272 Bulletin American Museum of Natural History. ["\"ol. XXVII, 



absent. In the latter genus the metastyle is enlarged, oblique; the hypo- 

 cone oblicjue, Ijroadly ledge-like and continuous with an internal cingulum 

 (absent in Tupaia); reduced metaconules are present and the molars as a 

 whole are of the general type seen in the Middle Eocene (Bridger) Nycti- 

 therimn nitidum {cf. Matthew, 1909, pi. 1, fig. 6). Resemblances to certain 

 of the Bridger Insectivores is especially clear also in the lower dentition, 

 which com|)ares well with that of Entomolestes grangerl {cf. Matthew, 1909, 

 pi. 1, fig. 2) in many significant details. 



In brief the dentition of Ptilocerciis has departed from the primitive 

 Insectivore type represented in Ictops in so far as it reveals a slight emphasis 

 of the tendencies toward diprotodonty, toward reduction of the premolars and 

 toward the omnivorous modification of the molars; but on the whole it is 

 a rather primitive Placental dentition. The dentition of Tupaia has avoided 

 these tendencies and has retained more completely the triangular form of the 

 upper molars; but on the other hand it has em})hasized the insectivorous 

 features, such as enlarged styles, sharply pointed cusps, etc., and has 

 acquired a large and more or less divided mesostyle. 



Skull (Figs. 21, 22). In many characters the skull of the Tupaiidw differs 

 widely from that of any Lipotyphlous Insectivore and approaches the lemuroid 

 type. The braincase is large and rounded, the skull-to[) broadly convex 

 between the orbits, the large orbits enclosed posteriorly by processes from 

 the frontal and malar; the skull as a whole is mesati- to brachycephalic, 

 contrasting strongly \\ith the Zalambdodont type, and further lemuroid 

 resemblances are shown in the tympanic region (see below), in the partial 

 bending down of the face iq)on the basicranial axis, in the presence of a 

 malar foramen etc. 



So far as concerns only general characters and proportions, there seems 

 little to refute the suggestion that the ancestral Tupaiid had a skull resem- 

 blincj that of Ictops in respect to the following characters: muzzle moder- 

 ately elongate, postorbital constriction gentle, breadth between orbits 

 considerable, brain case rounded, temporal crests well separated, anterior 

 part of zygoma stout, posterior j)art slender; small postglenoid well sepa- 

 rated from post-tympanic process, ectopterygoid fossae absent or incipient. 



As compared with this hypothetical norm or ancestral type Ptilocerciis 

 seems more primitive than Tupaia in the following characters: smaller 

 braincase and orbits, narrower interorbital region, somewhat stouter zygo- 

 mata, less reduced postglenoid process, greater space between the post- 

 glenoid and posttympanic, no secondary septa in the bulla, smaller bulla, 

 partly exposed petrosal, better defined tympanic wing of alisphenoid, optic 

 foramen well separated from foramen lacerum anterius; no supraorbital 

 foramen. In the skull figured by INIivart (18G7, p. 303) the postorbital 



