Scent Apparatus in the male of Amauwris niavius. 403 
dark and pale hairs, beyond that of colour, is that the 
former appear to be somewhat stiffer. Both have the 
distal ends rounded and often somewhat clavate. Their 
surface is longitudinally ribbed, and in section they have 
the appearance shown in fig. 7. Seen by transmitted hght 
the hairs have reticular markings as shown in fig. 6. 
We now come to the most peculiar structures in this 
complicated organ. If a brush be examined in section 
the spaces between the hairs are seen to be packed with 
very minute particles which have a stellate appearance 
when occurring singly, but which may also appear as 
elongate bodies covered with projections. If a brush be 
removed from a dried specimen, teased out on a slide and 
examined dry, with a very high power, the whole field is 
strewn with these objects, which then present the appear- 
ance shown in fig. 9. In sections mounted in balsam they 
appear to be smoother and are much more difficult to see. 
These particles arise from a special layer of cells forming 
the middle portion of the lining of the brush bag. Part 
of this layer is shown in fig. 11. The cells are very similar 
to those which produce the hairs of the brush except that 
they are much smaller. They terminate in chitinous sockets 
from each of which protrudes a delicate thread-like growth, 
the free end of which appears to be obtusely forked. In 
balsam-mounted specimens it is very difficult to observe 
any segmentation in these filaments, though from the 
appearance of dry preparations [ am convinced that they 
are segmented. It seems probable that the cells are in 
fact modified hair-producing cells and that the delicate 
hairs to which they give rise have a stellate section, and 
further that they divide transversely into a multitude of 
minute particles. The stellate appearance may also be 
due in part to the splitting of the hair at the broken edge. 
We may now compare the whole scent apparatus with 
the corresponding structures in D. erippus and D. gilippus, 
as described by Miiller in the paper already referred to. 
Highly magnified sections of the brushes and wing patches 
are not given, but the author figures and describes views 
of the inner membrane of the wing pockets of both these 
insects. The figures would seem to show that there are 
cup-like projections much as in A. niavius, and scales 
arising from sockets placed between them. 
We may here quote Dr. Miiller’s description. “ In 
D. erippus it [the patch] exhibits small circles of about 
