436 



Thorpe — Some Tertiary Camivora in 



Tephrocyon marshi, sp. nov. 



(Fig. 6.) 



Holotype, Cat. No. 12787, Y. P. M. Upper Miocene (Valentine 

 Cherry Co., Nebraska, along the Niobrara River, not far west of the mouth 

 of Minnechaduza Creek. Collected by Professor Marsh in 1873. 



This species is represented by a nearly complete left 

 ramus with P 2 , P 3 , M 1? and M 2 . It is nearly one half 

 larger lineally than T. hippophagus and in its analogous 

 parts seems to correspond most closely to the specimen 

 recorded by Matthew and Cook (1909, p. 376) from the 

 Snake Creek Pliocene. It differs, however, from all the 

 other species of the genus in that it shows the beginning 

 of a shortening of the ramus with a concomitant crowd- 

 ing, but not reduction, of the premolars, which overlap in 

 each case. P 2 is placed obliquely with the anterior part 

 outward, while P x was probably small, with a single root. 

 M 3 was set in the ascending ramus of the jaw, and in this 

 specimen is absent, with the alveolus partially closed, 

 resembling a specimen of Mlurodon haydeni in the Amer- 

 ican Museum (No. 9744, Upper Miocene, Montana). 



f278 7 TYPE 



Fig. 6. — Tephrocyon marshi, sp. nov. Holotype. X 2 / 3 - External 

 view of left ramus. 



From T. mortifer this new species differs in being about 

 one quarter smaller lineally, less robust, and in the crowd- 

 ing of the premolars. 



If, as we are disposed to believe, Tephrocyon is approx- 

 imately ancestral to Canis and 2Elurodon, then this new 

 species seems to be in the line from which 2Elurodon 

 developed, for here we have the robust sectorial showing 



