ANCESTEAL TRACES IN THE MIMIC 167 



traces of the white band may commonly be seen 

 along the inner edge of the persistent black 

 border. So far as my experience goes these 

 traces are only to be found on the upper surface 

 in the form hulsti (Edw.)- The modification of 

 the same marking in the fore wing is more in- 

 teresting. Here towards the costal margin the 

 black outer border is much expanded, invading 

 the white band and cutting off from two to four 

 white spots from its outer part. While the rest 

 of the band disappears except on the costa itself, 

 these black-surrounded white spots now repre- 

 sent the sub-apical pale-spotted black bar of the 

 model. The new marking is larger and more 

 conspicuous on the under surface, corresponding 

 with the strong development of white on this 

 surface of the model. The costal margin of the 

 fore wing of the latter is streaked with long 

 narrow white markings. In correspondence with 

 this we find, commonly on the under surface, 

 more rarely on the upper, that the extreme 



fact that entirely stripeless individuals were invariably males is 

 contrary to the rule that mimetic resemblance tends to develop 

 more rapidly and fully in the other sex. But in this species I have 

 observed another point in which the female tends to be more 

 ancestral than the male, viz. the more frequent and complete 

 development of the white spot in the cell of the fore-wing upper 

 surface (a common feature of Limenitis, although now generally 

 absent from L. arthemis). 



Mr. Cook's observations show that a single marking— and one 

 so simple that we might have expected it to act as a 'unit 

 character ', so small a fraction of the pattern that we could hardly 

 speak of its sudden disappearance as ' discontinuous ' evolution 

 — that even this behaves differently on the two surfaces of the 

 wing, while the individuals from which it has disappeared are 

 immensely outnumbered by those in which it is transitional. 



