164 Determinate Variation and Selection 



tribution ') as before^ yet has a mean which lies further in 

 the direction of the accomviodations themselves or in lines 

 consistent zvith them. This view is, therefore, quite con- 

 sonant with the negative answer which is probably to be 

 given to the question of fact as to determinate variation. 



The ancestors of the sole, for example, had one eye on 

 each side. Let us suppose that some of them also had a 

 certain power of adjusting the eyes by muscular strain. 

 Now those which could do this best in the way which 

 would bring the eyes closer together would have the bet- 

 ter chance of life.^ Then, in addition to the action of 

 natural selection upon those which were born with the 

 eyes closer together, there would be the further fact that 

 this acquired adjustment would save the lives of the 'ac- 

 commodating' soles.2 Not only would Weismann's intra- 

 selection have play to enable each successive generation to 

 make the same accommodation, in turn, as their fathers 

 had done before them, but there would be a directive ten- 

 dency given to the evolution of the eyes of the sole in the 

 matter of relative position. For while, originally, the strug- 

 gle had been between those which could adjust the eyes in 

 this manner and those which could not, the survival to 

 maturity of the accommodating ones only would bring 

 it about that only these would be fertile, all the next gen- 

 eration would have the power of some accommodation, and 

 the mean would thus be shifted in this direction. The 

 best accommodation would always be made by those whose 



1 By reason of some advantage, such as that arising from a flat position near 

 the bottom, with other adaptations for better concealment, as is explained be- 

 low, Chap. XIV. § 3. 



2 Professor C. B. Davenport suggests in a private letter that the principle 

 of organic selection might be described as * the survival of the accommo- 

 dating.' 



