One reason, I fancy, why nests begun in such distant trees are so often deserted before completion is that a solitary nest exposes both the building birds and their prospective offspring to grave danger from hawks. No hawk will attempt to approach a rookery--the rooks would attack him _en masse_ and easily put him to flight. Chickens are safer under or near a rookery from this cause: a hawk approaching them would alarm the rooks and be beaten away. The comparative safety afforded by numbers is perhaps a reason why many species of birds are gregarious. The apparently defenceless martins and swallows in this way dwell in some amount of security. If a hawk comes near the sand quarry (or the house--in the case of swallows) they all join together and pursue him, twittering angrily, and as a matter of fact generally succeed in sending him about his business. Even those birds which do not build in close contiguity no sooner find that a hawk is near than they rise simultaneously and follow and annoy him: so much so that he will sometimes actually drop the prey he has captured. It is astonishing with what temerity small birds, emboldened by numbers--chaffinches, finches generally, sparrows, swallows, and so on--will attack a hawk. The `quar-martins' that came to the orchard wall--emigrating from the quarry, and wandering about in search of a suitable habitation--if young birds, as we have supposed them to be, would naturally not yet have had much experience, and so might think the steep wall (roughly resembling the face of a quarry) available for their purpose till they had made the experiment. I have thought, from watching the motions of birds that go in flights, that most of them have a kind of leader or chief. They do not yield anything like the same obedience or reverence to the chief as the bees do to the queen-bee, and exhibit little traces of following his motions implicitly. He is more like the president of a republic; each member is individually free, and twitters his or her mind just as he or she likes. But it seems to be reserved to one bird to give the signal for all to move. So these martins, after lingering about the wall for hours--some of them, too, leaving it and flying away only to come back again--finally started altogether. It is difficult to account for such simultaneous and combined movements, unless we suppose that it is reserved to a certain bird to give the signal.