440 DISTRIBUTION OF ROUGH-WINGED SWALLOW 



Audubon discovered his species near Bayou Sara, in Louisi- 

 ana, October 20, 1819, but at that time did not perhaps 

 recognize it as distinct from the Sand Martin ; for he did not 

 describe it for many years afterward, and then did so from a 

 pair procured in South Carolina. He noted that the bird was 

 like the Bank Swallow, u but readily distinguishable by draw- 

 ing the finger along the edge of the wing, when the stiff pro- 

 jecting tips of the filaments are felt like the edge of a fine saw." 

 He knew nothing of the bird's habits, and surmised that its 

 most habitual residence might prove to be far westward, per- 

 haps the Valley of the Columbia River, which was a famous 

 ultima Thule in ornithology of the Audubonian period. Its 

 distribution is now known to include the entire breadth of the 

 United States, excepting some portions of New England, 

 whence we have no record as yet. But the bird certainly 

 enters New England. This fact was first announced, so far as I 

 know, by Mr. H. A. Purdie, who states that an -individual was 

 shot at Suffield, Conn., bj Mr. Shores, June 6, 1874; and Mr. 

 Merriam states that Mr. E. P. Bicknell found the bird in numbers 

 at Eiverdale, New York, within a few miles of the Connecticut 

 line. I had written in 1868 that it was singular there should 

 be no New England instances on record, " as the species cer- 

 tainly ought to be found there"; and some of the New England 

 ornithologists may learn in the course of time that every bird 

 known in a certain portion of the Middle States will also be 

 found in the Connecticut Valley. Determining thus the north- 

 easternmost point at which the Rough-winged Swallow has 

 been found, we may turn in another direction along its sup- 

 posed northern boundary. Its name appears in Gregg's 

 Elmira list, but not in Mcllwraith's Canada West, nor in 

 Trippe's Minnesota. I never saw the bird in Dakota or Mon- 

 tana; but west of the Rocky Mountains, Mr. J. K. Lord seems 

 to have met with it along the same parallel of 49 ; and we 

 also have Brown's Vancouver record. This exhibits a northern 

 limit coincident with that of Tachycineta thalassina, and we 

 may suppose that the northern border of the United States is 

 nearly the terminus of the species, excepting in New England, 

 where the bird is not known to go so far. 



In the other direction, the Rough-winged Swallow has been 

 traced through Mexico to Guatemala, though some of the 

 extraliinital quotations of " serripennis " may actually refer to 

 other species. In the Middle, Southern, and Western States, 



