IXFLUENCE OF ISOLATION ON TPIE FORMATION OF SPECIES 291 



or more species, but the areas of each are not sharply isolated, and 

 the division into species cannot, therefore, be due to isolation. But 

 it is not difficult to guess upon what it depends, when we know that 

 the males alone are of a beautiful yellow and black colour, while the 

 females are of a greenish protective colouring all over. 



Thus, in my opinion, sexual selection plays a part more or less 

 important in the origin of the numerous endemic species of diurnal 

 Lepidoptera which are characteristic especially of the islands of the 

 Malay Archipelago, and which make the Lepidopteran fauna there 

 so rich in individuality. A large number, indeed the majority of 

 the types of Papilionidse, have a peculiar species, a local form, on 

 most of the larger islands, which is sharply and definitely distinguished 

 from those of the other islands, usually in both sexes, but most 

 markedly in the much more brilliantly coloured males. 



Thus each of these types forms a group of species, each of which 

 is restricted to a particular locality, and lias usually originated where 

 we now find it, although of course the diffusion of one of these 

 large strong-flying insects from one island to the other is in no way 

 excluded. As an example we may take the Priamus group, the 

 blackish yellow Helena group, the blue Ulyssus group, and the pre- 

 dominantly green PerantJais group. 



If we inquire into the causes of this divergence of forms and 

 their condensation into numerous species, we shall find that their 

 roots lie in this case, as in that of all transformations, in germinal 

 selection and the variational tendencies resulting therefrom, but Ave 

 must reyard their fixation as the result of isolation, which prevented 

 the variational tendencies which happened to develop on any one 

 island from being neutralized and swamped by mingling with the 

 variations of other islands. But that sexual selection took control 

 of these strikinp- colour-variations and increased them still further 

 is obvious from the rarely absent dimorphism of the sexes. Even 

 if the females do not consciously select mates from among the males, 

 they will more readily accept as a mate the one among several suitors 

 which excites them most strono-lv. And that will be the one which 

 exhibits the most brilliant colours or exhales the most agreeable 

 perfume, for we know from their behaviour in regard to flowers how 

 sensitive butterflies are to both these influences. 



Although isolation has an important role in the formation of all 

 these species, it seems to me an exaggeration to maintain, as many 

 naturalists do, that the splitting up of a species is impossible without 

 isolation. Certainly the splitting up of species is, in numerous cases, 

 facilitated by isolation, and indeed could only have been brought 



U 2 



