( Ivii ) 



which was also perceptible on the fingers when they had been 

 used for rubbing the scales off the wing. A brush similarly 

 used for removing the scales emitted the odour strongly. The 

 body, though crushed, was odourless. A trial made with G. 

 brassica.', ^ , gave a negative result, but on subjecting a speci- 

 men of G. rapx, $ , to similar treatment, I fancied, but could 

 not be absolutely certain, that a faint sweetish odour was 

 perceptible. 



'* In the early autumn of 1900 I failed to detect any odour 

 in a specimen of Colias editsa, $ , which had been dead about 

 twenty-four hours, but a few days later I repeated the trial 

 with the following result : — A living C. echisa, 3 , was allowed 

 to flutter about the room for a few minutes, then taken in the 

 fingers and held gently. No odour was apparent, even when 

 the so-called ' glands ' on the costa of the hind-wing (which are 

 well known to be absent in G. hyale) were uncovered. The 

 ' glandular ' patch of the right hind-wing was then carefully 

 scraped with a penknife, and the little mass of scales thus 

 detached was immediately smelt. There was a distinct 

 sweetish and almost spicy odour which I should compare to 

 heliotrope. It soon passed off, or I ceased to appreciate it. 

 The patch on the left hind-wing was similarly scraped, with 

 the same result, except that the odour seemed fainter. A trial 

 was made with another specimen that had been dead at least 

 twenty-four hours. As in the first case, I could not be sure 

 that I detected the heliotrope odour. On crushing the thorax 

 there appeared to be a slight scent somewhat like that of fresh 

 varnish, and this seemed also to attach to the * scent-scales ' 

 when scraped off the glandular patch. But it bore no resem- 

 blance to the perfume found in the living specimen. A few 

 days later I convinced myself that the scales of the upper 

 surface of the wings in G. rapse, $ , about which I had at first 

 been doubtful, really possessed a sweetish, fragrant odour, 

 somewhat like that of mignonette. Two ladies, whom I had 

 not told what to expect, also immediately detected a scent in 

 the scales from the same upper surface, Avhich they described, 

 without prompting, as 'flowery' and like 'mignonette.' 

 Neither of the ladies had the least doubt or hesitation about 

 their verdict. In G. ra-ps!., $ , similarly treated, we could find 



