204 TRANSACTIONS OP SECTION D. 



evidence it seems clear that tHe geographical element must enter largely into 

 any explanation that may be offered. On the whole, it is certainly the case 

 that the forms which are supposed to be related by mimicry do inhabit the 

 same localities; the continental Prioneris, for example, is like the continental 

 Delias, and the island Prioneris recalls the island, not the continental, Delias. 

 Moreover, we find the differences between the Delias of Timor, of Sumbaw-a and 

 Sumba, reflected in the associated Huphinas of the same islands. If it be 

 granted that the geographical element is a factor, it is natural to inquire how 

 it works. 



It is no doubt true that external geographical conditions are occasionally 

 capable of producing, whether directly or indirectly, a community of aspect in 

 the animals or plants exposed to their influence. The prevalence of a sandy 

 coloration in the mammals and birds of a desert, and of whiteness in the 

 inhabitants of the arctic snow-fields, the spiny character so often assumed by 

 the plants of arid regions, and the general dwarfing of the vegetation that grows 

 close to the sea, may be given in illustration. At first sight these phenomena 

 may seem to be of the nature of direct effects of the environment; quite 

 pofsibly some of them are so, but I think that few observers would deny that 

 they are at least largely adaptive, being used for pui-poses of aggression or 

 defence. Still, even if we allow the direct effect of the environment, as perhaps 

 we may do especially in the case of the plants, can we frame any hypothesis 

 of the action of geographical conditions which shall lead directly to the assump- 

 tion of a common pattern in the case of the three or four butterflies from New 

 Guinea ? I confess that I am quite unable to do so. If the climate, or the soil, 

 or any other geographical condition in New Guinea is capable of directly 

 inducing so remarkable a combination of colour as we see in these Pierines and > 

 Nymphaline, why does it not affect other organisms in a similar way? Why do 

 not other Pierines, for instance, closely related to ornytion and abnormis, share 

 in the same coloration? And considering the characteristic aspect of the under- 

 side, which is supposed to be called into being by some unexplained condition 

 peculiar to New Guinea, we maj' well ask. Why should its most conspicuous 

 features belong in the one case to the forewing and in the other to the hind- 

 wing, and vire i^ersa. the general effect being the same? 



Fruhstorfer, we may note, does not feel these difficulties. 'Many Pierids,' 

 he says, 'present typical examples of that resemblance to other butterflies which 

 has been named Mimicry. The origin of this resemblance, however, is now 

 explained by the supposition that the mimics were modified by the same (as 

 yet unknown) influences under which the colouring of the models, mostly Danaids, 

 developed.' I think it will be generally agreed that this reference to 'unknown 

 influences ' is no explanation at all. 



It is necessary tn take into account the fact that the resemblances of which 

 we are speaking are independent of structural differences, being, in fact, merely 

 superficial. This is a point which is capable of much wider demonstration than 

 I am giving it to-day. But even from the instances now before us I think there 

 cannot be much diffictilty in coming to the conclusion that the resemblances are 

 an appeal to vision. They are meant to be seen, though by whom and for what 

 purpose may be open to question. Speculations as to recognition and sexual 

 attraction may, I think, in these cases be put out of court; but there remains 

 the theory of warning colours assumed in reference to the attacks of vertebrate 

 enemies. From the fact that the most striking and most conspicuous of these 

 common aposemes or danger-signals belong to the imder surface — that is to say, 

 the part chiefly exposed to view during rest — it may be inferred that the enemies 

 to be guarded against are mainly those that attack butterflies not on the wing, 

 but when settled in repose. Both birds and monkeys are known to feed on 

 butterflies, and there is a good deal of evidence as to their preference for one 

 kind of food over another. I will not stop to give details, but anyone who 

 wishes to studv the evidence may be referred especially to the Memoirs of 

 Dr. G. A. K. Marshall, Mr. C. F. M. Swynnerton. and Captain G. D. H. 

 Carpenter. 



If the warning-colour interpretation of these resemblances be the true one, 

 we see at once why they are so largely independent of structure and affinity. 

 Being meant to catch the eye, they ride rough-shod, so to speak, over incon- 



