( lix ) 



generally over the wing in the shape of the ' battledore scales ' 

 of LycjBnids, and the ' plumules ' of many Pierines, Satyrids 

 and Nymphalids, and in other cases collected into the definite 

 areas called variously ' chalky patches,' ' sex-brands ' or 

 * glandular spots.' The proof of this is that scales removed 

 from the wing are found to possess the odour strongly, 

 provided that the special scales are present; the absence of 

 special scales means the absence of scent. In Colias edusa, 

 where the special scales are segregated, I obtained direct proof 

 that the odour present in them was absent from the rest. The 

 wings after being well rubbed retain little or no scent, nor 

 does any appear to attach to the body. I reserve the case of 

 G. rhamni, which has neither plumules nor circumscribed 

 scent-patches, for later treatment. 



'* Though the special scent-scales seem to act as reservoirs 

 and distributors of the perfume, they are not the seat of its 

 manufacture. This is doubtless carried on by certain secreting 

 cells, described and figured by Giinther, which are found in 

 connection with the sockets of the scent-scales, lying between 

 the two layers of the wing-membrane. I have been investi- 

 gating the structure of the scent-distributing apparatus in 

 Pierines for some years past, and hope soon to be able to 

 communicate my results, together with some account of the 

 literature of the subject, which, though large, is scattered and 

 by no means exhaustive. 



" I may mention, in conclusion, that I have made several 

 attempts to preserve the scent of various male butterflies, with 

 more or less success. The first method I tried was extracting 

 the wings with rectified spirit. This certainly took up the 

 odour to some extent, but unless one uses the extract in a very 

 concentrated form, the characteristic scent, which is seldom 

 strong, is liable to be overpowered by the proper smell of the 

 spirit. I then tried rubbing up the wings in a mortar with 

 starch, choosing this as a practically odourless substance. In 

 this way I got enough material diffusing the pleasant lemon- 

 verbena scent of G. napi to make a small sachet. Unfortun- 

 ately, however, the scent in this form does not seem to be 

 lasting ; at any rate, I have not as yet been very successful in 

 pi'eventing its evanescence. It soon disappears, as might be 



