136 Determinate Evolution 



great objections to the 'adequacy of natural selection ' are 

 so impressive that the Neo-Darwinians have felt obliged 

 to deal with them. The first objection may be called that 

 from ' non-useful characters/ and the latter that from 

 'correlated variations.' * 



On the other hand, the doctrine of use-inheritance or 

 Lamarckism is open in my opinion to still graver diffi- 

 culties, (i) It is a pure assumption that any such inheri- 

 tance takes place. The direct evidence for it is practically 

 nothing. 2 No unequivocal case of the inheritance of the 

 definite effects of use or disuse has yet been cited. Again 

 (2) it proves too much, seeing that if it actually operated 

 as a general principle it would hinder rather than advance 

 evolution in its higher reaches. For, first, in the more 

 variable functions of life it would produce conflicting lines 

 of inheritance of every degree of advantage and disadvan- 

 tage, and these would very largely neutralize one another, 

 giving a sort of functional 'panmixia' of inherited habits 

 analogous to the panmixia of variations which arises when 

 natural selection is not operative. Again, in cases in which 

 the functions or acquired habits are so widespread and 

 constant as to produce similar ' set ' habits in the individ- 

 uals, the inheritance of these habits would produce, in a 

 relatively constant environment, such a stereotyped series 

 of functions, of the instinctive type, that the plasticity 

 necessary to the acquirement of new functions to any 

 great extent would be destroyed. This state of things 

 is seen in the case of certain insects which live by com- 



1 See the discussion of them with reference to Romanes' theory of instinct, 

 above, Chap. V. 



2 See the candid statement of Romanes, loc. cit., and cf. Lloyd Morgan, 

 Habit and Instinct, Chap. XIII. 



