122 rilOCEKDINGS OF THE ACADEMY OF [Jan., 



The sacrum in anthropoinorpha is composed of five or six 

 coalesced vertebra; in monkeys the normal number is two or three, 

 and a like number is shown by lemurs, except Indris, which ha.'i 

 four, and Perodlctlcus and Arctocebus, each of which has five.'* 



Now if we attempt, from Gratiolet's standpoint, to account for 

 the presence in anthropoids of so many of the above characters aa 

 their supposed ancestors do not possess, inheritance being excluded 

 by the very terms of the hypothesis, we are driven to analogous 

 variation as the only process with which we have any acquaintance 

 which might be held competent to explain them. 



But, so far from there being any good reason to assume that 

 analogous variation has been a frequent method in nature, there 

 is, on the contraiy, warrant for an a priori belief that the mere 

 mathematical chances against the occurrence of any single case of 

 it are very great ; so that where, as in the present circumstances, 

 seven cases of the independent development of almost exactly simi- 

 lar characters must have taken place in each of four sfenera (to say 

 nothing of man, who is not provided for by the hypothesis), the 

 improbability becomes so enonnous as to remove it from rational 

 consideration.'^ 



The theories of Darwin and of Cope remain to be examined, and 

 it may be said at once that no one of the homologies which have 

 been noted is excluded by either of them, but there is, in my belief, 

 a wide difference in their relative probability ; that of Cope being 

 so far the most simple, that it is logically indicated for our accept- 

 ance. 



Darwin's hypothesis requires us either to suppose that there has 

 been an extensive and complicated process of preservation of cer- 

 tain structures and suppression of others, in which the family 

 groups now differ, or to take refuge again in analogous variation. 

 Both are rendered difficult of acceptance by the reflection that of the 

 characters here advanced few, if any, can be believed to have been 

 adaptive. It is unsafe to dogmatically assert that a given structure 



^^Mivart, P. Z. -S'., 1805, p. 500. 



*® Furthermore, the same principle must he invoked to account for the 

 absence of the cusp-hearing heel on m^, possessed by each of the supposed 

 ancestral genera ; for the presence of three external [cusps on the lower 

 molars ; the pre.«ence of a vermiform appendix ; the independent origin of 

 the left common carotid froTu the arch of the aorta, and the converging di- 

 rection of the hair on the arms toward the elbow, all of which are peculiar 

 to anthropomorpha. 



