The following paper gives an account of a series of breeding experiments which were made with the butterfly Papilio polytes , Linn., in the course of one and a half years’ research on insect polymorphism in Ceylon. In the first instance the experiments on butterflies were not limited to this species, but unfortunately all others proved refractory in captivity and were one by one discarded. More time therefore could be given to this, the only amenable species, and it was possible to make a number of observations on the habits of the imago which are suggestive in several ways, and justify a more extensive reference than is usual in papers on heredity. The first chapter of this paper, therefore, is devoted to observations on the species in captivity, especially to the pairing habits of the imagines. The second chapter deals with the results obtained from the pedigree breeding and their Mendelian interpretation, and a final chapter may be added to allow of a comparison with similar work on other species and of a summary of the conclusions to be derived from the investigation as a whole. P. polytes , as a classical example of mimicry, is too well known to need a full description here. As a subject for a detailed investigation it has peculiar advantages. It is trimorphic; this trimorphism is confined to the female sex; of the three forms of female, one resembles the male in general pattern, while the other two are mimetic, their models being P. hector , Linn., and P. aristolochiæ , Fabr. It would naturally be supposed that the female like the male would be known as the polytes ♀, but this is not the case, as one of the mimetic forms was first described under this name. To avoid confusion and to facilitate subsequent reference, the various forms of the species may be stated thus:— P. polytes ♀ form cyrus , Fabr., resembles the male. P. polytes ♀ form romulus , Cram., resembles P. hector , Linn. P. polytes ♀ form polytes , Linn., resembles P. aristolocliæ , Fabr.