62 



AFRICAN MIMETIC BUTTERFLIES 



cenea, but that in this case both sexes are concerned. My reasons for so thinking are that 

 they have been taken in coitu several times, that specimens occur presenting intermediate 

 coloration, and also that the two forms are always found together whenever they are met 

 with in any number. I have not often been fortunate enough to see these congregations, 

 but I remember seeing some thirty or forty specimens, comprising about equal numbers of 

 each form, collected together on the side of a steep shady kraantz along the Palmiet River. 

 Mr. A. D. Millar, who has collected for many years in Durban, tells me this is by no means 

 uncommon, and that they congregate particularly in the afternoon when going to roost. 

 Mr. C. N. Barker teUs me that some years ago he came across a large number of both forms 

 on a large tree on which they had evidently bred,^ for many of them had only just emerged, 

 and some had not their wings fully developed. This shows that they are also associated in 

 their early stages.' 



The same writer subsequently had no opportunity of pursuing the investigation, and 

 some years were destined to elapse before proof of the species identity of the forms could 

 be obtained by breeding. At the meeting of the Entomological Society of London, in April 

 1910, Professor Poulton read a preliminary note communicated by Trimen, ' on Mr. A. D. 

 Millar's experimental breeding of forms of the Nymphaline genus Eur alia ' } The facts then 

 disclosed were as follows : — 



1. Ten ova from an example of wahlbergi produced two males, one female, and one 



deformed individual of the parent type, and also four males and one female 

 mima. 



2. Thirty-nine ova from an example of mima produced thirty- three imagines, twenty- 



four males and nine females, aU of which were of the parent type. 



3. Eleven ova from an example of mima produced eight of the parent type, five 



males and three females, and three wahlbergi, aU males. 

 None of the examples resulting from the above experiments shows any departure from the 

 respective type. 



Beyond the establishment of the species identity of the two forms mima and wahlbergi 

 it would perhaps be premature to discuss the results of these experiments. It was pointed 

 out in the communication referred to that in spite of the result of the investigation mima is 

 apparently the rarer of the two forms in Natal, a consideration which, if absolutely correct, 

 would preclude an explanation similar to that suggested for the dominant condition in that 

 region of the cenea form of P. dardanus. Professor Poulton at the same time suggested that 

 the two forms might be regarded as a Mendelian pair.^ He also exhibited six examples of 

 anthedon and four of dubius taken, in the same locality near Lagos in 1909, pointing out 

 that by analogy with the eastern forms mima and wahlbergi, their western representatives 

 might with confidence be regarded as specifically identical. 



The meaning of these polymorphic forms is very obscure. On the theory of mimicry 

 we should expect a mimetic butterfly to have developed an approach to that model the 

 pattern of which afforded the best protection. On MiiUerian grounds the tendency must 



^ This assumption must now be corrected in view 

 of the fact that Millar's experiments show that the 

 food plant is a species of nettle (Fleurya, sp. ?). 



2 Aurivillius includes Euralia in Hypolimnas. 



^ I have not in the present work attempted to deal 

 with the subject of Mendelism. Even were I qualified 

 by a special knowledge of Mendelian principles to 



discuss their bearing on the present subject, I should 

 hardly deem the time ripe for such an effort. The 

 intricacies of Mendelian heredity are being carefully 

 studied by scientists well qualified for the investiga- 

 tion, and their bearing, if any, on the phenomena of 

 mimicry will doubtless be duly discussed when 

 adequate data become available. 



