2 26 Annals Entomological Society of America [Fol. 11, 



In attempting to trace the past history, here again we have 

 the great advantage of knowing the more ancestral patterns 

 from which the three mimics were derived : — troilus from a pala- 

 medes-like form ; asteriiis from the pattern of its male, which again 

 leads back to the typical pattern of the Machaon Group: the 

 turniis female of glaiicits from the male and non-mimetic female 

 of the same species. 



It is highly probable that the earliest steps in the direction of 

 ]\Iimicry in asterius and glaiiciis were favoured by the appearance 

 of partially melanic varieties of the female, thus effecting sudden- 

 ly that essential change which enables a butterfly with a yellow 

 ground-colour to become the mimic of one in which it is black. 

 But this transformation, immensely important as it is, supplies 

 nothing more than a tinted paper for a new picture. That the 

 melanic varieties were partial is clearly shown by the persistence 

 (in glavicus) in a subdued and inconspicuous form of certain 

 ancestral features that do not contribute to the Mimicry', but 

 above all by the retention of every element in the original pattern 

 that can be worked up into the new. By the modification of these 

 elements in form or colour, — often in both form and colour, — the 

 detailed mimetic pattern has been wrought upon the darkened 

 surface. 



Valuable confirmation of the history suggested in the last 

 paragraph is to be found in the dark form melasina (Rothsch. 

 and Jord.) found in both sexes of P. polyxenes ameriats (Kollar), 

 extending from North Peru to Colombia and Venezuela. This 

 melanic variety probably represents the darkened form of asterius 

 before the initiation of the detailed mimicry of philenor. The 

 sub-species americus does not enter the range of philenor, and 

 those ancestral elements which have been retained by its melanic 

 form have not developed into the mimetic likeness seen in the 

 more northern sub-species asterius. 



It is well known that all four species (including philenor) fly 

 together. Even in my own limited experience I have taken three 

 of them in adjacent streets on the outskirts of Chicago on the 

 same day (Aug. lo, 1897), and the fourth in the same locality a 

 little earlier (July 28). But precise knowledge of their relative 

 proportions in different parts of their range would be of highest 

 interest. Again, troilus extends to the North-West Territory 

 of Canada, probably far beyond the area in which philenor occurs 

 as a straggler; and it would be very interesting to compare niin- 



