SERIAL ORGANIC EVOLUTION 243 



might prune an oak certainly, but would never 

 shape a single leaf of it. 



If we are to make selection the determinant of 

 species, and believe that, save for selection, variation 

 could never have produced the species, we have 

 to assume selective influences at least as subtle 

 and as sensitive as a sculptor's hands, and variations 

 almost infinite in number. And if we are consist- 

 ent, we must carry this principle back to the fire- 

 mist, and believe that the orbit of every electron 

 and the permanent position of every atom in the 

 earth was the result of numerous tentative com- 

 binations, with a survival of the fittest. We must 

 believe that the position of every atom and every 

 cell in an oak tree was the result of selective 

 addition and superposition. We must believe 

 that every tgg was an experiment ; and when we 

 remember that every Q,gg has to produce not only 

 an organism, but an organism with the capacity of 

 repeating itself, the odds against the adequacy of 

 variation and selection seem hopelessly great. 



There is nothing new about the idea of evolu- 

 tion ; it is at least as old as Empedocles, probably 

 much older ; and Aristotle stated the position 

 almost as it is to-day when he wrote : " Yet it may 

 be said that they (the teeth) were not made for this 

 purpose, but that this purposive arrangement came 

 about by chance ; and the same reasoning is applied 



