MIGRATION OF ANIMALS. While it is among birds (see BIRDS, MIGRATION oF) that migration is seen in its most typical form, it also occurs in many other animals, and it is to these that this article refers. But there has been an unfortunate tendency to apply the term in a loose way to many movements of animals very different from the migrations of birds. Migration in the strict sense, defined in reference to its most typical ex pression in birds, is a racial custom, enregistered in the animal's constitution, and takes the form of a periodic or seasonal mass movement between a breeding place and some other environment in which breeding does not occur. Migration therefore implies that the animals concerned have two haunts in which they reg ularly live at different seasons of the year or at different phases of their life. It is illustrated among seals, turtles, toads, by such fishes as salmon, and by such crustaceans as land-crabs. It is a pity that a convenient term should be blurred by an application to phenomena which it does not fit. The following should be excluded from the rubric of migration. (a) Roaming movements in search of food, whether at short intervals or at different seasons, are not migratory. The march of the Scandinavian lemmings, when they have exhausted the vegetation of a district, is not a migration, nor is the devastating swarming of locusts. (b) Mass-movements, of fishes for instance, that are not related to reproduction, but are instigated by marked changes in the physical conditions or in the dependent distribution of the food should not be mixed up with migrations. (c) The movements of larval animals from their birthplace to another more suitable environment, as when the larvae of shore animals become pelagic and return to littoral waters at metamorphosis. It would be just as unprofitable to apply the term migration to the movement of may-fly nymphs out of the water, or the movement of liver-fluke cercariae out of their water-snail host. The word has been similarly misapplied to the striking march of processionary caterpillars from the pine trees along the ground until they find suitable soft soil into which to burrow for pupation. (d) The various forms of extension of geographical range, which may be very impressive as in the in cursions or "invasions" of sand-grouse into Britain, or may be, as is more frequent, very gradual and hardly perceptible from gen eration to generation. The passive diffusion of gossamer spiders by the wind, or of marine animals by oceanic currents, is certainly not migration. Yet the term is persistently applied to the often striking mass-movements of butterflies and some other insects, which are usually, if not invariably, dispersal-movements.
migrant birds. In some cases, such as the common eel, the adults die after spawning, and the return journey is thus confined to the young.
The scope of migration has been admirably summarized by A. L. Thomson (Problems of Bird Migration, 1926). "To deserve the description 'migratory,' in its strict sense, movements need not necessarily have a very great geographical amplitude, but at the least they must involve a definite change of locality. They must be purposive in that the change of scene is associated with some definite advantage which serves as its raison d'etre, and there must be return movements to the original area. They must be periodic in that they correspond to some recurrent change either in the en vironmental conditions themselves or in the animal's reaction thereto. True migrations are changes of habitat, periodically re curring and alternating in direction, which tend to secure optimum conditions at all times." Mammals.—Perhaps the best instances of true migration in mammals are to be found among seals. The Alaska fur seal (Cal lorhinus alascanus) winters as far south as California, and returns in spring across the north Pacific for 2,000 miles to its breeding place on the Pribilof islands. Cetaceans, being more thoroughly adapted to marine life than the pinniped carnivores, do not need to come to the shore to breed, yet there is evidence of migratory movements. The same may be said in regard to reindeer, but the facts are not easy to understand. The Newfoundland caribou (Rangifer terranovae) moves in autumn from the stormy uplands, where the grazing is apt to be buried deep in snow, to the less strenuous conditions towards the south coast of the island. The southward movement takes place after the mating season, usually late in October, and is normally somewhat leisurely, at a smart walking pace, unless the wintry conditions set in very abruptly. The reindeer travel mostly by day, along more or less well-trodden paths, in relatively small companies, and in single file. The earlier companies consist mainly of does, fawns and young stags, while the later companies consist mainly of the big stags. In the spring there is a return movement, when the does are heavy with young, which is somewhat divergent from avian migration. Too much must not be made of this difference, yet it must be admitted that the so-called reindeer "migration" approximates to the mass-move ments of gregarious ungulates of steppe-like areas when the dry season compels them to "follow the grass." Moreover some cari bous are practically resident both in northern and southern New foundland. There is no doubt that some bats take gregarious flights on a large scale, while others may pass in crowds every evening from the mainland to an adjacent island, just as starlings do. But it is doubtful whether these mass-movements can be in cluded within the rubric of migration.