lo Darwinism Verified. [i. 



mistake entirely the conditions under which scientific 

 evidence can be procured. To estimate properly the 

 value of any hypothesis it is necessary that we should 

 know what kind and degree of proof to expect ; and 

 in the present case we must not look for a demon- 

 stration that shall be direct and simple. Instead of 

 a universal property of matter, so conspicuous as to 

 be recognised at once by the inspection of a few 

 striking instances, we have in the theory of natural 

 selection to deal with a very complex process, working 

 results of endless diversity throughout the organic 

 world, and often masked in its action by accompany- 

 ing processes, some of which we can detect without 

 being able to estimate their relative potency, while 

 others no doubt have thus far escaped our attention 

 altogether. Accordingly, while we may consider it 

 as certain that natural selection is capable of working 

 specific changes in organisms, we may at the same 

 time find it impossible to give a complete account of 

 the origin of any one particular species through natural 

 selection, because we can never be sure that we have 

 taken due notice of all the innumerable concrete cir- 

 cumstances involved in such an event. The theory, 

 therefore, cannot be adequately tested by any single 

 striking instance, but must depend for its sup- 

 port on the cumulative evidence afforded by its 



