188 MIMICRY IN N. AMERICAN BUTTERFLIES 



tion in tint is still to be seen in the great variation 

 of the ground-colour in artlicmis. 



Although, as Scudder rightly maintains (1. c., 

 287), L. astyanax is a very poor mimic of PJiarm. 

 philenor, it bears considerable resemblance to 

 the three Papilio mimics, especially troilus. Al- 

 though the iridescent blue or green of its upper 

 surface approaches rather more closely than the 

 Papilios to the brilliant, steely lustre of philenor, 

 it is still in this respect widely separated from 

 the primary model and near to the mimics. The 

 reddish spots of the under surface offer but a 

 rough likeness to those of any of the above- 

 named species, but there can be no doubt that 

 their emphasis is an element in the mimetic 

 resemblance. 



A careful examination of large numbers of 

 astyanax from the extreme south of the range 

 where it passes out of the area of glauctis and 

 troilus but remains within that of philenor and 

 asterius, might yield interesting results. An 

 investigation of the proportion it bears to the 

 four Papilionidae in various parts of their common 

 range would also be of deep interest. Of the 

 highest importance would be the attempt which 

 would probably be successful to breed astyanax 

 and arthemis and to ascertain whether the 

 Mendelian proportions appear in the offspring 

 of the hybrids. The pairing of astyanax and 

 archippus, although in this case failure is probable, 

 ought also to be attempted. 



