efore
long one of the Magpies flew at him, and put him to flight. The Crow
defended itself, but not, it would seem, very successfully, and in the
last attack upon him, made, with great spirit, in the air, a large black
feather floated to the ground, which I made no doubt was his. Yet this
did not drive him from the trees, and it was only on my approaching
nearer that he finally left them. Thus we see that both species look
upon the approach of the other to within a moderate distance of their
nest as an intrusion."

That the Rook suffers persecution from the Carrion-Crow is a
well-established fact, and there is reason to believe that it has
another dangerous enemy in the Hooded Crow. According to the late
Mr. Ussher, Choughs will attack both Hooded Crows and Ravens. "I once
saw," he says, "two Choughs energetically attacking a pair of Ravens;
they shot up into the air and darted down on the latter, whose heavy
flight made them helpless against their agile tormentors."

Birds of prey are often hostile to one another. The Merlin is
exceptionally pugnacious, and its boldness in attacking intruders is
well known. When, for example, a Kestrel approaches its territory, it
leaves the tree, bush, or rock upon which it was resting, utters its
characteristic cry, and soars rapidly upwards; then, rising to a
considerable height, it swoops down upon the Kestrel, and by
alternately stooping at and chasing its opponent, drives it away from
the immediate neighbourhood.

What we have, then, to consider is, Do these battles between different
species contribute towards the attainment of the end for which the whole
territorial system has been evolved?

Let us take the individual and see whether we can establish any relation
between the hostility it displays towards members of other species and
its general disposition to secure a territory. We must remember that a
male can have no knowledge of the prospective value of its behaviour,
nor is it likely that it has any ulterior purpose in ejecting other
males, beyond the pleasure it derives from satisfying its impulse to do
so. The proximate end of its behaviour is to attack, nothing more, and
this, of course, it can only do just in so far as the intruder evokes
the appropriate instinct.

Now the arguments we shall employ will, on the whole, be similar to
those which we made use of in the second chapter, wherein we attempted
to ascertain the conditions under which a male becomes intoleran