THEORIES OF EVOLUTION 117 



If we consider all of these selection theories care- 

 fully we find that such as are valid, in that they 

 deny any developmental role of the environment, 

 are almost equally well described as isolation 

 theories. The remainder are either insecure or dis- 

 tinctly Lamarckian. To the extent that choice is 

 involved, as in artificial and sexual selection, a proc- 

 ess may be truly selective, but an active selective 

 factor is usually lacking in natural selection. In 

 every case some condition tends to preserve a lim- 

 ited extent of the normal range of variation of a 

 species in the offspring of a limited number of in- 

 dividuals, and in the original statement of the 

 theory this factor is the survival of the favored 

 individuals and the death of the remainder. 



It is evident that the individuals which survive 

 in any generation must perpetuate a great many 

 characters which have no decisive part in their 

 survival. Out of the assemblage of inherited char- 

 acters found in a species, relatively few can be 

 regarded as useful in the Darwinian sense, to the 

 extent of preserving those individuals in which 

 they are slightly more effective in relation to a 

 given condition. Most characters must be regarded 

 as indifferent in relation to any particular factor 

 in the environment as long as they serve the in- 

 dividual normally, yet they vary like all char- 

 acters and since many individuals perish some 

 of the indifferent variations must inevitably be 



