the twofold exit, others, such as the
Solenius and the Leaf-cutter, behave like a flock of sheep and follow
the first that goes out. The entomological world is not all of a piece;
its gifts are very various: what one is capable of doing another cannot
do; and penetrating indeed would be the eyes that saw the causes of
these differences. Be this as it may, increased research will certainly
show us a larger number of species qualified to use the double outlet.
For the moment, we know three; and that is enough for our purpose.

I will add that, when the horizontal tube has one of its ends closed,
the whole string of Osmiae makes for the open end, turning round to do
so, if need be.

Now that the facts are set forth, let us, if possible, trace the cause.
In a horizontal tube, gravity no longer acts to determine the direction
which the insect will take. Is it to attack the partition on the right
or that on the left? How shall it decide? The more I look into the
matter, the more do my suspicions fall upon the atmospheric influence
which is felt through the two open ends. Of what does this influence
consist? Is it an effect of pressure, of hygrometry, of electrical
conditions, of properties that escape our coarser physical attunement?
He were a bold man who should undertake to decide. Are not we ourselves,
when the weather is about to alter, subject to subtle impressions,
to sensations which we are unable to explain? And yet this vague
sensitiveness to atmospheric changes would not be of much help to us in
circumstances similar to those of my anchorites. Imagine ourselves in
the darkness and the silence of a prison-cell, preceded and followed
by other similar cells. We possess implements wherewith to pierce the
walls; but where are we to strike to reach the final outlet and to reach
it with the least delay? Atmospheric influence would certainly never
guide us.

And yet it guides the insect. Feeble though it be, through the
multiplicity of partitions, it is exercised on one side more than on the
other, because the obstacles are fewer; and the insect, sensible to the
difference between those two uncertainties, unhesitatingly attacks the
partition which is nearer to the open air. Thus is decided the division
of the column into two converse sections, which accomplish the total
liberation with the least aggregate of work. In short, the Osmia and her
rivals 'feel' the free space. This is yet one more sensory faculty which
evolut