Determinate Variation 161 



work. Admitting * determinate evolution,' the question as 

 to the causes which 'determine' the evolution is never- 

 theless still open, and various answers have been given to 

 it. The Neo-Lamarckians say * use-inheritance ' (as Eimer, 

 who calls the determination secured by this means 'ortho- 

 genesis'); Weismann says 'germinal selection'; those who 

 accept ' organic selection ' say that it is a determining 

 factor (the resulting determination of evolution being 

 called ' orthoplasy ') ; others say 'determinate variation* 

 (continued in the same direction for successive genera- 

 tions) ; Professor Osborn says, ' determinate variation ' 

 with 'organic selection.' Determinate variation, then, in 

 the proper meaning of that term, is only one way of account- 

 ing for determinate evolution^ and to the writer it is not 

 the true way ; at any rate, it is not necessarily involved 

 in the theory of 'organic selection.' 



Let us look more closely at ' determinate variation.' 

 Supposing that by variation we mean ' congenital varia- 

 tion,' then we may ask : When are variations determinate ? 

 When for any reason they are distributed in a way different 

 from that required by the law of probability or chance. 

 The problem of determinate variations is purely one of 

 distribution; and is to be investigated for each gener- 

 ation, quite apart from its holding for a number of succes- 

 sive generations (and so giving 'determinate evolution'). 



Further, the possible determinateness of variation is to 

 be distinguished carefully from the extent or width of 

 variation. By ' extent ' of variation is meant the limits of 

 distribution of cases about their own mean ; while relative 

 determinateness means the distribution of cases, according 

 to some other law than that of probabilities, about a mean 

 established for the parents in the earlier generation. 



M 



