to need it; but until that moment it must remain more or
less well-filled, although becoming limper daily. It is important,
therefore, that the caterpillar's existence be not endangered by wounds
which, even though very tiny, would stop the working of the
blood-fountains. With this intent, the drainers of the bottle are, in a
manner of speaking, muzzled; they have by way of a mouth a pore that
sucks without bruising.

The dying caterpillar continues to lay the silk of his carpet with a
slow oscillation of the head. The moment now comes for the parasites to
emerge. This happens in June and generally at nightfall. A breach is
made on the ventral surface or else in the sides, never on the back:
one breach only, contrived at a point of minor resistance, at the
junction of two segments; for it is bound to be a toilsome business, in
the absence of a set of filing-tools. Perhaps the grubs take one
another's places at the point attacked and come by turns to work at it
with a kiss.

In one short spell, the whole tribe issues through this single opening
and is soon wriggling about, perched on the surface of the caterpillar.
The lens cannot perceive the hole, which closes on the instant. There
is not even a haemorrhage: the bottle has been drained too thoroughly.
You must press it between your fingers to squeeze out a few drops of
moisture and thus discover the place of exit.

Around the caterpillar, who is not always quite dead and who sometimes
even goes on weaving his carpet a moment longer, the vermin at once
begin to work at their cocoons. The straw-coloured thread, drawn from
the silk-glands by a backward jerk of the head, is first fixed to the
white network of the caterpillar and then produces adjacent warp-beams,
so that, by mutual entanglements, the individual works are welded
together and form an agglomeration in which each of the grubs has its
own cabin. For the moment, what is woven is not the real cocoon, but a
general scaffolding which will facilitate the construction of the
separate shells. All these frames rest upon those adjoining and, mixing
up their threads, become a common edifice wherein each grub contrives a
shelter for itself. Here at last the real cocoon is spun, a pretty
little piece of closely-woven work.

In my rearing-jars I obtain as many groups of these tiny shells as my
future experiments can wish for. Three-fourths of the caterpillars have
supplied me with them, so ruthless has been the toll of