From these and many other examples it is evident that in butter flies the female is more commonly mimetic than the male. Fur thermore when both sexes mimic, the female's likeness to a model is often the more complete. Apart from mimicry, female butter flies also appear in two or more forms far more commonly than males, and recent researches have shown that the females are more variable. It is therefore probable that the evolution of female mimicry has been facilitated by the quantity and variety of the material ready to be built up by natural selection.
Returning to a consideration of the advantages secured by mim ;cking the great tropical groups we find that their species possess conspicuous patterns made up of contrasted tints, the pattern of the under-surface of the wings being generally similar to that on the upper, so that when these butterflies fly with their character istic flapping or sailing flight, or rest with wings closed, the colours and patterns, once learned, are easily recognized. Many of the species when disturbed emit an unpleasant odour which in some species is known to be secreted by special glands. When offered to insect-eating animals they are rejected except under the stress of hunger, and they have been seen to be attacked and rejected by enemies in the wild state. Their tissues are soft and flexible so that they can often recover from the injuries of experimental tasting. Just as these groups of butterflies are foremost in pro viding models for mimicry so are they conspicuous in the posses sion and in the advertisement of qualities which protect against attack. Mimicry of the advertisement suggests possession of the qualities. That young insect enemies do learn to associate the advertisement with the qualities has been proved by the interest ing experiments of Lloyd Morgan. It must be remembered that the qualities are of the most varied kinds. Thus C. F. M. Swyn nerton has shown that large African butterflies Charaxes, although palatable, are rejected because of their toughness, thus accounting for the mimetic likeness to them borne by smaller and less tough species of the genus; also for the resemblance of larger species to one another.
The two figures in the second row on Plate I. show that the species of the great tropical groups are not only mimicked by other butterflies but also that they mimic each other. Thus several African Acraeines mimic Danaines, while in both groups the species of one genus sometimes mimic those of another; the same is true of the Oriental Danaines. In tropical America the few Danaines and many of the Heliconines mimic species of the domi nant group, the Ithomiines. This last mimetic association has been a fruitful source of confusion, for the superficial resemblance between the Ithomiines and an important section of the Helicon ines is so strong that both groups were united under the Heli conidae by the older naturalists. And even when H. W. Bates recognized the wide difference between them he still left them as a single family divided into the Danaoid Heliconidae (the Itho miinae) and the Acraeoid Heliconidae (the Heliconinae proper). Hence the models for mimicry, Ithomiine as well as Heliconine, are commonly called "Heliconidae." But in the meantime natural ists have recognized the true distinction, placing the two sections far apart, the Ithomiinae next to the Danainae, and the Heli coninae next to the Acraeinae. Now in the numerous mimetic associations between the two groups, the Ithomiinae, with hardly an exception, are the models and the Heliconinae the mimics. Thus the first figure in the second row represents the Ithomiine model, Melinaea mothone, and the second figure its beautiful Heliconine mimic H. aristiona, both from Peru. But as the latter is a Heliconine, it and other Heliconine mimics of Ithomiines are constantly referred to as models, thus reversing the true relation ship, which is well shown by an exhibit, in the Oxford University museum, of the butterflies indiscriminately captured on two days in central British Guiana. There were taken on the first day 216 specimens of the chief Ithomiine model with two Heliconine mimics, belonging to different genera, on the second day six months later, of the model 220, and no mimic.