We have assumed, in the imaginary case which we have just taken, that the conditions in the external world are such as enable the birds to endure throughout the year--in short, that there are no complications regarding the supply of food. But we must bear in mind that so long as conditions are favourable during the period of reproduction, which is of short duration, the breeding range can continue to expand, and that therefore, in the course of centuries, regions will come to be occupied wherein, owing to alternations of climate or physical changes in the surface of the earth, food will be impossible, or at any rate difficult to obtain at certain seasons. Hence there will come a time when the area of association ceases to follow in the wake of the expansion, and the breeding area begins to diverge from the subsistence area. How, then, is the gulf between these two areas to be bridged? We can of course say that those individuals which, in virtue of some slight variation of hereditary tendency, return to regions where food is plentiful will survive; whilst others, less well endowed, will perish. We can state the position in some such general terms, and doubtless there would be truth in the statement, but it does not carry us far; we wish to know more of the nature of the tendency, and of the manner in which it has evolved. Well now, in this new situation which arises, two things are apparent--that the struggle for existence becomes a struggle for the means of subsistence, and that anything in the inherited constitution of the bird which can be organised to subserve the biological end in view becomes of selection value. So long as food can always be procured in the new areas of association, the individuals that behave in accordance with ancestral routine gain thereby no particular advantage; but directly the breeding range extends into regions where the supply fluctuates, traditional experience becomes a factor in survival, and those individuals that come under its influence will, on the average, be more likely to endure and so to procreate their kind and maintain the tradition. Let it once be granted that there is an innate capacity to retain in later phases of routine the experience gained in earlier phases, and it is difficult to see how traditional guidance can be refused recognition as a factor in the developing situation. But only _a_ factor, and by no means the most important one; for observation has shown that the young are capable of performing the return journey without guidance. Something therefore _is_ inherited, some impulse which comes into functional activity at a specified time, and leads the bird to set forth in a given direction.