150 DR. J. B. HICKS ON A NEW STRUCTURE 
probably the external layer of the antennal wall continued over it; and this is Pee 
the condition in every variety of these structures (fig. 6, a 2). 
In fig. 7 is represented the antenna of Eucera longicornis ; the whole surface is occupied 
by a number of saucer-like depressions, which are arranged so close that they touch each 
other on all sides. In the centre of these is clearly seen the tubular perforation of the 
internal wall, closed in by the very thin membrane. The diameter of the perforation 
is about 7753 inch; the whole saucer-like depression is 4559 inch. If a fracture be made 
in the antenna, the above thin membrane will become detached and float about in the 
balsam. Beneath the perforation will be found, by very careful examination, a membra- 
nous sac (fig. 7, c). These sacs touch each other all round, the whole forming a thick 
layer on the internal surface of the wall of the antenna. The number of these structures 
in this insect is about 2000 for each of the ten joints furnished with them ; the total will 
therefore be about 20,000. 
In fig. 8 is shown the antenna of another Hymenopterous insect (Andrena fulva). 
It has the same construction as the last-mentioned, except that there are hairs between 
the saucer-like depressions; and each hair as well as the depressions being surrounded 
by a ring of very minute tubercles, the whole has a pretty appearance. i 
In the Ichneumons which I have examined, I find the appearance different, thon the 
real structure is not so. The perforation remains the same, but the thin transparent 
membrane closing it in, is elevated above the general surface, and extends lengthwise 
beyond the perforation, while it is generally not so wide. The shape of this membrane 
is not unlike that of an inverted canoe, and a keel-like ridge extending longitudinally 
through its centre adds to the similitude. The perforation beneath it is oval. I have 
endeavoured to show this at fig. 9, a, 5, c. ; 
> extreme length of the external arched membrane I have found in one specimen to 
be gig inch; the longest diameter of the subjacent perforation is 44555, the shortest 
3700 inch. 
The antenna of Vespa vulgaris and of V. crabro is arranged on the same principle, but 
the organs are more numerous and not quite so regularly em They are well 
shown in the pupa state. 
In Nomada the appearance is much the same as in Apis mellifica. i , 
But the most beautiful of all the structures I have as yet met with is that presented i in à 
the antenna of Libellula depressa (Tab. XXX. fig. 10). This antenna has six joints. The 
third, fourth and fifth have on one side the following organs. 1st, a simple, nearly round 
sac, formed in the internal membrane and projecting inwards; the external opening 
being closed in by the delicate thin membrane, as noticed in the other insects (fig. 10, å 
& 6 2). 2nd, a sac as above, but opening inwardly into another convoluted membranous 
sac, which is shown at fig. 10, 41. These are more simple towards the upper joints; and 
in some of the simpler sacs there seems to be å small cavity at the back. Whether there 
be any membrane separating the outer and inner sac, it is difficult to say ; but I faney there 
is no separation of the kind. 
The nerve may be distinctly seen to pass to these structures, and as the cuticle has 
scarcely any marking, the whole is plainly visible, and forms a very interesting object. 
