32 NOETH AMERICAN SHOEEBIKDS. 



Winter range.— The breeding knots of Siberia go south in winter 

 to southern Africa, India, Australia, and New Zealand; those from 

 Arctic America winter in South America, south to Tierra del Fuego, 

 where they were found to be common February, 1895 (Schalow). 

 The species is rfecorded locally from the coasts of South America, but 

 present data are not sufficient to define the northern limits of the 

 winter range. It is not probable that the knot winters regularly 

 anywhere north of South America, and all records to the contrary, 

 as the three seen in January, 1890, on Muskeget, Mass. (Mackay), 

 must be considered as accidental. On Barbados the species has been 

 noted as late as December 27, 1886 (Manning), but it is there con- 

 sidered to be only a migrant. 



Migration range. — ^The principal migration route is along the Atlan- 

 tic coast, where the knot is known locally from Florida to Newfound- 

 land, and was formerly quite common. There are notable gaips in 

 the records of this species. It is known locally as a tolerably com- 

 mon migrant throughout the Mississippi Valley east of the ninety- 

 eighth meridian, but apparently these Mississippi Valley birds pass 

 north and a little east to Hudson Bay and thence to the Arctic 

 islands, for the species is unrecorded in the whole interior of Canada 

 west of Hudson Bay, and has not been found even on the Arctic 

 coast of Mackenzie. Southward there is another break in the rec- 

 ords, for the specimen taken April 13, 1904, at Rivera, Veracruz 

 (Piper), seems to be the first and only record for Mexico, and there 

 is none for Central America, though the species is moderately com- 

 mon in Texas south to Corpus Christi (Sennett). A few individuals 

 of this species have been seen in migratipn on the Pacific coast from 

 San Diego, Calif. (Dwight), to Cape Blossom, Alaska (GrinneU). 



Spring migration. — The knot arrives on the United States coast in 

 April, but the larger flocks come about the middle of May, and there 

 is no apparent difference in the dates for the whole coast from Florida 

 to Massachusetts. An early date is March 28, at Grand Isle, Louisi- 

 ana (Beyer, Alhson, and Kopman). Near the northern limit of the 

 range some dates of arrival are: Point Barrow, latitude 71° 20' N., 

 May 30, 1883 (Murdoch); Fort Conger, latitude 81° 40' N., June 3, 

 1883 ^Greely); Floeberg Beach, latitude 82° 30' N., June 5, 1876 

 (Feilden), while far to the southward at Winter Island, latitude 66° 

 N., the first was not noted until June 16, 1822 (Parry), and the next 

 year at Igloolik, a few miles farther north, not until June 14 (Parry). 



Fall migration. — Birds from the north arrive on the coast of Mas- 

 sachusetts, on Long Island, and in some seasons, even on the coast 

 of South Carolina (Wayne), by the middle of July; the first was seen 

 at the Olympiades, Wash., July 7, 1905 (Dawson). It seems scarcely 

 possible that these early arrivals can have bred the same year, for 

 earliest dates of young are in July and that at places 2,000 miles or 



