The third main group of models characteristic of this region belongs to the Papilionidae. It was pointed out by Haase some 20 years ago that this great family falls into three definite sections, separable on anatomical grounds (see Appendix II). One of these sections he termed the _Pharmacophagus_ or "poison-eating" {26} group owing to the fact that the larvae feed on the poisonous climbing plants of the genus _Aristolochia_. It is from this group that all Papilios which serve as models are drawn. No mimics of other unpalatable groups such as Danaines are to be found among the Oriental Poison-eaters. In the other two sections of the genus mimics are not infrequent (cf. Appendix II), though probably none of them serve as models. To the Pharmacophagus group belong the most gorgeous insects of Indo-Malaya--the magnificent Ornithoptera, largest and most splendid of butterflies. It is not a large proportion of the members of the group which serve as models, and these on the whole are among the smaller and less conspicuous forms. In all cases the mimic, when a butterfly, belongs to the _Papilio_ section of the three sections into which Haase divided the family (cf. Appendix II). _Papilio aristolochiae_ (Pl. V, fig. 5), for example, is mimicked by a female form of _Papilio polytes_, and the geographical varieties of this widely spread model are generally closely paralleled by those of the equally wide spread mimic. For both forms range from Western India across to Eastern China. Another poison-eater, _P. coon_, provides a model for one of the females of the common _P. memnon_. It is curious that in those species of the poison-eaters which serve as models the sexes are practically identical in pattern, and are mimicked by certain females only of the other two Papilio groups, whereas in the Ornithoptera, which also belong to the poison-eaters, the difference between the sexes is exceedingly striking. {27}