food
in this case and less in that; she would build now a large and now
a small cell; and the future of the egg would be determined by the
conditions of food and shelter.

Let us make every test, every experiment, down to the absurd: the crude
absurdity of the moment has sometimes proved to be the truth of the
morrow. Besides, the well-known story of the Hive-bee should make us
wary of rejecting paradoxical suppositions. Is it not by increasing the
size of the cell, by modifying the quality and quantity of the food,
that the population of a hive transforms a worker larva into a female or
royal larva? It is true that the sex remains the same, since the workers
are only incompletely developed females. The change is none the less
miraculous, so much so that it is almost lawful to enquire whether the
transformation may not go further, turning a male, that poor abortion,
into a sturdy female by means of a plentiful diet. Let us therefore
resort to experiment.

I have at hand some long bits of reed in the hollow of which an Osmia,
the Three-horned Osmia, has stacked her cells, bounded by earthen
partitions. I have related elsewhere (Cf. "Bramble-bees and Others":
chapters 2 to 5.--Translator's Note.) how I obtain as many of these
nests as I could wish for. When the reed is split lengthwise, the cells
come into view, together with their provisions, the egg lying on the
paste, or even the budding larva. Observations multiplied ad nauseam
have taught me where to find the males and where the females in this
apiary. The males occupy the fore-part of the reed, the end next to the
opening; the females are at the bottom, next to the knot which serves
as a natural stopper to the channel. For the rest, the quantity of the
provisions in itself points to the sex: for the females it is twice or
thrice as great as for the males.

In the scantily-provided cells, I double or treble the ration with food
taken from other cells; in the cells which are plentifully supplied, I
reduce the portion to a half or a third. Controls are left: that is to
say, some cells remain untouched, with their provisions as I found them,
both in the part which is abundantly provided and in that which is more
meagrely rationed. The two halves of the reed are then restored to their
original position and firmly bound with a few turns of wire. We shall
see, when the time comes, whether these changes increasing or decreasing
the victuals have determined the sex.

H