revision of the genus Argynnis. 569 



of insects does not make his descriptions sufficiently 

 comparative and clear to be followed by others, he must 

 not be surprised if others refuse to accept them. Larval 

 characters alone, which are liable to vary like those of 

 the perfect insect, and which cannot be easily compared 

 by others, are not in my opinion sufficient. With 

 regard to A. nausicaa, however, I see a point not alluded 

 to by Mr. Edwards, which may be sufficient to separate 

 it, namely, the much less abundant and shorter tuft of 

 hairs on the subcostal vein of the hind wing in the 

 males. This tuft is prominent in all males of aphrodite, 

 alcestis, and halcyone which I have examined; in 

 nausicaa it is much less conspicuous, and, taken in con- 

 junction with the isolated habitat and deeper colour, is 

 probably enough to distinguish it. 



A. atlantis is another species which I find it uncom- 

 monlj'- difficult to decide about, not so much when the 

 eastern form alone is before me, but when the numerous 

 western species or forms have to be considered. Mr. 

 Edwards has got over the difficulty by naming them all 

 separately, and Mr, Scudder, though he was not perhaps 

 obliged to mention them in the ' Butterflies of New 

 England,' says nothing as to their very near relationship. 

 He remarks as follows : — " There is no need of con- 

 founding this si^ecies with either of the preceding 

 [aphrodite and cybele] : it is smaller than they, duller in 

 tint above, has a blackish border to all the wings in 

 both sexes, and more continuous mesial band on the 

 upper surface of the hind wings ; the darker colours of 

 the under surface of hind wings are deeper in hue than 

 in either of them, while the buff belt is wider than that 

 of aphrodite and narrower than that of cybele ; the buff 

 scales on the basal half of the wing also assume more 

 importance than in the other species ; finally the 

 costal border of the fore wings does not appear to be 

 quite so much arched. It is possible, perhaps even 

 probable, that this species is the true Papilio aphrodite 

 of Fabricius, but as it is quite impossible to be certain 

 of it, the names ought to stand as given by Mr. W. H. 

 Edwards, who first clearly distinguished the si^ecies in 

 this difficult group. The species w^ere still confounded 

 in the British Museum, after the publication of Butler's 

 Fabrician butterflies, as I myself saw, and notwith- 

 standing Butler's remarks on p. 108 of that work." Its 



2 R 2 



