practice of the aquatic rat to forsake the neighbourhood of the
water in the colder months?

Though I delight very little in analogous reasoning, knowing how
fallacious it is with respect to natural history; yet, in the following
instance, I cannot help being inclined to think it may conduce
towards the explanation of a difficulty that I have mentioned
before, with respect to the invariable early retreat of the hirundo
apus, or swift, so many weeks before its congeners; and that not
only with us, but also in Andalusia, where they also begin to retire
about the beginning of August.

The great large bat* (which by the by is at present a nondescript in
England, and what I have never been able yet to procure) retires
and migrates very early in the summer: it also ranges very high for
its food, feeding in a different region of the air; and that is the
reason I never could procure one. Now this is exactly the case with
the swifts; for they take their food in a more exalted region than the
other species, and are very seldom seen hawking for flies near the
ground, or over the surface of the water. From hence I would
conclude that these hirundines, and the larger bats, are supported
by some sorts of high-flying gnats, scarabs, or phalaenae, that are
of short continuance; and that the short stay of these strangers is
regulated by the defect of their food.
(* The little bat appears almost every month in the year; but I have
never seen the large ones till the end of April, nor after July. They
are most common in June, but never in any plenty; are a rare
species with us.)

By my journal it appears that curlews clamoured on to October the