cade of mud, of the leaf-putty of Latreille's Osmia, of the
Megachiles' barrier of leaves cut into disks. All these free tenants are
careful to shut tight the door of the dwelling, of which they have often
utilized only a portion. To watch the building of this barricade, which
is almost external work, demands but a little patience in waiting for
the favourable moment.

The Anthidium arrives at last, carrying the bale of cotton for the
plugging. With her fore-legs she tears it apart and spreads it out; with
her mandibles, which go in closed and come out open, she loosens the
hard lumps of flock; with her forehead she presses each new layer upon
the one below. And that is all. The insect flies off, returns the richer
by another bale and repeats the performance until the cotton barrier
reaches the level of the opening. We have here, remember, a rough task,
in no way to be compared with the delicate manufacturer of the bags;
nevertheless, it may perhaps tell us something of the general procedure
of the finer work. The legs do the carding, the mandibles the dividing,
the forehead the pressing; and the play of these implements produces the
wonderful cushioned wallet. That is the mechanism in the lump; but what
of the artistry?

Let us leave the unknown for facts within the scope of observation. I
will question the Diadem Anthidium in particular, a frequent inmate
of my reeds. I open a reed-stump about two decimetres long by twelve
millimetres in diameter. (About seven and three-quarter inches by
half an inch.--Translator's Note.) The end is occupied by a column of
cotton-wool comprising ten cells, without any demarcation between
them on the outside, so that their whole forms a continuous cylinder.
Moreover, thanks to a close felting, the different compartments are
soldered together, so much so that, when pulled by the end, the cotton
edifice does not break into sections, but comes out all in one piece.
One would take it for a single cylinder, whereas in reality the work
is composed of a series of chambers, each of which has been constructed
separately, independently of the one before, except perhaps at the base.

For this reason, short of ripping up the soft dwelling, still full of
honey, it is impossible to ascertain the number of storeys; we must
wait until the cocoons are woven. Then our fingers can tell the cells by
counting the knots that resist pressure under the cover of wadding. This
general structure is easily expl