346 COL. C. SWINHOE ON MIMICRY IN BUTTEBPLIES 



Different conditions under wJiich mimicry may appear : 

 attempted explanation. 



Knally our facts have an instructive bearing upon the very- 

 different conditions under which mimicry may appear in the 

 most closely related species. It seems clear that we have to do 

 with two species which are unable to exist without this deceptive 

 resemblance to some specially protected form, either in both 

 sexes or in the one which is chiefly exposed to attack. Wherever 

 we find these butterflies, whatever changes they may undergo, 

 the resemblance which enables them to live upon the reputation 

 of some local distasteful species is maintained. Mimicry being 

 equally necessary to both misippus and iolina in order to ward 

 off extermination, we nevertheless find that it pursued an utterly 

 different course in these two species. Hypolimnas misippus has 

 attached itself to a single well-known, conspicuous, wide-ranging 

 species of distasteful butterfly, resembling it with great fidelity, 

 and following it through the details of even minor changes. In 

 order to achieve this result, it has been compelled to depart very 

 widely from the ancestral form — even more so than is the case 

 with any of the holina group. But this extreme variation in one 

 direction appears to have deprived it of the power of developing 

 variations in other directions ; so that its existence and range 

 seem to depend upon the existence and range of a single butter- 

 fly, Danais cJirysip'pus and its varieties. In Hypolimnas Iolina, 

 on the other hand, we meet with much greater elasticity : its 

 range is almost unlimited as regards the conditions imposed by 

 mimicry, for it can vary in each locality into the semblance of 

 some local species. 



How is this wide divergence to be explained ? Many biolo- 

 gists would be inclined to lay stress on the amount and kind of 

 individual variation which has been at the disposal of the selective 

 process during the development of the mimetic resemblance ; 

 and it is certain that the results must have been largely influenced 

 by this. It is noteworthy that holina includes forms which are 

 both older and younger than those oi misippus, the latter repre- 

 senting but a single one out of the many phases of departure 

 from the ancestral type represented by the former. It may be 

 that this comparatively narrow limitation of misippus is merely 

 due to the exclusive predominance of a single specially advan- 

 tageous resemblance, Danais cJirysippus being so abundant and 

 well-known in the localities where it occurs, and its distribution 

 affording scope for a wide range. Or variation may liave carried 



