vin DARWINISM AND DESIGN 151 



to regard our action as one out of a series of actions 

 displaying no intelligence. He would cheerfully admit 

 that the action seemed intelligent, and by itself would 

 justify the inference to a real intelligence behind it. But 

 he would urge, if I take it as the one intelligent action 

 out of an indefinite number of unintelligent actions, there 

 is nothing in it that need cause surprise or calls for the 

 assumption of real intelligence. We might try to convince 

 him by multiplying the symptoms of intelligence, but in 

 vain. For, though he would admit the growing improb 

 ability of such a continuous series of apparently purposive 

 actions, he could still expand the context of non-purposive 

 actions rapidly enough to maintain his theory of their 

 chance origination. 1 



If, therefore, an indefinite number of non- adaptive 

 variations be really granted, no adaptations, however 

 numerous and complete, can ever prove an intelligent 

 cause of variation. Even if all the known facts testified 

 aloud to the operation of an adapting intelligence, the 

 Darwinian assumption might still be used to disprove all 

 teleology, if unbounded license were given for the invention 

 of hypothetical variations ! Now, of course it is not 

 contended that variations as known are all obviously 

 adaptive ; it is claimed rather that we do not know 

 enough about them to say what their actual character 

 is. But it must most strenuously be asserted that the 

 Darwinian theory cannot be quoted as destructive of the 

 action of purposive intelligence in organic evolution until 

 the occurrence of indefinite variation has been raised from 

 the position of a methodological device to that of an 

 incontestable fact. 



Even then it may be doubted whether the fortuit 

 ous character of the facts could ever be rendered incon 

 testable. To defy refutation by the facts the teleologist 

 has merely to adopt a device analogous to that of his 

 opponent. Just as the latter could always assume a 

 non-teleological extension of what seemed a teleological 

 ordering, so the former can always assume a secret 



1 Cp. pp. 71-2. 



