70 



GREGORY: XOTHARrTUS, AN AMERICAN EOCENE PRIMATE 



(2) The shaft as a whole is straighter, less S-shaped. 



(3) The capitelhini is less ball-like and is entirely confluent internally with the trochlea. 



(4) The entocondylar process is considerably stouter, this implying powerful pronators and flexors. 

 This indicates that Pararnys was less completely arboreal in its adaptations than was Notharctus. 



Paramys was provided with stout claws, operated by the powerful flexor muscles, which it used in climb- 

 ing. The humeri of later simplicidentate rodents diverge into easily recognizable arboreal, fossorial- 

 natatorial, fossorial, cursorial, and other types. But no rodent humerus could be mistaken for that of 

 any primate, if closely examined. 



Insectivora (Lipotyphla). Among existing Insectivora, Solenodoii has perhaps the most primi- 

 tive type of humerus. It is indeed of extremely primitive semifossorial, semiambulatory type, similar 

 to that of the most ancient Carnivora. The head is elongate anteroposteriorly and there is a promi- 

 nent warped deltoid plane, ending below in a prominent V-shaped eminence. The bicipital groove looks 

 inward rather than forward. The entocondylar process is large; the supinator crest, while well devel- 

 oped, does not flare out at the upper end. From this primitive type the humeri of other Liptotyphla 

 may readily be derived, chiefly by reduction of the crests, but in some phyla by new specializations, as 

 in the Talpidse. 



Menotyphla. I refer to this group the family Plesiadapida^, on the basis oi skeletal material of 

 N(illi(i(i('c(cs (jidlci/i Matthew (1*.)17) which the describer has generously ]ilaced in my hands for examina- 

 tion. 



The humerus ( I'late XXVII), while moi-e like that of Xollidrdu.s than any of those described above, 

 has the following special characters: 



(1) The delto-pectoral crest is very thin and acutely V-shaped, as seen from the outer side, ending 



below in a prominent pointed tip. 



(2) The supinator crest is not so large as it is in yotharcttis and does not extend up to the level of 



the deltoid tip. 



(3) The trochlea is relatively larger and more extended verticall^^ 



(4) The tuberosity for the teres major, on the inner side of the shaft, is much larger and more sharply 



defined. 



The humerus of the existing Ptilocercus might readily be derived from the Nothodectes type by the 

 lengthening of the shaft and slight reduction of the supinator crest. The Ptilocercus humerus still retains 

 nnich that is reminiscent of a Xolhodecles-Vike type, but in Tupaia the crests are further reduced; in Hhyn- 

 chocyon a cursorial ty])e of humerus, with a prominent greater tuberosity and hinge-like distal joint, is 

 finally e\'()lve(l. 



The humerus of Xothodectcs thus gives evidence, which is strengthened by that from many other 

 parts of the skeleton, for the following conclusions: 



(1) The Plesiadapida' are Menotyphla, not Lemuroidea. 



(2) The Menotyphla are the nearest known relatives of the primates. 



(3) They represent the specialized descendants of a Cretaceous pre-primate, unguiculate stock, less 



perfectly adapted for perching on the branches than were the Eocene Lemuroidea. 



Primates. The foregoing comparison of the types of humeri in various primitive mannnals also 

 adds some evidence for the following conclusions, which were long since arrived at from other evidence. 

 (1) The Mesozoic placental forerunners of the C'reodonta, Taligrada, Insectivora, etc., were ungui- 



