SELECTION 147 



Exposure to new conditions must be con- 

 tinued for generations to set up any great 

 variation; but this, once set up, continues 

 indefinitely. Changed conditions may di- 

 rectly influence the whole organization of 

 the creature or certain parts alone; or they 

 may act indirectly through the reproductive 

 system. With respect to the direct action, 

 the nature of the organization seems to 

 count for more than that of the conditions. 

 The effect on the offspring may be definite: 

 e.g. size may depend upon the amount of 

 food, colour upon quality of food, thickness 

 of skin and hair upon climate, etc. But 

 indefinite variability is a much commoner 

 result of changed conditions, and has prob- 

 ably played a more important part in the 

 formation of our domestic races. The re- 

 productive system is peculiarly sensitive to 

 very slight external changes. Many plants 

 and animals will not reproduce in domesti- 

 cation, even though individually vigorous; 

 others, though weak and sickly, breed freely. 

 Hence we need not be surprised at the 

 reproductive system acting irregularly and 

 producing variations. But that variation is 

 not exclusively associated with sexual re- 

 production is demonstrated by the case of 

 plants :< sporting' through bud- variation. 

 Such cases, moreover, prove that the nature 



