60 ORIGIN OF MIGRATION. tion in the southern hemisphere, where no evidences of glaeiation have as jet been discovered. As I have said, the existing conditions are the result of changes which have been active for ages. No species, therefore, has acquired its present summer range at one step, but by gradually adding new territory to its breed- ing ground. For example, certain of our Eastern birds are evidently derived through Mexico, and in returning to their winter quarters in Central America, they travel through Texas and Mexico and are unknown in Florida and the West Indies. Others have come to us through Floritla, and in returning to their winter quarters do not pass through either Texas or Mexico. This is best illus- trated by the Bobolink, an Eastern bird which, breeding from New Jersey northward to Nova Scotia, has spread westward until it has reached Utah and northern Mon- tana. But — and here is the interesting point — these birds of the far West do not follow their neighbors and migrate southward through the Great Basin into Mexico, but, true to their inherited habit, retrace their steps, and leave the United States by the roundabout way of Florida, crossing thence to Cuba, Jamaica, and Yucatan, and mn- tering south of the Amazon. The Bobolinks of Utah did not learn this route in one generation ; they inherited the experience of countless generations, slowly acquired as the species extended its range westward, and in return- ing across the continent they give us an excellent illustra- tion of the stability of routes of migration. They furnish, too, an instance of one of the most important factors in migration — that is, the certainty with which a bird returns to the region of its birth. This is further evidenced by certain sea birds which nest on isolated islets to which they regularly return each year. Of this wonderful " homing instinct," which plays so
