346 Appendix A 



individual concerned. They will constitute a congenital predis- 

 position to that strengthening of the part which is essential for 

 survival. Variations in the opposite direction, tending to thwart 

 the adaptive modification, will be disadvantageous, and wiU be 

 eliminated. Thus, if the conditions remain constant for many 

 generations, congenital variation will gradually render hereditary 

 the same strengthening of bone structure that was provisionally 

 attained by plastic modification. The effects are precisely the 

 same as they would be if the modification in question were 

 directly transmitted in a slight but cumulatively increasing 

 degree ; they are reached, however, in a manner which involves 

 no such transmission. 



'' To take a particular case : Let us grant that in the evolution 

 of the horse tribe it was advantageous to this line of vertebrate 

 life that the middle digits of each foot should be largely devel- 

 oped, and the lateral digits reduced in size ; and let us grant 

 that this took its rise in adaptive modification through the 

 increased use of the middle digit and the relative disuse of the 

 lateral digits. Variations in these digits are no longer sup- 

 pressed and eliminated. Any congenital predisposition to in- 

 creased development of the mid-digit, and decreased size in the 

 lateral digits, will tend to assist the adaptive modification and 

 to supplement its deficiencies. Any congenital predisposition 

 in the contrary direction will tend to thwart the adaptive modi- 

 fication and render it less efficient. The former will let adaptive 

 modification start at a higher level, so to speak, and thus enable 

 it to be carried a step farther. The latter will force it to start at 

 a lower level, and prevent its going so far. If natural selection 

 take place at all, we may well believe that it would do so under 

 such circumstances.^ And it would work along the lines laid 

 down for it in adaptive modification. Modification would lead ; 

 variation follow in its wake. It is not surprising that for long 

 we believed modification to be transmitted as hereditary varia- 

 tion. Such an interpretation of the facts is the simpler and 

 more obvious. But simple and obvious interpretations are not 



1 Professor Weismann's ' germinal selection,' if a vera catisa, would be a 

 cooperating factor, and assist in producing the requisite variations. 



