In explanation of the present case Dr. Weismann wrote to Mr. Edwards:--“_Marcia_ is the old primary form of the species, in the glacial period the only one. _Tharos_ is the secondary form, having arisen in the course of many generations through the gradually working influence of summer heat. In your experiments cold has caused the summer generation to revert to the primary form. The reversion which occurred was complete in the females, but not in all the males. This proves, as it appears to me, that the males are changed or affected more strongly by the heat of summer than the females. The secondary form has a stronger constitution in the males than in the females. As I read your letter, it at once occurred to me whether in the spring there would not appear some males which were not pure _Marcia_, but were of the summer form, or nearly resembling it; and when I reached the conclusion of the letter I found that you especially mentioned that this was so! And I was reminded that the same thing is observable in _A. Levana_, though in a less striking degree. If we treated the summer brood of _Levana_ with ice, many more females than males would revert to the winter form. This sex is more conservative than the male--slower to change.”
The extreme variability of _P. Tharos_ was alluded to more than a century ago by Drury, who stated:--“In short, Nature forms such a variety of this species that it is difficult to set bounds, or to know all that belongs to it.” Of the different named varieties, according to Mr. Edwards, “A appears to be an offset of B, in the direction most remote from the summer form, just as in _Papilio Ajax_, the var. _Walshii_ is on the further side of _Telamonides_, remote from the summer form _Marcellus_.” Var. C leads from B through D directly to the summer form, whilst A is more unlike this last variety than are several distinct species of the genus, and would not be suspected to possess any close relationship were it not for the intermediate forms. The var. B is regarded as nearest to the primitive type for the following reasons:--In the first place it is the commonest form, predominating over all the other varieties in W. Virginia, and moreover appears constantly in the butterflies from pupæ submitted to refrigeration. Its distinctive peculiarity of colour occurs in the allied species _P. Phaon_ (Gulf States) and _P. Vesta_ (Texas), both of which are seasonally dimorphic, and both apparently restricted in their winter broods to the form corresponding to B of _Tharos_. In their summer generation both these species closely resemble the summer form of _Tharos_, and it is remarkable that these two species, which are the only ones (with the exception of the doubtful _Batesii_) closely allied to _Tharos_, should alone be seasonally dimorphic out of the large number of species in the genus.