of these birds, or five thousand butterflies in all^[7]. But if these five species, instead of shewing five distinct warning patterns, all displayed the same one it is evident that the education of the birds would be accomplished at the price of but one thousand butterfly existences instead of five. Even if one of the five species were far more abundant than the others it would yet be to its advantage that the other four should exhibit the same warning pattern. Even though the losses were distributed pro rata the more abundant species would profit to some extent. For

the less abundant species the gain would of course be relatively greater. Theoretically therefore, all of the five species would profit if in place of five distinct warning patterns they exhibited but a single one in common. And since it is profitable to all concerned what more natural than that it should be brought about by natural selection?

Müller's views are now widely accepted by students of mimicry as an explanation of these curious cases where two or more evidently distasteful species closely resemble one another. Indeed the tendency in recent years has been to see Müllerian mimicry everywhere, and many of the instances which were long regarded as simple Batesian cases have now been relegated to this category. The hypothesis is, of course, based upon what appears to man to be the natural behaviour of young birds under certain conditions. No one knows whether young birds actually do behave in the way that they are supposed to. In the absence of any such body of facts the Müllerian hypothesis cannot rank as more than a plausible suggestion, and, as will appear later, it is open to severe criticism on general grounds.

Perhaps the next contribution to the subject of mimicry which must rank of the first importance was that of Erich Haase^[8], to whose book students of these matters must always be under a heavy obligation. It was the first and still remains the chief work of general scope. Since Haase's day great numbers of

fresh instances of mimetic resemblance have been recorded from all the great tropical areas of the world, and the list is being added to continually. Most active in this direction is the Oxford School under Professor Poulton to whose untiring efforts are largely due the substantial increases in our knowledge of African butterflies contributed by various workers in the field during the past few years. Whatever the interpretation put upon them, there can be no question as to the value of the facts brought together, more especially those referring to the nature of the families raised in captivity from various mimetic forms. With the considerable additions from Africa^[9] during the past few years several hundreds of cases of mimicry must now have been recorded. Some of the best known and most striking from among these will be described briefly in the next two chapters.

CHAPTER III

OLD-WORLD MIMICS

The earlier naturalists who studied butterflies made use of colour and pattern very largely in arranging and classifying their specimens. Insects shewing the same features in these respects were generally placed together without further question, especially if they were known to come from the same locality. In looking through old collections of butterflies from the tropics it is not infrequent to find that the collector was deceived by a mimetic likeness into placing model and mimic together. During the last century, however, more attention was paid to the anatomy of butterflies, with the result that their classification was placed upon a basis of structure. As in all work of the sort certain features are selected, partly owing to their constancy and partly for their convenience, the insects being arranged according as to whether they present these features or not. Everybody knows that the butterflies as a group are separated from the moths on the ground that their antennae are club shaped at the end, while those of the moth are generally filamentary and taper to a fine point.

Figs. 1-8. Terminal portion of front legs of butterflies belonging to different families. (After Eltringham.)

The butterflies themselves may be subdivided into five main groups or families^[10] according to the structure of the first of their three pairs of legs. In the Papilionidae or "swallow-tails," the first pair of legs is well developed in both sexes (Fig. 8). In the Pieridae or "whites," the front legs are also similar in both sexes, but the claws are bifid and a median process, the empodium, is found between them (Fig. 7). In the remaining three families the front legs differ in the two sexes. The females of the Lycaenidae or "blues" have well-developed front legs in which the tarsus is terminated by definite claws (Fig. 5), whereas in the males the terminal part of the leg, or tarsus, is unjointed and furnished with but a single small claw (Fig. 6). This reduction of the front legs has gone somewhat further in the Erycinidae (Figs. 3 and 4), a family consisting for the most part of rather small butterflies and specially characteristic of South America. In the great family of the Nymphalidae the reduction of the front legs is well marked in both sexes. Not only are they much smaller than in the other groups, but claws are lacking in the female as well as in the male (Figs. 1 and 2).

Though the structure of the fore limbs is the character specially chosen for separating these different families from one another, it is of course understood that they differ from one another in various other distinctive features. The chrysalis of the Nymphalidae for example hangs head downwards suspended by the

tail, whereas in the Pieridae and Papilionidae