286 Bulletin American Museum of Natural History. [Vol. XVI, 



small posterior cingular cusp. The second temporary pre- 

 molar is smaller, more compressed, two-rooted, with rudi- 

 mentary posterior cusp. The third is the carnassial, and is 

 three-rooted, composed of large protocone and strong posterior 

 blade, with a minute anterointernal basal cusp situate be- 

 tween the anterior and internal roots. The fourth milk pre- 

 molar is molariform, with no protoconule, strong hypocone (in 

 reality probably a metaconule), besides the three main cusps. 

 The second lower milk premolar is two-rooted, set obliquely 

 in the jaw, and has a small posterior cusp. The fourth (car- 

 nassial) has the trigonid of shearing protoconid and paraconid 

 blades, small metaconid, and three-cusped basin heel. Com- 

 pared with the corresponding teeth in the Coyote these teeth 

 differ in larger size, greater robustness, more sectorial char- 

 acter in the carnassials, proportionately smaller and narrower 

 heel on dp^, dp* of less transverse and greater longitudinal 

 width, the inner cusps less marginal, reducing the size of the 

 basin enclosed by them. The anterointernal cusp on dp^ is 

 smaller and situated much more anteriorly; the anteroexter- 

 nal cingular cusp is ha!rdly noticeable. The jaw is very much 

 shorter and deeper, ♦the premaxilla is carried much farther 

 back between maxilla and nasal. The cusp composition is 

 the same in both. 



No Canid has been described with which this can well be 

 identified. It is of the size of Mlurodon, but differs in ab-. 

 sence of anteroextexnal cusp on p* (and other characters). It 

 is much more modernized than any of the John Day dogs, and 

 the size is too great for C. hrachypus Cope, temerarius or vafer 

 Leidy, anceps Scott, all of which, moreover, seem more 

 typically cynoid. 



In a previous article Dr. Wortman and the writer attempted 

 to trace a line of descent from the Eocene Uintacyon and 

 Prodaphcenus through the Oligocene Daphcsnus and Tem- 

 nocyon to the modern Cyon. It seems not unlikely that the 

 Canid here described may nearly represent the Upper Miocene 

 stage of evolution of this race. Icticyon seems also to be more 

 or less nearly connected with it — and if this hypothesis of 

 descent be correct, this group parallels the Camels in their 



