some forty of the superb cells utilized by each of the
first two. The great majority are divided into two storeys by means of
a transversal partition. The lower storey includes the larger portion
of the Anthophora's cell; the upper storey includes the rest of the
cell and a little of the bottle-neck that surmounts it. The two-roomed
dwelling is closed, in the passage, by a shapeless, bulky mass of dried
mud. What a clumsy artist the Osmia is, compared with the Anthophora!
Against the exquisite work of the Anthophora, partition and plug strike
a note as hideously incongruous as a lump of dirt on polished marble.

The two apartments thus obtained are of a very unequal capacity, which
at once strikes the observer. I measured them with my five-millimetre
tube. On an average, the bottom one is represented by a column of sand
50 millimetres deep (1.95 inches.--Translator's Note.) and the top one
by a column of 15 millimetres (.585 inch.--Translator's Note.). The
holding-capacity of the one is therefore about three times as large as
that of the other. The cocoons enclosed present the same disparity. The
bottom one is big, the top one small. Lastly, the lower one belongs to a
female Osmia and the upper to a male Osmia.

Occasionally the length of the bottle-neck allows of a fresh arrangement
and the cavity is divided into three storeys. The bottom one, which is
always the most spacious, contains a female; the two above, both smaller
than the first and one smaller than the other, contain males.

Let us keep to the first case, which is always the most frequent. The
Osmia is in the presence of one of these pear-shaped hollows. It is a
find that must be employed to the best advantage: a prize of this sort
is rare and falls only to fortune's favourites. To lodge two females
in it at once is impossible; there is not sufficient room. To lodge two
males in it would be undue generosity to a sex that is entitled to but
the smallest consideration. Besides, the two sexes must be represented
in almost equal numbers. The Osmia decides upon one female, whose
portion shall be the better room, the lower one, which is larger,
better-protected and more nicely polished, and one male, whose portion
shall be the upper storey, a cramped attic, uneven and rugged in the
part which encroaches on the bottle-neck. This decision is proved by
numerous undeniable facts. Both Osmiae therefore can choose the sex
of the egg about to be laid, seeing that they