ramount importance; and it will perhaps make the
result clearer if I quote one instance from among a multitude of similar
cases. I give the preference to this particular instance because of
the rather exceptional fertility of the laying. An Osmia marked on the
thorax is watched, day by day, from the commencement to the end of her
work. From the 1st to the 10th of May, she occupies a glass tube in
which she lodges seven females followed by a male, which ends the
series. From the 10th to the 17th of May, she colonizes a second tube,
in which she lodges first three females and then three males. From the
17th to the 25th of May, a third tube, with three females and then two
males. On the 26th of May, a fourth tube, which she abandons, probably
because of its excessive width, after laying one female in it. Lastly,
from the 26th to the 30th of May, a fifth tube, which she colonizes
with two females and three males. Total: twenty-five Osmiae, including
seventeen females and eight males. And it will not be superfluous to
observe that these unfinished series do not in any way correspond with
periods separated by intervals of rest. The laying is continuous, in so
far as the variable condition of the atmosphere allows. As soon as one
tube is full and closed, another is occupied by the Osmia without delay.

The tubes reduced to the exact length of two cells fulfilled my
expectation in the great majority of cases: the lower cell was occupied
by a female and the upper by a male. There were a few exceptions.
More discerning than I in her estimate of what was strictly necessary,
better-versed in the economy of space, the Osmia had found a way of
lodging two females where I had only seen room for one female and a
male.

This experiment speaks volumes. When confronted with tubes too small to
receive all her family, she is in the same plight as the Mason-bee
in the presence of an old nest. She thereupon acts exactly as the
Chalicodoma does. She breaks up her laying, divides it into series as
short as the room at her disposal demands; and each series begins with
females and ends with males. This breaking up, on the one hand, into
sections in all of which both sexes are represented and the division, on
the other hand, of the entire laying into just two groups, one female,
the other male, when the length of the tube permits, surely provide us
with ample evidence of the insect's power to regulate the sex of the egg
according to the exigencie