( Iv ) 



means of sexual attraction and possibly of recognition. The 

 odours belonging to this class, in all the species with which I 

 am here concerned, are confined to the male sex. They have 

 a general family likeness, recalling in most cases the perfumes of 

 flowers, or of various aromatic vegetable products such as lemon- 

 peel, spices, vanilla and chocolate. They are, I think, in nearly 

 every instance agreeable to the average human perception. 



" The scents included in the second class are probably made 

 use of for defensive purposes. They occur, as a rule, in 

 species which we have grounds for believing to be dis- 

 tasteful to some at least of the usual enemies of butterflies. 

 They are found, as would be expected on this supposition, 

 in both sexes; and tbey are for the most part disagreeable 

 or even disgusting to the human observer. It is perhaps 

 conceivable that both kinds of odours may occur in the 

 same individual ; though such instances, if they exist at all, 

 would appear to be rare. 



" The Pierine scents that have been examined belong to 

 the first class — probably serving the purpose of sexual 

 attraction and perhaps also of recognition. I subjoin a list 

 of the African species in which they have been noted. 



" Catopsilia Jlorella, Fabr. When the tufts of hair-like 

 scales present in the male are exposed by separating the 

 fore- from the hind-wing, they emit a very strong, sweet 

 scent, which Dr. Longstaff compares to that of 'tuberose' 

 or Freesia, and to my mind suggests jasmine. (A similar 

 scent was detected by Wood-Mason in the Indian C. glioma 

 [pyranthe]. See Journal of the Asiatic Society of Bengal, 

 vol.lv., Part II, 1886, No. 4, p. 371.) 



" Colias electra, Linn. The wings of a male specimen were 

 separated, and the costal patches of the hind-wing smelt, 

 with a negative result. The left costal patch was then 

 scraped with a penknife and the detached scales tested, 

 again with a negative result. The right costal patch was 

 then gently scraped and the patch itself smelt. I immedi- 

 ately perceived a distinct flowery fragrance reminding me 

 of heliotrope, with the added suggestion of chocolate. The 

 scent, which was quite pleasant, corresponded very fairly 

 with my recollection of that of Colias edusa, Fabr., obtained 



