EVOLUTIONARY TELEOLOGY, 3G5 



possible, except on tlie supposition that the phenomena 

 would of themselves give rise to such an inference. 

 That it is now possible to make the inference, and, 

 indeed, hardly possible not to make it, is suflScient 

 warrant of its relevancy. • 



It may, of course, be rejoined that, if this impor- 

 tant factor is given, the inference yields no indepen- 

 dent argument of a divine creator ; and it may also 

 be reasonably urged that the difference between things 

 that are made under our observation and comprehen- 

 sion, and things that grow, but have originated be- 

 yond our comprehension, is too wide for a sure infer- 

 ence from the one to the other. But the present 

 question involves neither of these. It is simply 

 whether the argument for design from adaj^tations in 

 JS^ature is relevant, not whether it is independent or 

 sure. It is conceded that the argument is analogical, 

 and the parallel incomplete. But the gut is in the 

 points that are parallel or similar. Pulleys, valves, 

 and such-like elaborate mechanical adaptations, can- 

 not differ greatly in meaning, wherever met with. 



The opposing argument is repeated and pressed 

 in another form : 



" Tlie evidepce of design afforded by the aiarks of adapta- 

 tion in works of human competence is null and void in the caso 

 of creation itself. . . . Nature is full of adaptations ; but these 

 are valueless to us as traces of design, unless we know some- 

 thing of the rival adaptations among which an intelligent being 

 miffht have chosen. To assert that in Nature no such rival 

 adaptations existed, and that in every case the useful function 

 in question could be established by no other instrument but 

 one, is simply to reason in a circle, since it is solely from what 

 we find existing that our notions of possibility and impossi- 



