KNOT. 81 



June 6, 1893 (Russell) ; the last one noted at Indian Head in 1892 was 

 June 5 (Macoun). Eggs were taken on Anderson River, Mackenzie, 

 June 24, 1863; at Rendezvous Lake, June 27, 1865; and young at 

 Franklin Bay, July 8, 1865 (MacFarlane). 



Fall migration. — That fall migration begins as early as possible is 

 evidenced by the presence of the species in Peru and Brazil by 

 August. It is said to arrive on Grenada and Barbados, West Indies, 

 in July, and the earliest records on the Atlantic coast of each of the 

 New England States is in the same month — earliest July 6, 1874, 

 near Newport, R. I. (Sturtevant). As late as 1879 Doctor Brewer 

 stated that it was not yet known to be a regular migrant in this 

 region, while a gunner near Newport, R. I., had listed 279 individu- 

 als, shot in 1867-1874, the dates ranging from July 6 to September 

 19 (Sturtevant), but these records were not published until 1901. 

 Along this part of its course the species is most common in August, 

 and most have departed by early September. Some late dates are: 

 Newfoundland, one September, 1867 (Reeks); St. John, New Bruns- 

 wick, September 8, 1881 (Chamberlain); Portland, Me., October 13, 

 1906 (Eastman); Key West, Fla., November 1, 1888 (Scott); Bar- 

 bados (Feilden), Grenada (Wells), and in Trinidad (Leotaud), a few 

 remain until October. The species has been noted as casual or acci- 

 dental in Colorado (Thorne), Montana (Coues), British Columbia 

 (Brooks), and Bermuda (Hurdis). 



Knot. Tringa canutus Linn. 

 Breeding range. — The summer range of the knot is almost circum- 

 polar, extending from Iceland across the whole of Arctic America and 

 westward to northwestern Siberia. The species has also been taken 

 once in Spitzbergen, but seems in general to be lacking in the Arctic 

 regions north of Europe. There is every reason for believing that 

 the species breeds locally throughout its summer range, but authentic 

 eggs are a great desideratum. Eggs claimed to be positively identi- 

 fied have been collected at four localities: Fort Conger, Grinnell Land, 

 June 9, 1883 (Greely); Disko, Greenland, 1875 (Seebohm); Iceland, 

 June 17, 1898 (Ottosson); Taimyr Peninsula, Siberia, July 6, 1901 

 (Dresser). These eggs differ so much in size, shape, and coloration 

 that some of them must have been wrongly identified. One of the 

 latent expeditions into the Arctic regions saw young at Goose Fiord, 

 latitude 76° 30' N., but found no eggs (Sverdrup). The species was 

 found breeding commonly at Igloolik on Melville Peninsula, latitude 

 69°, in the summer of 1823 (Parry), but none of the many eggs col- 

 lected were preserved. Three years earlier it had been noted as an 

 abundant breeder on Melville Island (Sabine). At that time the bird 

 was undoubtedly many times more common than now. The Iceland 

 record at 66° latitude is the most southern breeding record, and nest- 

 lings have been taken on Grinnell Land at 82° 44 ' latitude. These 

 represent the extremes of the breeding range. 



