Appendix. 1 2 7 



perature of the last place rising daily, as the ice had all 

 thawed by August 2Oth. On opening the box it was 

 found (probably owing to the cold not having been 

 sufficiently severe) that the butterflies had commenced 

 to emerge. Twenty -seven dead and crippled specimens 

 were removed, together with several dead pupae. One 

 butterfly that had just emerged was taken out and 

 placed in a box, and when its wings had fully expanded 

 it was found to be a " Telamodides of the most pro- 

 nounced type." The experimenter then states : 

 " Early in the morning I made search for the dead and 

 rejected butterflies, and recovered a few. It was not 

 possible to examine them very closely from the wet and 

 decayed condition they were in, but I was able to dis- 

 cover the broad crimson band which lies above the 

 inner angle of the hind wings, and which is usually lined 

 on its anterior side with white, and is characteristic of 

 either Walshii or Telamonides, but is not found in 

 Marcellus. And the tip only of the tail being white in 

 Walshii, while both tip and sides are white in Telamo- 

 nides, enabled me to identify the form as between these 

 two. There certainly were no Walskii, but there seemed 

 to be a single Marcellus, and excepting that all were 

 Te lamonides. ' ' 



The remaining pupae were kept in a light room where 

 3 Telamonides emerged the following day, and by Sep- 

 tember 4th 14 specimens of the same variety had 

 emerged, but no Marcellus or intermediate forms. From 

 the 4th to the 2Oth of September a few more Telamo- 

 nides appeared, but between the 4th and I5th of the 

 month 12 out of 26 butterflies that had emerged were 

 intermediate between Telamonides and Marcellus, some 

 approximating to one form and some to the other form. 

 The first pure Marcellus appeared on September 4th, 

 and was followed by one specimen on the 6th, 8th, 

 1 3th and I5th respectively. From this last date to 



