The suggestion that migration owes its origin to the Glacial Epoch, "that supposed solution of so many difficulties," to quote Mr Gadow (28), has had many exponents. Some take for granted that the Polar Regions were the original home, the centre of dispersal, of all northern birds, and consequently that migration originated in the gradual pushing back of avian life as the ice gained more and more land each year. During the summer, the birds, urged by an irresistible love of home, travelled as far north as the ice allowed them, but gradually they were driven to nest further and further south until they found refuge in the unglaciated parts of the earth. The individuals and the species, if not the whole families of birds, which failed to retreat, went the way of the "thousand types." On the retreat of the ice, the birds, impelled by a mysterious hereditary memory of home and of the good times enjoyed by their remote ancestors, for very very many generations must have been born under more or less sedentary conditions during the Ice Age, began the same pushing forward each year to the limits allowed them. In this case they travelled nearer and nearer to the original home instead of constantly being driven further from it. Surely the question of original home, at any rate of the home in pre-Glacial days, may be entirely left out of the question. No one can ever prove that this wonderful memory did or could exist. Post-Glacial dispersal northwards, and the foundation of migratory habits of advancing to the new food-producing areas, suitable also for the rearing of young, was doubtless a fact, but would have taken place in any case. The congestion due to the increased numbers driven to a restricted area, would involve a rebound outwards, and the uninhabited areas northward of the refuge would be the natural bourn towards which the birds would travel. The seasonal return of cold would drive them southwards in winter, and the periodical migration habit would thus be originated. The intense love of home during the spread of glacial conditions would tend rather towards extinction than the formation of any new habits. The birds which possessed the greatest attachment to the particular district would be less likely to fly from adverse conditions, and the reduction of their numbers through the ordinary physiological changes in habit--reduction of the number of young produced, and possibly disinclination to pair--would inevitably end in extinction. The stronger the attachment to home the more likely the bird to remain to the bitter end, and if driven away by increasingly severe winters, to return and attempt to nest in the locality which had become unsuitable for nesting. The spread of glaciation would be gradual and so would be the annihilation of the species, but the end would be sure. Birds which are cited as species which have shown this remarkable attachment to home, have disappeared before adverse circumstances--the great auk and the Labrador duck.