86 BULLETIN 16 9, UNITED STATES NATIONAL MUSEUM 



It is this character also that is reflected in the greater width of this 

 tooth in mediaevus. Median cusps of P^ not significantly different, 

 but internal cusps mode 9 in montanus, and 8 in one specimen of 

 mediaevus. 



The two species are certainly very closely related, but cannot be 

 considered synonymous. 



No other described species could be confused with P. montanus. 



Thanks to the fine specunen found by Silberling and prepared and 

 described by Gidley, Ptilodus montanus is the best-known multituber- 

 culate and typifies this order, the longest lived and among the most 

 widespread of all mammalian orders, despite its extinction in the 

 Lower Eocene. Gidley (1909) published an excellent, but explicitly 

 provisional and preliminary, description of the best specimen, and it 

 was later redescribed summarily, with new reconstructed sketches, by 

 Broom (1914). It has become a classic specimen and is mentioned in 

 practically all and figured in many of the general works on fossil 

 mammals (e. g., Schlosser, 1923; Osborn, 1910; Romer, 1933; etc.; 

 Scott, 1913, adds a life restoration, and Abel, 1912, a modified but 

 incorrect reconstruction based on Gidley 's figures). 



The species is here briefly redescribed, as typical of family and order 

 (or at least suborder). By taking into consideration numerous other 

 specimens of this species, and with the help of more recently described 

 specunens of other species, it is possible to add a few points to those 

 previously described and also to remove the discrepancies involved in 

 the previous descriptions. 



Dentition. — The dental formula is rMl- Gidley gives yjM- He 

 considered the second upper tooth as a canine, but it is surely an 

 incisor. His inclusion of a lower canine is doubtless a lapsus calami, 

 as there is no suggestion of such a tooth, and he does not mention it 

 in his description. His premolar-molar division seems to me to be 

 the most suitable one, although the real criterion, replacement, is not 

 available. The ancestors of Ptilodus probably had five upper pre- 

 molars, but it is uncertain which one was lost, and hence it is conven- 

 ient to call those of Ptilodus simply P^''*. Its lower premolars, how- 

 ever, are certainly P3.4 of the ancestral series and are so designated. 



I^ is a large, high-crowned, but apparently rooted tooth with a 

 completely enameled crown. It is directed downward, forward, and 

 inward, so that the tip must have been nearly in contact with that 

 of its mate on the other side, although the alveoli were well spaced. 

 The anterior face is convex and the posterior concave vertically and 

 slightly convex transversely except for excavations at the sides. 

 There are sharp vertical external (proximal) and internal (distal) 

 crests, and near the tip is a more rounded posterior (lingual) crest, 

 so that the tip is triangular in section. There are no accessory cusps. 



