The length of migration varies greatly; the pine warbler withdraws in winter only a short distance from the southern limit of its breeding range, whereas the most northerly breeding black-polls migrate from Alaska to the Tropics. In spring many species migrate at nearly the same time, apparently advancing northward in intermittent waves of great numbers during favorable nights. Flocks made up of sometimes a dozen species together flash about in their bright plumage during the week or two at the height of the migration and furnish days of great excitement to ornithologists. Their return in late summer and autumn is more leisurely and regular; in loose flocks they drift slowly by for several weeks, their southward passage evident even in daytime. The flocking begins early, soon after nesting is over, and to the north is apparent early in July, if closely watched for, even before the leaves begin to wither. The mixed fall flocks, with adults in winter plumage and young birds in duller colors, present many fascinating problems in identification as the birds move quietly along. [AUTHOR'S NOTES: When I asked Dr. Tyler to contribute these remarks we discussed Professor Cooke's (1904) theory of trans-Gulf migration, which has been generally accepted until recently, when it was challenged by George C. Williams (1945). This paper started a discussion in which George H. Lowery, Jr. (1945), has taken a prominent part, and of which we have not yet heard the last. Routes of migration from South America to the United States are evidently well established through the West Indies and the Bahamas to the southeastern States; across the Caribbean to Jamaica, Cuba, and Florida; through Central America and directly across the Gulf from Yucután to the Gulf States; through eastern Mexico and Texas; and through western Mexico to the southwestern States. Professor Cooke was probably correct in assuming that the majority of wood warblers breeding in eastern North America migrate directly across the Caribbean or the Gulf. Some species may confine themselves to only one of the routes named, but we need more data to say just which species uses what route.]