numbers increased from minute to minute. Part drifted above, in clouds,
twisting round like soot in a smoke-wreath. Thousands kept sweeping just
over the tops of the willows, skimming so thickly that the sky-line was
almost blotted out for the height of from three to four feet. The quarter
from which these armies of swallows came was at first undiscoverable. They
might have been hatched, like gnats, from the river.

In time I discovered whence they came. They were literally "dropping from
the sky." The flocks were travelling at a height at which they were quite
invisible in the cloudy air, and from minute to minute they kept dropping
down into sight, and so perpendicularly to the very surface of the river
or of the eyot. One of these flocks dropped from the invisible regions to
the lawn on the river bank on which I stood. Without exaggeration I may
say that I saw them fall from the sky, for I was looking upwards, and saw
them when first visible as descending specks. The plunge was perpendicular
till within ten yards of the ground. Soon the high-flying crowds of birds
drew down, and swept for a few minutes low over the willows, from end to
end of the eyot, with a sound like the rush of water in a hydraulic pipe.
Then by a common impulse the whole mass settled down from end to end of
the island, upon the osiers. Those in the centre of the eyot were black
with swallows--like the black blight on beans.

Next morning, at 6.30 a.m., every swallow was gone. In half an hour's
watching not a bird was seen. Whether they went on during the night, or
started at dawn, I know not. Probably the latter, for Gilbert White once
found a heath covered with such a flock of migrating swallows, which did
not leave till the sun dispelled the mists.

The migration routes of birds follow river valleys, when these are
conveniently in line with the course they wish to take. There is far more