length of the distance traversed a criterion of migration? Surely not. The distance traversed is merely a collateral consequence of the process as a whole. The annual life-history of a bird presents, as we have seen, two distinct phases--the one in which the individual dominates the situation, the other in which it is subordinated to the welfare of the community. Let us take these two phases separately and endeavour to see how they may have influenced the seasonal movements; and first let us take the more important of the two, namely that one which is directly concerned in the continuance of the race. In this phase we must consider the three factors to which allusion has already been made:--(1) the internal impulse, (2) the innate ability to return to the former breeding ground, (3) the conditions in the external environment. These three work in close relation and, as I shall endeavour to show, lead to important results. (1) If there were nothing in the inherited nature beyond an impulse to seek the breeding ground, if, that is to say, when the appropriate locality were reached, the bird took no further interest in the developing situation, the attainment of reproduction would become largely a matter of chance. A male in a congested district, having no incentive to seek fresh ground, would remain inactive until a female happened to cross its path and stimulate its sexual impulse, when its activity would take another form. Hence some districts would be over-populated, whilst others would remain unexplored. But the system of reproduction does not consist merely of a search for the breeding ground, and of the discharge of the sexual function; it is a much more complex business, yet withal more complete. Nothing is left to chance; the end is attained step by step; and each successive stage marks the appearance of some specific factor which contributes towards the success of the whole. We start with the appropriate organic condition under which, when adequate stimulation is provided, the disposition to secure a territory comes into functional activity. Within the field of this disposition we can distinguish certain specific impulses. In sequential order we have the impulse to seek the breeding ground; the appropriate situation which gives rise to an impulse to dwell in it; and the act of establishment which supplies the condition under which the impulse to drive away intruders is rendered susceptible to stimulation. Grouping these impulses, for the convenience of treatment, under one general heading, I speak of an impulse to seek isolation. It implies some kind of action with some kind of change as its correlated effect; and from it there flows a ceaseless energy directed towards a definite end which for us, who can perceive its prospective value, is isolation in an appropriate environment. The emphasis here is on "isolation," for it involves competition, and there cannot be competition without some