SIMILARITY OF SUCCESSIVE TYPES. 83 



Darwin attributes these gaps to the imper- 

 fection of the geological record, and believes 

 they will be filled up by future discoveries. 

 But even then the difficulty would not be 

 removed, because, as we have seen, the 

 same objection applies in principle in the 

 case of the smallest gap for example, where 

 variation arises only after birth. 



Further, the argument based on the sim- 

 ilarity of successive types is negatived by 

 the facts of retrogression that is, where the 

 successor is less highly specialised than its 

 antecessor. But we discuss retrogression in 

 another chapter. 



The generally higher specialisation of suc- 

 cessive types, and the position of their re- 

 spective fossil remains in successive geological 

 strata, we accept as conclusive evidence of a 

 process of evolution, and their similarity 

 suggests the idea that there is something in 

 common between them ; but similarity does 

 not prove that the one was evolved from the 

 other by secondary causes. To infer that the 

 steam-engine of to-day was evolved from the 

 first machine of the kind, without human in- 

 tervention, would be an absurdity unworthy 

 of notice, and the difference between a mam- 

 mal and a fish, to go no lower, is far greater 

 than between any steam-engines. 



