442 W. K. GREGORY NOTHARCTUS AND LEMUROIDEA 



cranial region of this specimen lias served to emphasize its resemblance 

 to Tarsius, especially in the formation of the bulla and in the pattern 

 of the cheek teeth. It is, however, far more primitive than Tarsius in 

 haviiiir a much smaUer cochlea, or auditory prominence, a less expanded 

 bj'aiii-easc, and smaller orbits. Its Middle Eocene relative Omomrjs is 

 extremely like Tarsius in the cheek teeth, as noted by Wortman, and, in 

 the o])inion of J)()ctoi' Matthew and the writer, it is so difficult to sep- 

 ai-ate the Anaptomorphidte and the Tarsiidse as distinct families that we 

 prefer to unite them in a single family — Tarsiida?. 



IMie LoM'er Eocene members of the Tarsiiformes were thus rather 

 widely different from the contemporary Lemuriformes of the family 

 \otharctina\ Erom the characters of the jaw and dentition in the 

 earliest Notharctina? it seems right to infer that the general architecture 

 of the skull was not dissimilar to that of NotJimrtus. This differs widely 

 from " Anaptomorplius" in having a relatively narrow, unexpanded brain- 

 cast', a far larger face, and smaller orbits. These features may have been 

 h'ss ])ronounced in the earlier Xotharctina?, while the opposite Tarsiiform 

 characters may have been less pronounced in some of the other Lower 

 Eocene genera of the Anaptomorplms group; but still the contrast l)e- 

 tween the representatives of the Lemuriformes and Tarsiiformes must 

 have been sufficiently great in the Lower Eocene to Avarrant us in looking 

 for the common stem form of the Lemuroidea in the Paleocene or even 

 earlier. 



Mkruclmr'uUr. — A study of several excellently preserved skulls of 

 Nerrolemur shows that the basicranium closely resembles that of Tarsius 

 ill many respects, especially the mode of formation of the bulla?, character 

 (>r the glenoid, position of the cranial and carotid foramina; so that, in 

 view of further evidence offered by the facial region and dentition, refer- 

 ence of Necrolemur to the Tarsiiformes seems well warranted. On the 

 other hand, Necnilninir has more complex sexitubercular upper molars 

 than any of the Anapt(jinin-/)]i us-Tdrsiiis group, and its nearest affinities 

 ai-e undoubtedly with M i( roi Im-nts. as suggested by Eorster Cooper: but 

 iieitlief of these genera have anything to do with the Hyopsodontidai, 

 with which earlier authors placed Microchoerus. 



The relationships of the Tarsiiform series oji the one hand to the 

 Mici'osyopsida^ and on the other to tlie Anthropoidea seem at present 

 hio-hlv doiihtful. 



'fe 



