100 Plant-Breeding 



themselves, while less adaptive forms would tend similarly 

 to disappear. Now the same thing would take place 

 if this individual or its adaptive offspring were to cross 

 with the main stock of the parent species ; for all the 

 offspring of such a cross which is intermediate in character 

 and therefore less adapted to the new conditions would 

 tend to disappear, and the two types would, as a result, 

 become more and more fixed and the tendency to cross 

 would constantly decrease. 



The refusal to cross, the result of natural selection. 

 The refusal to cross, therefore, becomes a positive character 

 of separation, and the " missing links" that result from 

 crossing are no more or no less inexplicable than the 

 ''missing links" due to simple selection; or, to state the 

 case more accurately, natural selection weeds out the tend- 

 ency to promiscuous crossing, when it is hurtful, in the 

 same way that it weeds out any other injurious tendency. 

 It makes no difference in what way this tendency ex- 

 presses itself, whether in some constitutional refusal to 

 cross, if such exists, or infertility of offspring, or 

 in different times of blooming: all equally come under 

 the power of natural selection. We are likely to look upon 

 infertility as the absence of a character, a sort of negative 

 feature which is somehow not the legitimate field of 

 natural selection; but such is not the case. We are 

 perhaps led the more to this feeling because the word 

 infertility is itself negative, and because we associate 

 full productiveness with the positive attributes of plants. 

 But loss of productiveness is surely no more a subject 

 of wonder than loss of color or size, if there is some corre- 

 sponding gain to be accomplished. In fact, we see, in 



