with the mouth of the hole. For a while it seemed as if the rabbits
declined to use the hole at all; presently, however, the loop was pushed
back, showing that one must have got his nose between it and the bank
and so made a safe passage sideways. A run that crossed the field was
then selected, and the wire erected at about the middle of it,
equidistant from either hedge. Near the entrance of the buries the
rabbits moved slowly, sniffing their way along and pausing every yard or
so. But they often increased their speed farther away, and sometimes
raced from one mound to the other. When going at that rate it appeared
natural to conclude that they would be less careful to pick and choose
their road.

The theory proved so far correct that next day the upright was down, but
the wire had snapped and the rabbit was gone. The character of the
fracture clearly indicated how it had happened: the rabbit, so soon as
he found his head in the noose, had rolled and tumbled till the wire,
already twisted tight, parted. Too much twisting, therefore, weakened
instead of strengthening. Next a single wire, somewhat thicker, was
used, and set up nearly in the same place; but it broke again.

Finally, two strands of medium size, placed side by side, but only
twisted once--that is, just enough to keep them together--were employed.
The lesser loop--the slip-knot, as it might be called--was at the same
time eased in order to run quicker and take a closer grip. Experiments
with the hand proved that this style of wire would bear a great strain,
and immediately answered to a sudden jerk. The running noose slipped the
more easily because the wires were smooth; when twisted the strands
checked the noose, the friction causing a slight sound. The wire itself
seemed nearly perfect; but still no rabbit was caught.