ESKIMO CURLEW. 75 



considerable numbers on the coast of Massachusetts (Mackay) and 

 less often on Long Island (Giraud) and the New Jersey coast (Turn- 

 bull). The Eskimo curlew was absent, except as a straggler, from 

 the whole coast of the United States south of New Jersey, from the 

 Bahamas and from the Greater Antilles. In its southward flight it 

 passed through the Lesser Antilles (Feilden) and along the eastern 

 portion of Brazil (Pelzeln) to its winter home. Throughout the 

 whole line of its fall migration it was unknown in spring, at which 

 season it was traveling northward some thousands of miles farther 

 west over the prairies of the Mississippi Valley. The exact route 

 between its winter home and the United States is unknown, for 

 along the whole 4,000 miles from Argentina to northern Mexico 

 and southern Texas the species has been recorded only twice — once 

 in Costa Rica (Zeledon) and once in Guatemala (Salvin). Its prin- 

 cipal migration route in spring was a comparatively narrow belt 

 crossing the prairies on both sides of the meridian of 97°. The course 

 is well known from southern Texas (Merrill) to southern South 

 Dakota (Agersborg) and thence data are wanting. There seem to be 

 no records of the species from about latitude 44° in the Mississippi 

 Valley until Great Slave Lake is reached, a thousand miles to the 

 northward. 



This curlew is unknown in' the Rocky Mountain States or any- 

 where on the Pacific slope or coast south of Alaska, and the specimen 

 taken April 8, 1892, at Lake Palomas, Chihuahua (specimen in 

 United States National Museum), was far out of the usual course of 

 the species. The species has been taken a few times in western 

 Alaska, south to St. Michael (Nelson) and west to the Pribilof Islands 

 (Palmer) and Bering Sea (specimen in United States National 

 Museum). It has occurred accidentally several times in Europe and 

 on the western coast of Greenland, north to Disco Bay (Winge). 



Spring migration. — The Eskimo curlew arrived in Texas in March — 

 Boerne, March 9, 1880 (Brown); Gainesville, average March 17, 

 earliest March 7, 1884 (Ragsdale); and reached central Kansas about 

 the middle of April— April 14, 1884 (Kellogg); April 13, 1885 (Kel- 

 logg). Most of the records in the central Mississippi Valley are in 

 April. One of the latest and most northern is that of Coues, 

 who says that he saw them in large flocks the second week in May, 

 1873, between Fort Randall and Yankton, S. Dak. Then there is no 

 further news of them until they arrived at Fort Resolution, Mac- 

 kenzie, May 27, 1860 (Kennicott); Fort Anderson, May 27, 1865 

 (MacFarlane) ; Point Barrow, May 20, 1882 (Murdoch). 



Eggs were taken at Point Lake, Mackenzie, June 13, 1822 (Richard- 

 son), and on the Barren Grounds near Fort Anderson, June 13, 1863, 

 June 16, 1864, and June 16, 1865 (MacFarlane). 



Fall migration. — The Eskimo curlew started so early in August 

 that by the middle of the month the old birds reached the eastern 



