 

 Mylagaulus devel- 

 oped large digging 

 claws on the fore 

 foot, and this speci- 

 men is of appropri- 

 ate size and propor- 

 tions to belong to 

 the Mylagaulus. It 

 is found in the same 

 horizon, and could 

 not belong to any 

 other rodent known 

 from those beds, for 

 all the others are 

 quite nearly allied 

 to still existing gen- 

 era. Edentates have 

 not been found in 

 the Loup Fork, I and 

 our specimen 

 no 



Fig. 2. Mylagaulid, indet. Hind limb bones and claw, 

 natural size : /•", anterior view of femur ; F^'. T. posterior view 

 shows of tibia and fibula ; superior view of pes ; .-J , .4, ' superior and 

 lateral views of ungual phalange of fore foot. No. 8336. Loup 

 resemblance to Fork, (Republican R. Beds) Kansas. 



The supposed exception, Caryoderma snovianum, is, according to Williston, a turtle. 



