Now a male may leave its territory for three reasons--to pursue an intruder, to join the flock on neutral ground, or to find the necessary means of subsistence on other feeding grounds. On each of these occasions it hears the song of, and is in close contact with, other males; and if the relationship of which we are speaking be really exclusive of cross-correlation, its instinct ought to respond with the customary freedom. But what happens? A male pursues its rival, betraying much emotion and singing extravagantly, until the boundary is passed, when emotion subsides and it is silent; or, it flies to the flock on neutral ground, and, although surrounded by the very males that a short time previously evoked response, is there unresponsive; or again, it goes in search of food and collects with other males bent on a similar errand, and in presence of what we know would be an exciting influence under other circumstances, it nevertheless remains silent. Hence the relationship between the song and a male rival seems, as in the case of the headquarters, to depend in the first instance upon the occupation of a territory. So that the relationship between the song and the territory as a whole is clearly of a different order from that which obtains between the song and the headquarters, or the song and a male rival; for the first, as far as can be judged by observation, is exclusive of, whilst the second and the third involve, cross-correlation. How are these facts to be explained? We have already seen that it belongs to the nature of the male during the season of reproduction to establish itself in a definite place, and this action is just as much a part of its hereditary nature as the building of the nest is of that of the female, and it is just as necessary for the successful attainment of reproduction. What exactly the stimulus is to this mode of behaviour we do not know; we can go no further back than the internal organic changes which are known to occur and which we assume, not without some reason, are responsible for its initiation. Granting, then, that there is this congenital disposition, what relation does it bear to the song? Without a doubt the song is likewise founded upon a congenital basis; it is truly instinctive, and as such requires appropriate stimulation; furthermore the male sings only when in occupation of its territory. Having regard to these two facts we might say that the territory is the stimulus to the song. But this can scarcely be a true interpretation, for inasmuch as the stimulus would be relatively constant, a relatively constant response ought to follow, and even a slight acquaintance with the daily round of behaviour will furnish plenty of evidence to the contrary, seeing that the song, though persistent, is never continuous--in fact there are long periods of silence during the daytime, and only in the morning and the evening does the male become really vociferous. What then is the