the Coloration of Eggs. 573 



from detection. The Hawks lay very variable eggs ; some 

 few of tliem regularly lay white eggs, and white is not an 

 unusual variation even amongst those species that do not. 

 Still, w^e have here yet another possible counteragent to 

 variability, a limited one, in the general family resemblance 

 that runs through the many variations. One finds the same 

 thing in the Acraeinse, a subfamily of variable butterflies, 

 and hence it was undoubtedly sometimes of use in preventing 

 attack by the birds on which I have experimented. The Field- 

 fare, a very "fightable^' bird (Wallis), and one that tends 

 also to nest in colonies, lays rather specially variable eggs, 

 and so do those notorious warriors the Drongos, though 

 here, again, an alternative or supplementary explanation is 

 possible: for in some Drongos, perhaps in all, the variability 

 is of the more definite kind that I am disposed to attribute 

 rather specially to active selection and may perhaps have 

 been in part selected in relation to Cuckoos ■^. 



Finally, there is the probable factor of abundance in a 

 species. It is obvious, I think, that an enemy may become 

 acquainted and remain acquainted with a considerable range 

 of variation in a highly abundant species with, actually and 

 proportionately, a far smaller loss to the latter than would 

 fall to the lot of a scarce species, and that, other things 

 being equal, the selection will be less severe, and the conse- 

 quent permission to vary greater, for an abundant species 

 than for a scarce one. Certainly most of the more highly- 

 variable species I am acquainted with are also common. 



* An objection to this explanation for polymorphism would he that 

 many species in the eggs of which it occurs do not need it for protection 

 from Cuckoos, seeing that none of their eggs are matched by that of any 

 local Cuckoo. It must be remembered, however, that the explanation 

 is only regarded as one of several ; that in some cases, too, the form 

 resembled by a Cuckoo may have died out ; that other cases may repre- 

 sent a Cuckoo defeated and forced to follow other channels, especially if 

 Cuckoos' eggfi should possess the relatively ready adaptability that 

 would seem to be possessed by members of mimetic genera in butterflies, 

 and that, in any case, even a quite distant resemblance should be taken 

 into account, seeing that I found in my experiments that such a 

 resemblance usually sufficed to secure adoption. 



