the rearing of
Silkworms and, in autumn, for the drying of figs. At the end of April
and during May, which is the time when the Osmiae work, the canisses
are indoors, in the Silkworm nurseries, where the Bee cannot take
possession of them; in autumn, they are outside, exposing their layers
of figs and peeled peaches to the sun; but by that time the Osmiae have
long disappeared. If, however, during the spring, an old, disused
hurdle is left out of doors, in a horizontal position, the Three-horned
Osmia often takes possession of it and makes use of the two ends, where
the reeds lie truncated and open.

There are other quarters that suit the Three-horned Osmia, who is not
particular, it seems to me, and will make shift with any hiding-place,
so long as it have the requisite conditions of diameter, solidity,
sanitation and kindly darkness. The most original dwellings that I know
her to occupy are disused Snail-shells, especially the house of the
Common Snail (Helix aspersa). Let us go to the slope of the hills thick
with olive-trees and inspect the little supporting-walls which are
built of dry stones and face the south. In the crevices of this
insecure masonry we shall reap a harvest of old Snail-shells, plugged
with earth right up to the orifice. The family of the Three-horned
Osmia is settled in the spiral of those shells, which is subdivided
into chambers by mud partitions.

The Three-pronged Osmia (O. Tridentata, Duf. and Per.) alone creates a
home of her own, digging herself a channel with her mandibles in dry
bramble and sometimes in danewort.

The Osmia loves mystery. She wants a dark retreat, hidden from the eye.
I would like, nevertheless, to watch her in the privacy of her home and
to witness her work with the same facility as if she were nest-building
in the open air. Perhaps there are some interesting characteristics to
be picked up in the depths of her retreats. It remains to be seen
whether my wish can be realized.

When studying the insect's mental capacity, especially its very
retentive memory for places, I was led to ask myself whether it would
not be possible to make a suitably-chosen Bee build in any place that I
wished, even in my study. And I wanted, for an experiment of this sort,
not an individual but a numerous colony. My preference lent towards the
Three-horned Osmia, who is very plentiful in my neighbourhood, where,
together with Latreille's Osmia, she frequents in particular the
monstrous n