s, the female, twice or thrice
as bulky, will consume three to six at least. If the male Tachytes
requires three Mantes, his consort's meal will demand a batch of
something like ten. With her comparative corpulence, the female Osmia
will need a heap of paste twice or thrice as great as that of her
brother, the male. All this is obvious; the animal cannot make much out
of little.

Despite this evidence, I was anxious to enquire whether the reality
corresponded with the previsions of the most elementary logic. Instances
are not unknown in which the most sagacious deductions have been found
to disagree with the facts. During the last few years, therefore, I have
profited by my winter leisure to collect, from spots noted as favourable
during the working-season, a few handfuls of cocoons of various
Digger-wasps, notably of the Bee-eating Philanthus, who has just
furnished us with an inventory of provisions. Surrounding these cocoons
and thrust against the wall of the cell were the remnants of the
victuals--wings, corselets, heads, wing-cases--a count of which enabled
me to determine how many head of game had been provided for the larva,
now enclosed in its silken abode. I thus obtained the correct list
of provisions for each of the huntress' cocoons. On the other hand, I
estimated the quantities of honey, or rather I gauged the receptacles,
the cells, whose capacity is proportionate to the mass of the provisions
stored. After making these preparations, registering the cells, cocoons
and rations and putting all my figures in order, I had only to wait for
the hatching-season to determine the sex.

Well, I found that logic and experiment were in perfect agreement. The
Philanthus-cocoons with two Bees gave me males, always males; those with
a larger ration gave me females. From the Tachytes-cocoons with double
or treble that ration I obtained females. When fed upon four or five
Nut-weevils, the Sand Cerceris was a male; when fed upon eight or ten, a
female. In short, abundant provisions and spacious cells yield females;
scanty provisions and narrow cells yield males. This is a law upon which
I may henceforth rely.

At the stage which we have now reached a question arises, a question of
major importance, touching the most nebulous aspect of embryogeny.
How is it that the larva of the Philanthus, to take a particular case,
receives three to five Bees from its mother when it is to become a
female and not more than two when it is