196 



GREGORY: NOTHARCTUS, AN AMERICAN EOCENE PRIMATE 



of the dentition in many phyla. Between the extreme forms, Cryptoprocta and Bdeogale, the differences 

 in the dentition are of so pronounced a character that at first sight the association of these genera in the 

 same family seems unnatural; for in Cryptoprocta the dentition is cat-like with blade-like carnassial 

 teeth and much reduced single upper molars, while in Bdeogale the carnassials are submolariform and the 

 molars are of the low-cusped omnivorous type. (Figs. 72, 1,2.) A markedly different style of dentition 

 is exhibited by the Ichneumon (Herpestes) in which the narrow transversely widened molars seem to be 

 adapted perhaps for snake-eating. But amid all this diversity in the dentition the whole basicranial region 

 remains singularly constant in essentials throughout the family. 



Among the Mustelidse it has also proved possible to work out the divergent phyla from a study of 

 the dentition and skulls of recent and extinct types; and, here again, the extreme forms, Latax and 

 Putorius, differ so widely in the dentition that if they were not connected by many intermediate genera 

 it is unlikely that modern systematists would allow them to remain in the same family. But the basi- 

 cranial region throughout the family is essentially identical, although differing in minor characters such 

 as the size and degree of inflation of the bulla, the development of accessory sinuses in the surrounding 

 parts, the partial deflation and flattening of the bulla. 



Perhaps even more striking is the difference in the dentition between jEIutus and Bassariscus, which 

 are referred to the same family and reveal their relatively close relationship in the construction of the 

 basicranial region. j.'Eluropus, which according to Lankester ' and Lydekker is also a member of the 

 same family, as shown by many important characters, differs from the procyonid type in the region of 

 the bulla only in the flattening of the bulla, a difference which can be matched between two genera of 

 the Mustelidse {Latax, Zorilla) or of the Lemuridse {Megaladapis, Microcehus). 



Among the primates themselves it will be recalled that the whole platyrrhine series have one char- 

 acteristic basicranial pattern, while the Catarrhini, including the Old World monkeys, baboons, apes and 

 man, have another which is divided into two well-marked sub-types: first, that which is characteristic 

 of the Cercopithecidse and the Hylobatidse and, secondly, that which is characteristic of the Great-ape- 

 Man group. Among the primates the evidence offered by these characteristic basicranial patterns, as 

 seen from the ventral surface, is fully substantiated by the evidence of the encephalic surface of the 

 whole under side of the skull as will be shown later. 



A third reason for associating Adapts and Nothardus in the same family is as follows: the detailed 

 studies of the two groups which have already been made, besides revealing the divergent tendencies in 

 them and the numerous fundamental characters which they still retain in common, have also made it 

 possible, by projecting backward the divergent trends to an assumed common origin, to reconstruct 

 with reasonable probability the food habits and general structure of the common ancestral stock,^ as 

 described below (p. 229). That such a common ancestral stock existed is plainly indicated by the evidence 

 at hand. When and if discovered in the Paleocene of Europe, Asia, or North America, it will surely 

 establish the already patent fact that there is at least no greater degree of structural divergence between 

 the extremes representing Adapis parisiensis and Nothardus crassus than there is between Megaladapis 

 insignis and Microcehus which are commonly placed along with many genera of intermediate character 

 in a single family, Lemuridse. 



After all this has been said in favor of uniting Adapis and Nothardus in the same family Adapidse, 



1 1901, Trans, Linn. Soc. Zool., (2) VIII, p. 171. 



^ Such reasoning will of course not appeal to the advocates of "I'ecole des faits," who disapprove of reconstructions and prefer to 

 await the discovery of complete material. The discerning student ought to be able to distinguish plaster from bone and inference from 

 observation, but to make perfectly sure the present writer has always tried to label consistently the two kinds of material. 



