From experiments with cultivated plants such as primulas and sweet peas, we have learnt that this discontinuous form of inheritance which occurs in _P. polytes_ is the regular thing. Moreover, we have plenty of historical evidence that the new character which behaves in this way is one that has arisen suddenly without the formation of intermediate steps. The dwarf "Cupid" form of sweet pea, for instance, behaves in heredity towards the normal form as though the difference between them were a difference of a single factor. It is quite certain that the "Cupid" arose as a sudden sport from the normal without the intervention of anything in the way of intermediates. And there is every reason to suppose that the same is true for plenty of other characters involving colour and pattern as well as structure, both in the sweet pea, the primula, and other species. Since the forms of _polytes_ female behave in breeding like the various {92} forms of sweet pea and primula there is every reason to suppose that they _arose_ in the same way, that is to say, as sudden sports or mutations and not by the gradual accumulation of slight differences. But if we take this view, which is certainly most consonant with the evidence before us, we must assign to natural selection a different rôle from that which is generally ascribed to it. We cannot suppose that natural selection has played any part in the _formation_ of a mimetic likeness. The likeness turned up suddenly as a sport quite independently of natural selection. But although natural selection may have had nothing to do with its production, it may nevertheless have come into play in connection with the _conservation_ of the new form. If the new form possesses some advantage over the pre-existing one from which it sprang, is it not conceivable that natural selection will come into operation to render it the predominant form? To this question we shall try to find an answer in the next chapter. * * * * * {93} CHAPTER VIII THE CASE OF _PAPILIO POLYTES_ (_cont._)