A FINAL WAENIKG. 405 



universal difficulty of deciding whether a certain modification 

 which has taken place is to be ascribed to some direct determin- 

 ing and transforming cause, or to the enhancing of a previously 

 modified character, which is frequently connected with selection, 

 that it becomes imperative that we should in the first place 

 carry out the most exact research possible by means of experi- 

 ment, and also wean ourselves of the convenient — but, as it 

 seems to me, highly pernicious — habit of theoretical explana- 

 tions from general propositions. Otherwise there is great danger 

 that the bright expectation which Darwin has opened out to us 

 by his theory may be baffled — the prospect of gradually bring- 

 ing even Organic Being within reach of that method of inquiry 

 which seeks to discern mechanical efficient causes. 



