GREGORY: NOTHARCTUS, AN AMERICAN EOCENE PRIMATE 



201 



than that of pi; it is separated from the small pi by a slight diastema. Pi, po, ps increase rapidly in size; 

 there is a sudden increase in size and complexity as we pass from ps to p4, and an equally wide structural 

 gap between p4 and mi; m2 is markedly wider than m^. All these characters joined to the stoutness of 

 the jaw impart a primitive facies, suggestive of the conditions in Pelycodus frugivorus and many other 

 Eocene and Paleocene mammals, which according to Dr. Stehlin's hypothesis must all be specialized in 

 these characters. In Adapis rutimeyeri also (cf. Stehlin, 1916, Taf. xxi, fig. 24), which has a very 

 primitive type of molars, the lower canines (as shown by the alveolus, cf. Taf. xxi, fig. 24) were very 

 much larger than pi. In Pronycticebus, which has an extremely primitive type of premolars and molars, 

 the upper canines were decidedly larger than p\ (Fig- 81.) In this primitive genus with comparatively 

 well developed canines the zygomatic arches are stout, while in its more specialized Tarsius-Mke relatives 

 with very small canines the zygomata and jaw are weak. 



Dr. Stehlin is so sure of his conclusion, cited above, that he says (p. 1531) : 



Es ist mir keine Thatsache bekannt, welche niit ditser meiner AufTassiing iiii WidiTstreit stiindc; wolil aher lassen 

 sicli solche namhaft machen, welche sie noch besser zii stiitzen ^'e^mogen. 



Vorerst ist von Belang, dass die von mir als urspriinglich betrachtete Einrielitung dcs \'()r(kTgclMssi's nicht bios ein 

 theoretisches Postulat, sondern schon thatsachlich nachgewiesen ist. 



He then cites the condition of the front teeth in Parapithecus Schlosser from the Upper Eocene or Lower 

 Oligocene, a catarrhine primate with primitive molars and canines intermediate between the incisors 

 and the anterior premolars. But it must be objected that Parapithecus probably belongs to a very differ- 

 ent section of the primates from that of any of the known lemuriform Eocene primates ; it may very well 

 be true, as the writer has elsewhere maintained,'^ that in the catarrhine series the typical caniniform 

 and tusk-like canines have been derived from short canines ; but the marked reduction of the ante-molar 

 formula in Parapithecus indicates that even although this genus may be primitive as compared with the 

 Simiidse it is very far from primitive as compared with the Lower Eocene lemuroids of the subfamily 

 Notharctinse which retain the primitive formula of l|, Cj, P|, Mf and in which the dentition as a whole 

 approaches that of the most primitive Paleocene representatives of the Insectivora, Carnivora, Creo- 

 donta, Condylarthra, Taligrada and other primitive orders. 



A differentiation of the canines both from the incisors and the premolars had occurred even in the 

 Permian Therocephalia and C'ynodontia and in the Mesozoic mammalian orders Protodonta, Tricono- 

 donta, Trituberculata and Polyprotodontia. In these very early mammals the lower canines seem to 

 be associated functionally with the "incisor series and yet they are fully caniniform and much larger 

 than either the incisors or the first premolars. The upper canine being the first tooth l)ehind the pre- 

 maxillary suture and nearer the fulcrum would probably be subjected to different mechanical conditions 

 from those affecting the incisors. In referring to the incisors and caniniform canines of early Eocene 

 mammals Dr. Stehlin assumes (1916, p. 1529) that it is much more probable that all these ancient forms 

 were not primitive but early specialized. From the evidence supplied by extensive collections of 

 Paleocene and Eocene mammals in this Museum, however, one might equally well infer that in the 

 remote ancestors of many placental orders the lower canines were subcaniniform and much larger than 

 either the incisors or pi; such teeth are borne in a fairly stout mandible, not elongate distally, with a 

 curved lower border, a stout backwardly prolonged angular process; the zygomata are stout, the brain- 

 case small, with well-marked muscle areas. 



' See the discussion of this genus in Part II of the i^resent series, 1916, Bull. Ainer. Mus. Nat. Hist., XXXV, pp. 280-28-1. 

 2 See the figures given in Osborii, H. F., 1907, Evolution of Manunalian Molar Teeth, pp. 19-30. New Y(M-k. 



