direction opposite to that which I finally take; in addition, three
rotations on the road; a fifth rotation at the place where they are
set free. If they do not lose their bearings this time, it will not be
for lack of twisting and turning. I begin to open my screws of paper
at twenty minutes past nine. It is rather early, for which reason my
Bees, on recovering their liberty, remain for a moment undecided and
lazy; but, after a short sunbath on a stone where I place them, they
take wing. I am sitting on the ground, facing the south, with Serignan
on my left and Piolenc on my right. When the flight is not too swift
to allow me to perceive the direction taken, I see my released
captives disappear to my left. A few, but only a few, go south; two or
three go west, or to right of me. I do not speak of the north, against
which I act as a screen. All told, the great majority take the left,
that is to say, the direction of the nest. The last is released at
twenty minutes to ten. One of the fifty travellers has lost her mark
in the paper bag. I deduct her from the total, leaving forty-nine.

According to Antonia, who watches the home-coming, the earliest
arrivals appeared at twenty-five minutes to ten, say fifteen minutes
after the first was set free. By twelve o'clock mid-day, there are
eleven back; and, by four o'clock in the evening, seventeen. That ends
the census. Total: seventeen, out of forty-nine.

I resolved upon a fourth experiment, on the 14th of May. The weather
is glorious, with a light northerly breeze. I take twenty Mason-bees,
marked in pink, at eight o'clock in the morning. Rotations at the
start, after a preliminary backing in a direction opposite to that
which I intend to take; two rotations on the road; a fourth on
arriving. All those whose flight I am able to follow with my eyes turn
to my left, that is to say, towards Serignan. Yet I had taken care to