1894 THE FACTOES OF ADAPTIVE EVOLUTION 337 



that, even if it does, you are not concerned with 

 this fact — i.e. of the development of the adaptation, 

 but only with its origin. 



(b) All the same, however, we must remember 

 that where high elaboration of mechanism is con- 

 cerned, the question as to the causes of its develop- 

 ment become of more importance than those of its 

 origin; e.g. even if self-adaptation be conceived 

 capable of making a first step towards producing 

 the exquisite mechanism of a bivalve shell, by 

 discriminate variation, how is it conceivable that it 

 should go on through the odd millions of successive 

 steps of improvement needed to produce the perfect 

 mechanism in which the great wonder of adaptation 

 really occurs ? 



I can conceive of no natural process to accomplish 

 this development even in one such case of mechanism 

 other than natural selection. Let alone the ' endless 

 variety ' of elaborate mechanism elsewhere. 



(c) Of course, if you coxddL prove that indiscrimi- 

 nate variations have not occurred in wild plants, but 

 only under cultivation, you ivould destroy Darwinism 

 — intoto. But is the proposition ctedlhle a priori ; 

 or sustainable a posteriori, &c. ? 



I suppose you have read Wallace on the subject 

 as regards wild animals, and if you were to make 

 similar measurements ^ith regard to wild plants, you 

 would obtain analogous results. 



I remember as a boy having a game of who could 

 find most specimens of fern-leaved clover in a given 

 time, or even two leaves of clover which would be 

 exactly ahke in all respects. But I have already 



