The young are mostly on the wing toward the end of July, and the birds begin to gather into flocks along the muddy edges of the brackish pools and the banks of tide creeks. Very soon after this they begin to lose their summer plumage, and the molt continues until the last of September or first of October. During the first of October they are very common in flocks and singly among the lakes and streams; a little later and the borders of these situations are edged with ice and most of the birds leave for the south, but some of the hardier ones betake themselves to the seashore, where they join with Coues's sandpiper and remain as late as the 12th or 13th of the month. The southward migration separates into two widely divergent main routes, with only stragglers between. One route is southward along the Pacific coast and one southeastward along the west coast of Hudson Bay, through the eastern Great Lakes, and to the coast of New England and farther south. E. A. Preble (1902) saw them on the west coast of Hudson Bay, just commencing the migration, on July 19, and "present by thousands" south of Cape Eskimo on August 3 to 13. It seems to be a rare bird in the interior Provinces of Canada; my Manitoba correspondents have no fall records, and Professor Rowan has only one for Alberta. Mr. Brewster (1925) saw it regularly at Lake Umbagog, Maine, in October; and W. E. Clyde Todd (1904) calls it common in Erie County, Pennsylvania; probably these two points represent the north and south limits of the eastward route. Mr. Todd (1904) quotes from Samuel E. Bacon's notes as follows: In former years extensive flights took place about the 1st of November, upon which occasions bushels of them are said to have fallen to a single gun. During these great flights the flocks were accustomed to follow the outside beach of the peninsula (having presumably come directly across the lake) to its southeastern extremity, thence crossing over to the sand beach east of the mouth of Mill Creek, where, after having been sadly depleted by dozens of guns, they would finally rise high in the air and pass southward over the mainland, flock following flock, all day long. I know this by hearsay only, but am positive that this is the bird that used to arrive in such numbers late in the fall. On October 29, 1897, I killed 53 of these birds out of two flocks, comprising in all perhaps as many more, and this is the nearest approach to a flight that has occurred of late years.