TOMES, ON A TRICHOPTEROUS LARVA. 249 
certain of the joints being greatly expanded and abruptly cut 
off at their point of junction with the succeeding joint, may 
adapt them to separate filaments of Conferva at their joints, 
so as to preserve the integrity of the cells. Whether this be 
so, fresh observation of living specimens alone can decide ; 
I have not observed the limbs so used, though I have often 
seen filaments bitten across by the strong mandibles, the legs 
being employed to gather together and hold a bundle of 
filaments which were sorted over by the mouth. ‘The legs 
are fringed with fine hairs, which doubtless greatly increase 
their efficacy as swimming organs, and are terminated by 
long hooks. 
The abdominal segments are much larger than the thoracic ; 
none of them carry appendages, nor are there any external 
branchie. 
The caudal segment is expanded and flattened, so as to 
form a quadrangular plate, the terminal corners of which are 
each armed with a hook; it is by means of these hooks that 
the larva retains its hold of the case. 
The case is of; oblong form, with rounded ends forming 
slit-shaped openings ; it is much flattened from side to side, 
and is about three times as long as it is broad. It is formed 
of closely woven, silky fibres, and perfectly translucent. 
Upon the outer surface of the sides, which are slightly con- 
vex, is a layer formed by concentrically arranged filaments of 
Conferva, entirely concealing the inner silky case, except for 
a small space at the centre of each side, where the Conferva 
is less closely coiled. 
The general direction of the filaments is parallel to the 
outline of the case; diagrammatically, the arrangement of the 
individual filaments may be represented by two letters J placed 
thus f. And herein lies the clue to the manner in which 
the Conferva is applied—the larva, which is represented in 
the accompanying figure in its habitual position, with the 
head and legs protruded from the slit-shaped opening at either 
end of the case, never voluntarily quits its shelter ; and if 
removed from it, manifests the greatest anxiety to. regain it. 
Hence the filaments of Conferve are all applied to the outer 
surface of the silky case (to which they are attached by nu- 
merous threads) by the larva working always from one end. 
Now, the larva cannot reach from either end much further 
than to the middle of the case ; accordingly, we find the ends 
of the majority of the filaments near that point; the filament 
is carried in a U-shaped curve to the end of the case, returns 
down the other side, and terminates at a point nearly opposite 
