404 Mr. H. Eltringham on the 
0:01 mm. diameter, a httle more transparent than the rest 
of the membrane. From the centre of each rises a straight 
hair, about 0°06 mm. long. The circles are placed in 
regular lines, about 0°03 to 0°05 mm. apart. Alternating 
with these circles are opaque grey scales, distinguished 
from the ordinary ones by their smaller size and by their 
shape. 
In D. gilippus the circles are much closer together— 
so much so that in places they almost touch; although 
more transparent than the rest of the membrane, they are 
less so than those of D. erippus. The hairs are wanting, 
but one sees in the centre of each circle a small spot, the 
last vestige which proves their former existence. 
Miiller ‘regarded these “small circles” as scale sockets. 
The evidence for this view is very strong, since in D, 
erippus hairs still arise therefrom. In the Pierine butter- 
flies, as we know from the researches of Dr. F. A. Dixey, 
the special plume scales themselves act as distributors of 
the scent which passes into the substance of the scale 
through the footstalk. We may therefore with some 
degree of confidence regard the scent cups in the wings 
of A. niavius as highly ‘specialised scale sockets, the pore 
in the centre of the cover being the vestige of the insertion 
of the scale stalk and now functioning as an exit for the 
secretion of the gland cells. 
We may suppose that the insect brushes out the secretion, 
the stiffer hairs probably assisting in lifting the covering 
scales, which are specially adapted to withstand this 
treatment. It should be noted that these hairs are on the 
outer side of the brush, and would thus naturally come 
first into contact with the scent patch. The completely 
expanded brush then diffuses the scent. The ribbed 
structure of the hairs probably serves not only to increase 
their surface, but also to aid in the retention of the secretion. 
It still remains to explain the purpose of the stellate 
fragments which occur in such profusion in the brush bags, 
and for the production of which so many special cells exist. 
The most probable explanation appears to be that they 
float off the brushes during flight and carry the scent, thus 
diffusing it to a greater area around the insect than could 
be accomplished ‘by the brushes alone. 
The occurrence of hairs which split transversely into 
fragments is not unknown in connection with analogous 
organs in other Lepidoptera. 
