DIFFICULTIES OF INDUCTIVE VERIFICATION. 437 



incubation, and the young afterwards. Evidently all these 

 differences affect the proportion between the total cost of re- 

 production and the total cost of individuation. 



Whether the species is monogamous or polygamous, and 

 whether there are marked differences of size or of structure 

 between males and females, are also questions not to be over- 

 looked. If there are many females to one male, the total 

 quantity of assimilated matter devoted by each generation to 

 the production of a new generation, is greater than if there 

 is a male to each female. Similarly, where the requirements 

 are such that small males will suffice, the larger quantity of 

 food left for the females makes possible a greater surplus 

 available for reproduction. Another cause has a like effect. 

 Where the habits of the race render it needless that both 

 sexes should have developed powers of locomotion where, 

 as in the Glow-worm and sundry Lepidoptera, the female is 

 wingless while the male has wings the cost of Individuation 

 not being so great for the species as a whole, there arises a 

 greater reserve for Genesis: the matter which would other- 

 wise have gone to the production of wings and the using of 

 them, may go to the production of ova. 



Other complications, as those which we see in Bees and 

 Ants, might be dwelt on; but the foregoing will amply serve 

 the intended purpose. 



333. To ascertain by comparison of cases whether Indi- 

 viduation and Genesis vary inversely, is thus an undertaking 

 so beset with difficulties, that we might despair of any satis- 

 factory results, were not the relation too marked a one to be 

 hidden even by all these complexities. Species are so ex- 

 tremely contrasted in their degrees of evolution, and so 

 extremely contrasted in their rates of multiplication, that the 

 law of relation between these traits becomes unmistakable 

 when the evidence is looked at in its ensemble. This we shall 

 soon find on ranging in order a number of typical cases. 



In doing this it will be convenient to neglect, temporarily, 



