i STRING OF ASSUMPTIONS 61 



the variations so originated are almost 

 infinite in variety, and that some of them 

 are almost sure, at some time or another, 

 to "turn up trumps," or in other words 

 to be accidentally in a useful direction ; 

 we have only to assume, again, that 

 these will be somehow continued and 

 developed through embryotic stages 

 until they are fit for service ; we have 

 only to assume, again, that there are 

 adjustments by which serviceability, 

 when transmuted into actual use, has 

 power still further to improve all 

 adaptations by some process of self- 

 edification ; then, making all these 

 assumptions, we may explain anything 

 and everything in the organic world. 

 But in such a series of assumptions 

 we do not speak the language of 

 true physical causation. This is what 

 Mr. Spencer now confesses. " Natural 

 selection," he says, " could operate 



