and then stops short. His tail, which is broader than it looks, is bent slightly downwards; his wings beat the air, at the first glance, just as if he was progressing. Sometimes he seems to oscillate to one side, sometimes to the other; but these side movements do not amount to any appreciable change of position. If there be little or no wind (note this) he remains beating the air, to the eye at least perfectly stationary, perhaps as much as half a minute or more. He then seems to slip forward about half a yard, as if a pent-up force was released, but immediately recovers himself and hovers again. This alternate hovering and slipping forward may be repeated two or three times: it seems to depend on the bird's judgment as to the chance of prey. If he does not think a mouse is to be had, at the first slip he allows himself to proceed. If the spot be likely, or (what is still more tempting) if it is near a place where he has taken prey previously, he will slip and bring up several times. Now and then he will even fetch a half-circle when his balance or impetus (or both) is quite exhausted, and so return to the same spot and recommence. But this is not often, as a rule, after two or three slips he proceeds on his voyage. He will repeat the same round day after day, if undisturbed, and, if the place be at all infested with mice, he will come to it three or four times a day. There is, therefore, every chance of watching him, if you have once found his route. Should he spy a mouse, down he comes, quick but steady, and very nearly straight upon it. But kestrels do not always descend upon prey actually in view. Unless I am much mistaken, they now and then descend in a likely
