Appendix //. 317 



characters whose utility is unkr-own are nevertheless useful, 

 and therefore due to natural selection. In other words, 

 the question is not — Are there not many useful specific 

 characters whose utility is unknown? but — Does it follow 

 from the theory of natural selection that all specific 

 characters must necessarily be useful ? Well, it appears to 

 me that without going further than the above passage, 

 which Mr. Wallace has quoted, we can see clearly enough 

 what was Darwin's opinion upon the subject. He did not 

 believe that it followed deductively from his theory that all 

 specific characters must necessarily be useful; and therefore 

 he regarded it as a question oi fad — to be determined 

 by induction as distinguished from deduction — in what 

 proportional number of cases they are so. Moreover he 

 gives it as his more matured opinion, that, "as far as we can 

 at present judge" (i.e. from the present state of observation 

 upon the subject : if, with Mr. Wallace, his judgement were 

 a priori, why this qualification?), he had not previously 

 sufficiently considered the existence of non-adaptive characters 

 — and this he ended by believing was one of the greatest 

 oversights as yet detected in his work. To me it has always 

 seemed that this passage is one of the greatest exhibitions of 

 candour, combined with solidity of judgement, that is to be 

 met with even in the writings of Darwin. There is no talk 

 about any deductive " necessity " ; but a perfect readiness to 

 allow that causes other than natural selection may have been 

 at work in evoking non-adaptive characters, so that the fifth 

 edition of the Origin of Species was altered in order to 

 confine the theory of natural selection to ".adaptive changes" 

 — i.e. to constitute it, as I have said in other words, 

 "a theory of the origin, or cumulative development, of 

 adapiaiions," 



If to this it be said that in the above passage there 

 is no special mention of species, the quibble would admit 

 of a threelold reply. In the first place, the quibble in 



