234 



GREGORY: NOTHARCTUS, AN AMERICAN EOCENE PRIMATE 



some of them, especially the line leading to the primates, became more agile and began to leap from 

 branch to branch. 



But little is known concerning the evolution of the claws, which are usually of assistance in primitive 

 climbing types. The claws on the oldest and most primitive known therapsid limbs {Galepus, Galechirus 

 Broom) were of primitive reptilian type, somewhat compressed, gently curved and downwardly pointed. 

 In the cynodont genus Microgomphodon Seeley (1895, plate I) the ungual phalanges were blunt and partly 

 flattened, much like those of Echidna. In Mlurosuchus broivni Broom (1906, PI. X) the single claw 

 phalanx preserved is moderately narrow, but less so than those of Ornithorhynchus. The change from 

 compressed claws to flat claws is not a profound one, and in many groups (e. g., Chelonia, Dinosauria, 

 Monotremata) we find both types represented. 



In the arboreal ancestors of the marsupials the claws became compressed, curved and pointed, as 

 they are in the opossums. The diversified descendants of the primitive Cretaceous opossums have 

 the feet variously adapted for running (e. g., Thylacinus), swimming (Chironectes) hopping {Peramelidoe, 

 Macropodidw), digging (Phoscolomys, Notoryctes), and so forth; meanwhile the primitive compressed 

 claws and the characters of the palms and soles have undergone appropriate modifications, which have 

 been described by Dollo, Bensley and others. Thus very probably the compressed claws of the opossum 

 are near the primitive marsupial type. Accordingly the marsupials furnish some instructive hints as to 

 the way in which the most jirimitive arboreal adaptations were first evolved, and the history of the group 

 shows that this arboreal stage is capable of giving rise to a great diversity of terrestrial types. 



The precise relations of the various placental orders to the marsupials are still unsettled, but it is 

 evident that after setting aside a few aberrant specializations, the marsupials as a whole represent an 

 earlier and lower grade of evolution. And since the earliest known placentals have a great many char- 

 acters in common with the primiti\'e arboreal marsupials, it seems that all the evidence so far examined 

 tends to support Dr. Matthew's view (1904) that the placentals also were originally arboreal in habit.^ 

 The construction of the hands and feet in many Eocene mammals (e. g., Hyopsodus, Meniscotherium^ 

 Thryptacodon, Vulpavus, Limnocyon, Dissacus, Periptychus, etc.) seems especially to favor this view, 

 since their hands and feet were spreading and the pollex more or less divergent — an apparent remnant 

 of better developed grasping power. The pads on the palms and soles of primitive placental mammals- 

 are homologous with those of arboreal types (Whipple) and may owe their peculiar placing to a primary 

 arboreal adaptation. As shown by Matthew arboreal life makes possible the preservation of primitive 

 characters in the limbs, but terrestrial habits, if long continued, result in evident specializations: such, 

 as enlarged claws and powerful flexor muscles for digging, compressed elongate feet for running, etc. 



Long before the opening of the Paleocene record, the placentals had differentiated into distinct 

 orders, many of which had already become terrestrial. Assuming provisionally the ultimate derivation 

 of the placentals as a whole from an arboreal stock, the main outstanding questions with regard to the 

 origin of the primates is: were the primates eventually derived from terrestrial unguiculate placentals,. 

 and were they thus secondarily arboreal mammals? Or were they derived from a primitive marsupio- 

 placental group which had not yet come down out of the trees? 



Categorical replies to these questions may not be safely given, but the evidence available at present 

 seems to point to the following hypothesis: at a very early period the opossum-like arboreal placentals 



' Some time after this memoir was set up in type I received a paper from Mr. J. W. Gidley, " Significance of divergence of the first 

 digit in the primitive mammaUan foot " (Journ. Washington Acad. Sci., IX, May 19, 1919, pp. 273-280), in which he opposes the hypo- 

 thesis of the arboreal ancestry of the placental mammals. After a candid study of Mr. Gidley's paper I see no reason for modifying, 

 either the present discussion or the conclusions stated above on pages 70, 71. W. K. G. 



