ORIGIN OF MIGRATION. 59 tance in search of a place where they may conceal their nests. Many species of tropical sea birds resort each year to some rocky islet, situated perhaps in the heart of their range, where they may nest in safety. This is not migra- tion as we understand the word; but, nevertheless, the object is the same as that which prompts a Plover to travel to the arctic regions; moreover, the movement is just as regular. These sea birds pass their lives in the tropics, their presence or absence in any part of their range being largely dependent upon the supply of food. But, as in the case of the Warbler which migrates from South America to Labrador, they are annually affected by an impulse which urges them to hasten to a certain place. This impulse is periodic, and in a sense is com- mon to all birds. There is a regular nesting season in the tropics, just as there is a regular nesting season in the arctic regions. There is good reason, therefore, for the belief that the necessity of securing a home in which their young could be reared was, as it still is, the cause of migration. It must be remembered, however, that birds have been migrating for ages, and that the present conditions are the result of numerous and important climatic changes. Chief among these is doubtless the Glacial period. In- deed, Dr. Allen has stated, and the theory has been gen- erally accepted, that the migration of birds was the out- come of the Glacial period. That their journeys were greatly increased and the habit of migration extended during the ice age is apparently undeniable, but, although previous to the Glacial period a warm temperate climate prevailed nearly to the northern pole, there must even then have been sufficient difference between winter and summer climate to render a certain amount of migration necessary. Furthermore, there is a well-defined migra-
