62 THE EVOLUTIONARY THEORY 



culty that the utiHty of many variations does not 

 appear until they have become somewhat pronounced, 

 so that in their incipient and shghter beginnings they 

 afford no apparent basis for natural selection. 



Other difficulties attend the supposition that species 

 are evolved by the selection of sHght and continuous 

 variations. One has been mentioned in another 

 connection — the swamping effect of inter-breeding 

 between new forms and other members of the parent 

 stock. Isolation would indeed remove this difficulty, 

 but the chances of such isolation are not sufficiently 

 numerous to meet the requirements of natural evolu- 

 tion. Moreover, continuous variations fluctuate, and 

 are limited in range. They show a tendency to slow 

 down at certain stages, and reversion to type is a 

 frequent phenomenon. They are also purely linear. 

 That is, they constitute quantitative rather than qual- 

 itative changes, and cannot produce a difference in 

 kind, such, for instance, as is represented by the evo- 

 lution of an eye from a blind spot. At any rate con- 

 tinuous variations can produce a difference in kind 

 only by a coincidence and co-ordination of several 

 such variations, which can hardly be explained without 

 supposing directive forces to be at work in the organ- 

 ism of which the theory of natural selection takes no 

 account. To this should be added the necessity that 

 similar variations, with all their complexity of co-ordi- 

 nation, should appear at once in a sufficient number 

 of individuals, and with sufficient persistence for the 

 process to be accomplished of propagating and estab- 



