1911] on the Scents of Butterflies. 79 



of scent ; and to get some insight into the mode of production of 

 the latter, it is evident that we must pursue our researches into the 

 structure of the wing itself. 



It was noticed by Weismann over thirty years ago, and more 

 recently by Giinther, that in the hypodermis, as it is called, or 

 cellular layer immediately underlying the homogeneous surface- 

 membrane of the wing, there occur certain cells which appear to 

 be specialised for the production of a secretion. These cells were 

 described and figured by Giinther under the name of " Driisenzelle," 

 or " gland-cells." In this diagram, which is copied from one of 

 Giintiier's figures, we see two of these " gland-cells " in direct con- 

 nection with the sockets in the wing-membrane into which fit the 

 footstalks of two scales partly seen in section. These scales are not 

 plume-scales, for they possess no disc ; but if secreting cells are found, 

 as here, in connection with scales of the ordinary kind, there seems 

 to be no reason why we should not also find them in relation with 

 the plume-scales, supplying in that position the living and working 

 protoplasmic element by means of which the scent-bearing secretion is 

 elaborated. The clear spaces in these cells of Gtinther's figure are 

 rather suggestive of the oil or fat vacuoles seen especially in growing 

 cells of adipose tissue ; and it may be conjectured that the scent- 

 bearing secretion is of the nature of a volatile oil. In the case of the 

 ordinary scales the secretion may still be oily. Probably most of us 

 know how difficult it is to wet a butterfly's Aving with water. This 

 is no doubt partly due to the mechanical conditions involved in the 

 coating of minute scales, but it is possible that, as in the case of the 

 plumage of aquatic birds, some additional power of resistance to wet is 

 afforded by the presence of an oily secretion, which may be conveyed 

 to the surface by the scales of ordinary character. It is also possible, 

 as has been suggested by Weismann, that the secretion formed in con- 

 nection with the ordinary scales may bear an odour, though of a 

 different nature from that of the plume-scales, and, at least in many 

 cases, imperceptible by our senses. All this is a matter of more or 

 less probable conjecture, and it is very clear that there is a good 

 deal more work to be done before we can be sure that we know all 

 about the various functions of the scales and their associated struc- 

 tures. 



Before we go on to the next part of our subject I should like to 

 call your attention to some figures that will be thrown upon the screen 

 of various forms of plume scales. In these figures the chitinous 

 sculpturing of the scale will be seen. It differs in character from 

 species to species, but in all there is more or less visible a longitudinal 

 striation of the lamina, which we have seen reason for interpreting 

 as an indication of channels along which pass the odoriferous secre- 

 tions or exhalations from the gland-cells buried in the substance of 

 the wing to the fimbriiB and so into the open. 



The question will no doubt have occurred, Are these plume-scales 



