454 DIRECTIVE FACTORS IN EVOLUTION: 



must not be used hurriedly in depreciation of the role of 

 selection in natural wild conditions. (a) Pure or inbred 

 lines are not typical of wild stocks, in which cross-fertilisa- 

 tion is of frequent occurrence, (h) It is dangerous to argue 

 from very short-lived experiments to the age-long processes 

 of Nature, (c) It is premature to deny the possibility of 

 stable germinal variations occurring in a pure or inbred 

 line. If one did, it might be the starting-point of a new 

 advance. In any case there remains a great role for Natural 

 Selection in eliminating certain lines or races and favouring 

 others in its ceaseless sifting. 



§ 6. Subtlety of Selection Theory. 



Natural Selection is a technical expression for a manifold 

 and almost ubiquitous process of sifting, which discriminates 

 in life and in death between the relatively more fit to the 

 given conditions and the relatively less fit. It must always 

 be thought of in the Here and Now, i.e., in reference to 

 particular conditions of space and time. There are three 

 reasons why it is important to keep this obvious fact in view, 

 (a) It is a frequent and pernicious error to suppose that 

 there is any sort of ceaseless winnowing towards an ideal 

 of fitness, except perhaps self-consistency. The only common 

 character of surviving variants is that they survive, — they 

 must have consistent viable constitutions suited to particular 

 conditions, which may be those of parasitism or putridity. 

 The fallacy of supposing that Natural Selection necessarily 

 works towards ^ fitness ' in the colloquial sense is largely due 

 to thinking of the process abstractly and hypostatising it, 

 and to misunderstanding the word ' fit ', which means merely 

 relatively advantageous in given conditions, making for sur- 

 vival in short. But the error is also due to a shrewd per- 



