292 Heredity. 



be admitted that new conditions of life do sometimes 

 definitely affect organic beings, it may be doubted 

 whether well-marked races have often been produced by 

 the direct action of clianged conditions, Avithout the 

 aid of selection, either by man or nature" ( Variation, 

 347-352). 



AVhile we acknowledge the great weight of this reason- 

 ing we must bear in mind that evidence to show that 

 new forms of life are not produced, without the aid of 

 selection, by direct modification, is not necessarily proof 

 that the causes of variation have no relation to the p^:»r- 

 j)ose of the modification — that variations are, so f^r as 

 their use goes, purely fortuitous. 



Even if external changes do not give rise to useful 

 modifications, unless they are aided by natural selection, 

 it may still be true that they play an important and 

 essential part. 



It may be true that a change of conditior-s does not 

 necessarily produce a change of structure, and yet true 

 that when a change of structure does take place it is due 

 to the changed conditions. 



It may be true that an unfavorable change in the 

 environment has no power to produce a compensating 

 chan2:e of hereditarv structure without tne aid of natural 

 selection, and yet true that this external change may be 

 the cause of variation in the part affected. 



If this latter supposition be a fact the work of natural 

 selection will be almost infinitely simplified, for in place 

 of an indefinite number of fortuitous variations, it will 

 be furnished with variation of the part in which change 

 is needed, and it is only an even chance whether a change 

 in a part which is out of harmony with its environment 

 . be favorable or unfavorable. 



According to our theory of heredity, when an organ- 



