140 Dr. H. Eltringham on Specific and 



Thus nanna and rosina are much rarer than phyllis and 

 colombinus, whilst, on the other hand, xenoclea is much 

 commoner than notabilis microclea. He points out, how- 

 ever, the possibility that in some or, perhaps, most of the 

 localities, members of one section may be always more 

 dominant than the other. Also that the exact times of 

 appearance may not coincide. 



Hitherto it has been the custom to assign all such cases 

 of mimicry either to the Batesian or Miillerian category. 

 Those of us who have heard Mr. Swynnerton's remarks 

 on his experiments in South Africa, given some time ago 

 before the Linnean and Entomological Societies, know that 

 the relative distastefulness of insects is a variable character 

 dependent on factors not entirely related to the insect 

 itself. Mr. Swynnerton's papers are not yet published, 

 so that I cannot utilise his results for the present discussion. 

 Meanwhile, we know that, although no direct experiments 

 have been made with Heliconius, the genus gives great 

 evidence of being a distasteful one, and we may therefore 

 regard the existing resemblances as being of a Miillerian 

 character. We are, however, faced with the difficulty 

 of the multiplicity of patterns. If Section II contains 

 the models it would appear to be a great disadvantage to 

 the forms to have so many different appearances, since 

 each form is not necessarily confined to a special locality 

 but considerable overlapping occurs. Where mimicry 

 occurs between separate species of separate genera, then 

 multiplicity of patterns in the models may be compensated 

 for by diversity of habit of the different species. Thus 

 Mr. Kaye, in discussing mimetic groups in the Potaro 

 district of British Guiana, states that the species of Helico- 

 nius with patterns resembling Ithomiinae frequent flowers 

 of a certain plant, whilst other species of Heliconius are 

 never, in his experience, found on these flowers. This is 

 an extremely valuable point and one which should be 

 remembered in considering mimetic phenomena. 



It does not, however, seem probable that forms of the 

 same species will have different habits corresponding 

 to those of their respective models. Apart from the 

 preferences of insectivorous enemies, whether absolute or 

 conditional, the unpalatability of the insect is, of course, 

 a relative factor. Thus resemblance of a species of Heli- 

 conius in Section I to a species in Section II may be a 

 measure of protection to both, assuming the two species 



