of its task, has steeped
thoroughly in a sort of varnish prepared not by the silk-glands but by
the stomach. The cocoons of the Sphex have already shown us a similar
varnish. This product of the chylific ventricle is chestnut-brown. It is
this which, saturating the thickness of the tissue, effaces the bright
red of the beginning and replaces it by a brown tint. It is this again
which, disgorged more profusely at the lower end of the cocoon, glues
the two wrappers together at that point.

The perfect insect is hatched at the beginning of July. The emergence
takes place without any violent effraction, without any ragged rents. A
clean, circular fissure appears at some distance from the top; and the
cephalic end is detached all of a piece, as a loose lid might be. It is
as though the recluse had only to raise a cover by butting it with her
head, so exact is the line of division, at least as regards the inner
envelope, the stronger and more important of the two. As for the outer
wrapper, its lack of resistance enables it to yield without difficulty
when the other gives way.

I cannot quite make out by what knack the Wasp contrives to detach the
cap of the inner shell with such accuracy. Is it the art practised by
the tailor when cutting his stuff, with mandibles taking the place of
scissors? I hardly venture to admit as much: the tissue is so tough and
the circle of division so precise. The mandibles are not sharp enough to
cut without leaving a ragged edge; and then what geometrical certainty
they would need for an operation so perfect that it might well have been
performed with the compasses!

I suspect therefore that the Scolia first fashions the outer sac in
accordance with the usual method, that is, by distributing the silk
uniformly, without any special preparation of one part of the wall more
than of another, and that it afterwards changes its method of weaving in
order to attend to the main work, the inner shell. In this it
apparently imitates the Bembex (Cf. "The Hunting Wasps": chapters 14 to
16.--Translator's Note.), which weaves a sort of eel-trap, whose ample
mesh allows it to gather grains of sand outside and encrust them one
by one in the silky network, and completes the performance with a cap
fitting the entrance to the trap. This provides a circular line of least
resistance, along which the casket breaks open afterwards. If the Scolia
really works in the same manner, everything is explained: the eel-trap,