of fieldfares, redwings, and some of the finches which come to winter in the British Islands wander continually from feeding ground to feeding ground, remaining in one place only so long as the food supply is plentiful. When there is a plentiful harvest of beech-mast, chaffinches and bramblings will linger near one clump or avenue of beeches for many weeks, but when, as often happens, the mast crop fails, they become nomadic, and pass from place to place in their hunt for food. They visit fields top-dressed with manure, glean the refuse of the harvest, frequent the farm-yards, and in early spring, visit the budding larches to prey upon their insect pests. On the other hand golden plovers and lapwings are remarkably local in their winter habits, and so long as the weather remains open will frequent the same fields throughout the winter. Severe weather, especially snow, which effectually closes their chance of obtaining food, at once drives them away. They will migrate to the unfrozen mud-flats of the coast, or to those parts of England, generally the south-west, and Ireland, where the climate is normally milder, or they will even leave our islands altogether under great stress.

The wandering habit, except during the breeding season, is confirmed in most birds, and experience shows that the same species of birds visit the same districts again and again when there is some particular food supply to attract them. Memory and experience guide them from place to place. This regular visitation of certain food bases, being of the greatest importance to birds which have a long period of travel or wandering before them, tends to originate the so-called route by which they travel. The fact that as a rule these stages are in consecutive steps southward is surely due to the fact that the temperature is falling in the north more rapidly than in the south. That they are not always due south is certain. The American golden plover, as Mr Wells W. Cooke so lucidly demonstrates, at first travels eastwards from its home in western Arctic America to the fruit-laden lands of Labrador and Nova Scotia, where it feeds for some time, stoking up for its long oversea journey due south. Mr Cooke says, "It can also be said that food supplies en route have been the determining factor in the choice of one course in preference to another, and not the distance from one food base to the next. The location of plenty of suitable provender having been ascertained, the birds pay no attention to the length of the single flight required to reach it" (21). During the evolution of the route many bases would be found which were superior to others, and skipping and the gradual shortening of the journey from one to another would result. The final goal, the food base which in any weather or season provides the safe sufficiency of food, having been reached by the birds, this becomes the winter quarters. They return to this secure retreat each winter, instead of aimlessly wandering in search of a better, and thus the long-distance migratory habit is formed. Heredity tends to confirm this and it becomes an instinct.

Any observer may verify the assertion that birds regularly visit certain favourable food-bases by paying attention to the occurrences of birds of passage. The study of a county, for instance, shows that certain species show partiality for particular localities. Thus in Cheshire goldeneyes pass through every spring and autumn, and may be met with occasionally on any of the meres; but at Oakmere, in the Delamere district, one may be almost certain of seeing parties of this species any time during the periods of passage. The curlew may be heard or seen passing over any part of the county, but only in the Delamere fields do we frequently meet with flocks feeding in inland Cheshire. Before the winter resident golden plovers have arrived in autumn and after they have departed in spring, the favourite fields are regularly visited by passing flocks, and the lower reaches of the Mersey, where the common sandpiper is rare as a summer resident, are visited every autumn by parties of birds on passage. Chance may lead a casual wanderer to a good food-supplying spot, but the regularity of appearance suggests habit and memory.

A fact which supports the theory that birds ramble far in search of food in their winter quarters, is that in many species the winter range is more extensive than the breeding area. Thus Mr Cooke shows that the known breeding area of the Pacific golden plover has an east and west extension of some 1700 miles, but in winter it ranges over an area with an east and west extension of about 10,000 miles. The scarlet tanager, however, has a breeding range extending for some 1900 miles across eastern Canada and a winter home in north-western South America of only some 700 miles in extent.

The winter quarters, or the outermost limits of the individual but not necessarily the specific range, having been reached, the bird spends its time in seeking food, remaining in one place if food is plentiful, or wandering, according to necessity or the habit of the species. The assertion that some birds have a second breeding season in their southern home is either unsupported by any direct evidence or is the result of a mistake in identification; the bird which has been found breeding has in several instances been shown to be a southern form or a related species of the one it was thought to be.

As the northern spring approaches, the strongest of all animal instincts, on which reproduction and the very existence of the species depend, overcomes all other desires, and the bird grows restless. The hereditary instinct, the origin of which we have endeavoured to show, urges the bird to seek the breeding area which has by degrees become so far removed from the winter quarters. The bird returns home.

But here is a serious difficulty urged by some writers as a powerful argument against the sexual impulse as the great factor in the return journey. Many of the birds which migrate northwards or homewards are sexually immature, and others of them are undoubtedly to be classed as "non-breeders," which means that during that particular summer they will not be engaged in the work of reproduction; why, then, should young birds or non-breeders migrate from the winter base. Possibly in the early days of migration only the mature birds did return; that we cannot state one way or the other. But it is reasonable to argue that once a regular migration habit has become not only confirmed by heredity but a very true advantage to the species, its influence will be felt by each and every individual. Again it is clear that the sexual impulses, in an undeveloped form, are appreciated by the adolescent, and in many animals by even the most juvenile. The play of all young animals is either an imitation or reflection of the search for food—the hunting instinct—or the love-making and sexual quarrels pertaining to reproduction, the pretended competition by the young for the favours of the opposite sex. They may play at and actually perform a migration which is so closely bound up with the life of the species. That this impulse has not always sufficient strength to force them to perform the whole journey is apparent from the fact that many non-breeders, young or sexually mature, on their northward journey through our islands or along our coasts, never reach the breeding area; the food supply on the way attracts them more than the memory of home; they linger with us until the breeding season is over and the return journey has begun. Knots, sanderlings, turnstones