242 SCIENCE AND IMMORTALITY 



differences between, say, a starfish and a man. 

 It may be easy to conceive of the translation of 

 two similar species — to picture the circumstances 

 that might make an ape into a man ; but if varia- 

 tions at present tend to swing to and fro about 

 a mean, and if we cannot conceive of variations 

 and selective action sufficient to translate species 

 into species all along the supposed genealogical 

 line, we cannot accept selective evolution as a 

 whole. It must be realised that it is a very differ- 

 ent thing to believe that one species may turn into 

 another species, and that hundreds of species, by a 

 process of variation and selection, may be consecu- 

 tively transformed along the arborescent line from 

 amoeba to man. Each succeeding transformation 

 becomes more and more mathematically improbable, 

 since, species by species, the number of possible 

 variations involved in the evolution increases, and, 

 species by species, some new variation in the 

 selective process must be invented. Say that the 

 chances are ten to one against an amoeba being 

 converted into a starfish ; they are about a million 

 to one against its conversion into a codfish, and so 

 on in increasing ratio. 



Granting infinite germinal variations, we have 

 yet to find machinery sufficient to select the right 

 ones, and such machinery we have not found. 

 The selective agencies that have been suggested 



