170 MIMICRY IN N. AMERICAN BUTTERFLIES 



selection based on a likeness to certain white 

 spots present in the new model (berenice) but 

 absent from the old (plexippus). But many 

 hundreds of specimens from different localities 

 scattered over the total area of distribution re- 

 quire to be examined from this point of view. 

 An even more interesting inquiry would be to 

 trace the range of the floridensis form northward 

 and determine the relationship of its limits to 

 the zone in which berenice becomes scarce and 

 disappears, and above all to ascertain whether 

 floridensis on the borders of its range interbreeds 

 with archippus and how far transitional varieties 

 occur. Interbreeding between the two forms, 

 if possible, would be of extraordinary interest. 

 It is also of importance to ascertain precisely 

 how far the one form penetrates the area of the 

 other. Scudder indeed states that floridensis 

 ranges into the Mississippi Valley and Dakota, 

 far beyond the limits of Danaida berenice. It 

 would be deeply interesting to make an exact 

 comparison between such specimens and those 

 from Florida, and also to ascertain the proportion 

 which they bear to typical arcliippus. By far 

 the most important feature in the evolution of 

 floridensis is the general darkening of the ground- 

 colour, and the material for such a transformation 

 certainly exists freely in archippus, for the shade 

 of brown varies immensely and may often be 

 seen of as dark a tint as in floridensis, but not 

 in my experience of precisely the same shade. 



