metamorphosis takes place with the chrysalis attached by the tail but supported also by a fine girdle of silk round the middle so that the head is uppermost. The larvae also afford characters by which some of the families may be distinguished—those of the Papilionidae for example having a process on the back which can be extruded or retracted.

Owing to the great size of the family of the Nymphalidae, in which the number of species approaches 5000, it is convenient to deal with the eight sub-groups into which it has been divided. The characters serving to mark off the sub-groups from one another are various. Sometimes it is the minuter structure of the tarsus, at others the form of the caterpillar or the chrysalis, at others the arrangement of the nervures that form the skeleton of the wing. Into these systematic details, however, we need not enter more fully here^[11]. What is important from the standpoint of mimicry is that these divisions, made solely on anatomical structure, correspond closely with the separation of models from mimics. Of the eight sub-families into which the Nymphalidae are divided four, viz. the Danainae, Acraeinae, Heliconinae, and Ithomiinae, provide models and some, but far fewer, mimics; two, the Satyrinae and Nymphalinae, provide many mimics and but few models, while two groups, the Morphinae and Brassolinae, practically do not enter into the mimicry story.

Simple mimicry, explicable, at any rate in theory, on the lines laid down by Bates, is a phenomenon of not infrequent occurrence in tropical countries, though rare in more temperate lands. In each of the three great divisions of the tropical world we find certain groups of butterflies serving as models, and being mimicked by butterflies belonging as a rule to quite different groups. Speaking generally the models of any given region are confined to a few groups, while the mimics are drawn from a greater number. In Asia the principal models belong to the Danaines, the Euploeines, and to a group of swallow-tails which from the fact that their larvae feed on the poisonous Aristolochia plant are generally distinguished as the "Poison-eaters," or Pharmacophagus group. Of these the Danaines and Euploeines are closely related and have much in common. They are usually butterflies of medium size, of rather flimsy build and with a somewhat slow and flaunting flight. In spite, however, of their slight build they are toughly made and very tenacious of life. Most butterflies are easily killed by simply nipping the thorax. There is a slight crack and the fly never recovers. But the collector who treats a Danaid in a way that would easily kill most butterflies is as likely as not many hours after to find it still alive in his collecting box or in the paper to which it may have been transferred when caught. They give one the impression of being tougher and more "rubbery" in consistence than the majority of Lepidoptera. Moreover, the juices of their bodies seem

to be more oily and less easily dried up. In general colour scheme they vary a great deal. Some, such as Danais chrysippus (Pl. IV, fig. 1), are conspicuous with their bright fulvous-brown ground colour and the sharp white markings on the black tips of their fore wings. Others again such as Danais septentrionis (Pl. I, fig. 3), with a dark network of lines on a pale greenish ground, are not nearly so conspicuous. Of the Euploeines some have a beautiful deep blue metallic lustre (cf. Pl. II, fig. 4), though many are of a plain sombre brown relieved only by an inconspicuous border of lighter markings (cf. Pl. I, fig. 10).

Both Danaines and Euploeines serve as models for a great variety of species belonging to different groups. Danais septentrionis (Pl. I, fig. 3) is a very abundant species in India and Ceylon, and in the same region there are several other very similar species. Flying with them in Northern India are two species of Papilio, P. macareus and P. xenocles (Pl. I, fig. 4), which resemble these Danaids fairly closely. In Southern India and Ceylon one of the two forms of Papilio clytia (Pl. I, fig. 7) is also regarded as a mimic of these Danaids. In the same part of the world there is a Pierine of the genus Pareronia, whose female is very like these Danaines on the upper surface (Pl. I, fig. 1). The male of this Pierine is quite distinct from the female (Pl. I, fig. 2).

The common Danais chrysippus (Pl. IV, fig. 1), found in this region, has been described as probably the most abundant butterfly in the world, and serves

as a model for several species belonging to different groups. It and its mimics will, however, be described in more detail later on. Mention must also be made of the striking case of the Danaid, Caduga tytia and its Papilionine mimic P. agestor from Sikkim (Pl. II, figs. 2 and 3). In both species the fore wings are pale blue broken by black; while the hind wings are pale with a deep outer border of rusty red. Not only in colour but also in shape the swallow-tail bears a remarkable resemblance to the Danaid. C. tytia is also mimicked by a rare Nymphaline Neptis imitans, which exhibits the same striking colour scheme so very different from that of most of its allies.

No less remarkable are some of the cases in which the Euploeines serve as models. E. rhadamanthus, for example, is mimicked by the scarce Papilio mendax, and a glance at Figs. 8 and 9 on Plate II shews how well this butterfly deserves its name. Euploea rhadamanthus also serves as a model for one of the several forms of female of the Nymphaline species Euripus halitherses. In some Euploeines the sexes are different in appearance—a somewhat unusual thing among butterflies serving as models in cases of mimetic resemblance. Such a difference is found in Euploea mulciber, the male being predominantly brown with a