THE LAMARCKIAN PRINCIPLE 61 



our present subject. In each ovary of the Yucca flower there 

 are about 200 ovules. The grub of the Yucca moth does not 

 eat by any means all of these and quite enough are left. 

 Apparently, cross-fertilisation is so important to the plant that, 

 in order to obtain it, it is worth sacrificing half the ovules 

 produced. 



It is among insects that we find such instincts most perfectly 

 developed. In every case the fate of the species depends upon 

 the exact performance of a difficult work, and it would seem 

 that Nature dare not entrust this work to intelligence that can 

 learn only by blundering. But though such instincts are found 

 in their most elaborate perfection among insects, it is upon 

 instincts among other animals also that the propagation of the 

 species depends. And for the existence of these Lamarckism is 

 unable to account. 



In speaking of the antlers of the stag and the various co- Lamarck- 

 adaptations I pointed out how difficult it is for Lamarckism and ^"jg"* 1 

 Weismannism to go in harness together. We must trust either mannism 

 to the inheritance of acquired characteristics or to Natural Selec- J|J ; by 

 tion, always ready to make use of so-called spontaneous varia- 

 tions, but not to the co-operation of the two principles. It is 

 true they are not diametrically opposed to one another, and 

 Darwin himself did not entirely exclude the former. But if not 

 mutually antagonistic, it is difficult to make them pull together 

 when we find them at work in competing species, or in one and 

 the same organism. 



I shall show in the chapter on Natural Selection that when 

 one species preys upon another the two must advance pari passu, 

 else one of two things must happen : either the preyed upon will 

 no longer be caught and killed (with the result that their enemies 

 will starve), or the preying species will annihilate its means of 

 subsistence. Suppose that a species of butterfly protects itself 

 from birds who live upon it by resemblance to the leaves of trees. 

 If the resemblance becomes too exact the bird will search in vain, 

 and must either go unfed, or seek some other food. If, on the 

 other hand, the bird's eyesight improves more rapidly than the 

 protective coloration of the butterfly, then it is likely that this 



