OF THE GENUS HYP0LIMNA3. 345 



The ancestral non-mimetic form from which the mimetic varieties 

 have been derived : various phases of development of mimicry. 



The ancestral form of both groups is preserved in the closely 

 sirailar non-mimetic males, and tlie rare cases of reversion to the 

 same type exhibited by the females. But the beautiful evidence 

 supplied by the existence of the ancestral non-mimetic form of 

 both sexes in certain islands is wanting here, although so well 

 seen in the inerope group. 



The most ancestral form described in this paper is probably 

 the Fijian holina, in which the females exhibit a transition from 

 non-mimetic to mimetic forms ; then would follow the Indian 

 holina, in which the female is not a very perfect mimic of 

 Euploea core, and still retains traces of the blue spots so charac- 

 teristic of the non-mimetic males, culminating in the Celebes 

 form, in which the mimicry of the female is fairly complete and 

 has entailed a more marked divergence from the normal type 

 than any other form in this group : at this stage misippus must 

 be placed, with its non-mimetic male and females with extremely 

 perfect and detailed mimicry. We finally reach the climax of 

 change in those island forms of holina in which the males also 

 are mimetic, and in Africa, where no more ancestral phase is at 

 present known. 



Bearing upon mimetic resemblance to different species in one 

 locality. 



The well-known mimetic resemblance to two or more very 

 diflerently coloured species of distasteful insects in the same 

 locality is not well exemplified, although it appears probable that 

 some varieties of the females from Fiji bear this interpretation, 

 which may also in part explain the occurrence of all three 

 varieties of the female misippus at Aden, where the three corre- 

 sponding forms oi Danais are also found (viz. chrysippus, alcip- 

 pus, and dorippus). But here, too, we meet with nothing that 

 approaches the condition of some species of the merope group of 

 the S. -African Papilio cenea for example, in which four forms of 

 the female respectively mimic such differently coloured species 

 as Danais chrysippus, Amauris dominicanus, and two varieties of 

 Amauris echeria, thus widening the area of possible mistake so 

 far that the mimetic species can become comparatively numerous 

 without the risk of extermination. 



