274 gidl^y: primitive; mammalian foot 



placentals are likewise of arboreal origin. This latter belief was 

 apparently based, principally at least, on the fact that the first 

 digit, in both the fore and hind feet of the early Eocene mammals 

 of generalized type, is so frequently found in a divergent posi- 

 tion, and on the condition found in the earliest known Primates. 

 But unhke Huxley and Dollo, who considered arboreal adaptation 

 in the marsupials a specialization and not a primitive condition, 

 Matthew advanced the hypothesis that opposability of the first 

 digit in the early mammals was a primitive condition. While I 

 do in the main indorse Matthew's characterization of a hypo- 

 thetical "common ancestral group," I am unable to agree with 

 him regarding arboreal ancestry as applied to placentals in gen- 

 eral, and therefore differ from him in the interpretation of the 

 conditions found in the primitive mammalian foot. In order 

 more clearly to set forth my own view I here quote the prin- 

 cipal arguments advanced by Matthew in support of his inter- 

 pretation : 



(i) In the first place, as far as we can trace back the history of each 

 of the arboreal groups, we find their first ancestors with the first digit 

 as fully opposable as in the modern representatives {e. g., the Middle 

 Eocene primate Northarctus). 



(2) Second, in those groups which have not an opposable thumb, we 

 find as we trace back their ancestry that the trapezium, whose form and 

 facets give the surest indication on this point, approaches more and 

 more nearly to the type preserved in the Primates, etc. 



(3) In the four Basal Eocene mammals (Pantolambda, Euproto- 

 gonia, Claenodon, and Dissacus) in -which this part of the skeleton is 

 known, the form of the bone [trapezium] is surprisingly uniform, and 

 when the manus is put together, the first digit is thrown partly out- 

 ward from the rest of the hand, and permits of much freer motion than 

 the remaining digits, with a considerable degree of opposition. 



Following this, in the same article, Matthew states that the 

 primitive opposability of the hallux is less clearly indicated but 

 suggests this is due to the probability that the evolution of the 

 hind foot for terrestrial locomotion "began earlier or proceeded 

 more rapidly." He further suggested that "the hypothesis 

 that all mammals passed through a stage when the pollex and 

 hallux were opposable," would explain among other things, 

 "(a) the presence of but two phalanges on digit i, three on each 

 of the others ; (6) the epiphysis of digit i being proximal as in the 



