Temperature Experiments on tivo Tropical Butterflies. 457 



it. In the example before us, remembering the short ten 

 days in which Natural Selection has to act, and presuming 

 that in some time past the misippus % was somewhat 

 similar to the ancient type of the ^ , we may ask, how 

 could a specimen, or specimens, with a few red scales 

 scattered over the wing, be noticed and subsequently 

 avoided by birds, by any lesson they may have previously 

 learnt from the capture of Danais chrysippns ? 



On the assumjDtion that this is a case of Batesian 

 mimicry, a bird tasting a red speckled specimen, would 

 ascertain at once that it Avas palatable, and the red scales 

 in others would not save them from destruction. The 

 likeness towards chnidpims, therefore, could not progress. 



If it is a case of Mlillerian mimicry, where both species 

 are unpalatable, a red speckled unpalatable one, as also its 

 unspeckled companions, after a few experiments, would be 

 left severely alone and nothing would be gained, for if the 

 bird could discriminate it Avould neglect them, and if it 

 could not, both would equally suffer. 



Lastly, if the resemblance is due to the experimental 

 attacks of young birds, the emergence of the butterflies 

 should coincide with the time that tasting experiments 

 are taking place, but in Ceylon it so happens that the 

 young birds are olf the nest and foraging for themselves 

 in May, and misippus is not on the wing until October. 



We could the more readily understand the process if 

 the mutation was sudden and large, but the specimens 

 exhibited negative such a supposition. 



Some supporters of the mimicry theory, among others 

 Mr. Pocock and Prof, Poulton, consider " that the first 

 steps towards a mimetic hkeness are not caused by a few 

 differently coloured scales," but " by a large colour varia- 

 tion which was enough to produce a rough resemblance, 

 and that Natural Selection gradually produced out of this 

 a detailed resemblance." At first sight this looks like 

 mutation pure and simple, but it is not necessarily so. 



The evolution of the species is internal, and the large 

 variation Prof. Poulton speaks of may be the cumulative 

 effects of an increasing number of differently coloured 

 scales in many generations. We have only to assume 

 the disappearance of such intermediates to arrive at a 

 "large colour variation." 



I see nothing that prevents our believing that if internal 

 evolution can produce a large variation, a continuance of 



H H 2 



