October 31, 1907] 



NA TURE 



677 



of the hindwing, near the base. These marks reach a 

 high state of developEiient in some members of the Eastern 

 genus Delias, and relics of them are to be seen in the 

 common white butterflies of our own country. 



Now no one who accepts mimicry at all will be inclined 

 to doubt the existence of a mimetic relation between 

 Hclicotiius guaricus and the Pierine Pereute leucodrosifue. 

 How has it been brought about? The dark colour and 

 red band are not at all characteristic of Pierines, and 

 have no doubt been copied from the Hcliconius ; but the 

 Pierine red spots have passed the other way, being taken 

 up by the Heliconius from the Pierine. This, I believe, 

 was the first case of mimetic interchange noticed. 



Another instance. White is not an ancestral colour in 

 Heliconius ; it is ancestral in the Pierines. The Pierine 

 P. lociista falls by its undersurface into mimetic associ- 

 ation with the group of Heliconius represented by 

 H. alithea and Tl . galaitihus. The white colour has 

 passed from Pieris to Heliconius ; the dark, in great 

 measure, from Heliconius to Pieris. 



Much the same has happened in the rase of Heliconius 

 /(■lire and the female of Ficiis noctipennis. These two 

 have undergone reriprnca! chanfe. The white comes from 

 the Pieris, the black from the Heliconius. 



There is another case where two species belonging to 

 widelv separated sections of the same subfamily are in 

 uuestion. The hindwings of the island form of the 

 Pierine IJtihhina nfrissa have been drawn awav into 

 '"litntion of the hindwing of Ixias halicttsis, also a 

 Pierine ; while the forewing of the Ixias, leaving the usual 

 aspect of its genus, has been assimilated to the forewing 

 of Hunhina. 



.\nother good example of interchange is afforded by 

 the " swallowtail " Papilio rex from Uganda, which is 

 in undoubted mimetic relation with the Danaine Melinda 

 jormosa. The brown at the base of the forewings is a 

 Danaine character adopted by the Papilio : the pale areas 

 at the base of the hindwings are a Papilionine character 

 adopted by the Daniiine. Each has in one of these re- 

 spects acted as a model to the other. The two African 

 genera .Mylothris and Phrissura, the species of which form 

 'A parallel series of mimetic pairs analogous to the 

 Ithomiines and Heliconiines of tropical America, furnish 

 what is probably another instance of the same pheno- 

 menon. For this mutual approach by a process of give 

 and take on both sides. Prof. Poulton has proposed the 

 apt term " diaposematism," the idea of reciprocity being 

 conveyed by the Greek particle "dia." 



Let us now look at the working of this reciprocal 

 princiole in another direction. 



It is well known that where the sexes differ in the 

 extent to which they are protected, whether by power of 

 concealment or by other means, it is almost invariably 

 the female that has the advantage. This was pointed out 

 long ago by Dr. Wallace, and was, no doubt rightly, 

 attributed by him to the fact that the continuance of the 

 life of the female, as the guardian of the early stages 

 of the future brood, was of greater importance than 

 that of the male to the welfare of the species. So we 

 find many cases in which the female alone is mimetic, 

 not the male. Shall we say that " Nature abhors the 

 unorotected female"? 



We have already noticed the case of PapHio dardanus, 

 with its non-mimetic male, and three or four different 

 forms of female, each form in mimetic relation with a 

 Danaine model. 



The case of the Pierine Leiiceronia art^in is in manv 

 respects parallel to that of the Panilio. .Again we have a 

 non-mimetic male, and several different forms of female, 

 each being in mimetic relation with another butterfly of 

 quite different affinities. 



Once more : the female of the Nvmnhaline butterflv 

 Hypolimnas hnlinn falls into mimetic association with 

 Danaines and with a Panilio, leaving its own male out- 

 side the group. Tt will he remembered, also, that much 

 the same thine occurs with certain of the South American 

 Pierines which we considered in an earlier part of the 

 present lecture. 



These instances are sufficient to show the readiness of 

 the female, as distinct from the male, to enter a mimetic 

 combination. 



NO. 1983, VOL. 76] 



Bearing these facts in mind, when we look at such a 

 combination as that of Papilio iphidamas with Euterpe 

 approximata, we shall have little or no difficulty in 

 recognising that here we have an analogous case. The 

 se.xes of the Euterpe (a Pierine) and the female of the 

 Papilio all resemble each other, while the male of the 

 Papilio stands apart. We have just seen how readily the 

 female of a given species may be drawn away into a 

 mimetic relation apart from its own male, and we have 

 every reason to suppose that the same has occurred here, 

 only that in the case of P. dardanus, L. argia, and 

 //. bolina the pull has been mainly or entirely away 

 from the dardanus, argia, and bolina standard, while here 

 there is no doubt that the female Papilio has pulled the 

 Pierine away from the usual Pierine standard, though it 

 has in turn been pulled away from its own male. The' 

 male, it is true, belongs to a synaposematic group of its- 

 own, but the female has joined the stronger combination. 

 The pull has been mutual between female Papilio anJ 

 Pierine, and the association must therefore be Mullerian. 



(2) Hero is one more piece of evidence. We have seerr 

 that from the nature of the case the attraction (so to 

 call it) in a Miillerian assemblage acts, or may act, iir 

 all directions, for each member of a Mullerian group is 

 potentially both mimic and model. 



It ought then sometimes to happen, if the Miillerlarr 

 theory is correct, that although one dominant species, dis- 

 tinguished perhaps for its hardiness and distasteful 

 qualities, may act as the centre of a group, influencing 

 all the other members, yet that these other (subordinate) 

 members of the group should show signs of having in- 

 fluenced each other, apart from the dominant species. 



Does this ever happen? Certainly it does. 



Limnas chrysippus, a Danaine, is one of these dominant 

 forms, numerous in individuals, hardy, conspicuous, proved 

 to be distasteful, and accompanied by mimics wherever 

 it goes. Among its mimics in .Africa are an Acra^a, and 

 a Lycajnid (allied to our common blues). Now it is quite 

 evident that there is a mimetic relation between the 

 Acrsa and the Lycaenid apart from that which exists 

 between them both and the Limnas. In short, they 

 resemble each other in some respects more than either 

 of them resembles the common model. Hence one or 

 other of these two, or perhaps both, must be distasteful, 

 and therefore there must be a Mullerian element in the 

 whole group, if, indeed, it be not entirely Mullerian. 



It has been pointed out by Prof. Poulton that many 

 beetles, belonging to different families, are all in a sense 

 mimics of the hymenopterous group of the Mutillidse, and 

 vet they have become assimilated to each other in non- 

 Mutilloid points. 



Facts of this kind prepare us for a further consider- 

 ation of great importance, with which I shall conclude. 



(3) We find that it is impossible to regard the mimetic 

 assemblages of a given region as so many isolated groups. 

 .As a matter of fact, though there are certain dominant 

 forms which act, so to speak, as centres of attraction, 

 we often find that the mimetic forms constitute a nexus, 

 models of the most dissimilar aspect being held together, 

 as it were, bv a kind of connection which runs from 

 group to group, gradations from one group to another 

 being formed in the most unexpected ways. From the 

 nature of things this point is difficult to illustrate within 

 the limits of a lecture like the present; a few examples 

 mav suffice. 



Starting from a white Pieris of ordinary aspect like 

 P. phaloe, we ran pass by easy stages through 

 P. calydonia Q and P. demophile 9 into a well-marked 

 distasteful group of which the Ithomiine Aeria agna is 

 a good example. From P. demophile O we can also 

 pass through P. viardi Q to the pattern of Heliconius 

 charilonia, or in yet another direction bv wav of 

 P. tithorrides Q to Heliconius atthis and Tithorea 

 pavonii. P. calydonia 9. ''S^'". S'ves us a fresh starting 

 point from which to proceed by way of P. ki(aha. 

 P. pandosia, and P. leptalina up to a well-marked 

 Ithomiine group tvpified by Nnpeogenes inachia. We 

 have already seen how P. locusta r^ , which presents the 

 Pierine characters of P. phaloe in an intensified form, 

 comes into association with another group of Heliconii, 

 while the same butterflv forms an early link in the chain 



