174 



Bulletin American Museum of Natural History. [Vol. XXVII, 



Molar characters seen in Spalaco- 

 therium and the remaining Triconodonta. 



Metaconid postero-internal springing 

 obliquely inward and backward from 

 internal base of the protocone. 



Paraconid and metaconid sub-equal; 

 tips on same level. 



A strong internal cinguUnn rising 

 into a centro-internal prominence. 



No talonid on the cheek teeth. 



Characters believed to be universal in 

 the Trituberculata. 



Metaconid centro-internal directly in- 

 ternal to protoconid and usually con- 

 necteil with it by a prominent transverse 

 ridge. 



Paraconid smaller than metaconid; 

 tip on level lower than that of metaconid. 



Internal cingulum absent (or possibly 

 grown up into paraconid and metaconid). 



Talonid on molars seemingly always 

 present, at least on inner side. 



Accordingly there seems to be considerable evidence in favor of Osborn's 

 view (1888, p. 230) that SpaJacoiherium belongs with the Triconodonta 

 and not with the Trituberculata. 



Peralestes prohahhj not a member of the Triconodonta. 



The next important question is: What are the affinities of the upper teeth 

 called Peralestes'^ Were they "probably associated with lower molars of 



the type seen in S palacotheriimi" (Osborn, 1907, 



p. 25, 



fig. 11), or was 



Peralestes "obviously related to Peraspalax" (Osborn, 1888, p. 206) ? 





efos 



Fig. 9. Cheek teeth of Peralestes longirostris. 



A. Upper teeth after Osborn; lower teeth hypot helically reconstructed, showing talonid 

 for the reception of the high protocone and anterior fossa for the reception of the metacone 

 (compare Didelphis). The tip of the protoconid is hidden. External view, much enlarged. 



B. Internal view of a lower molar hypothetically reconstructed, showing high protoconid, 

 asymmetrical arrangement of trigonid, low talonid. The protoconid as represented may be too 

 high. In A the talonid may be too large. 



A careful study of the descriptions and figures of the Peralestes teeth 

 seems to enable one to interpret and harmonize the apparent differences 

 in the drawings given by Owen (1871, pi. i) and Osborn (1888, pi. viii, 

 fig. 8; 1907, p. 26, fig. 12 i). 



1 The diagram of the Peralestes molar on page 43 of the work cited does not agree with 

 Osborn's detailed drawings and descriptions and represents the Peralestes tooth as almost typi- 

 cally tritubercular. 



