V °i8*6 ni ] Dwight, The Sharp-tailed Sparrows. 275 



and we can only suppose they come from unknown breeding 

 grounds. When we realize that nelsoni has not been recorded as 

 breeding east of about 87 W. long, while subvirgatus has not 

 been found west of about 70° W. long., there is ample ground for 

 such supposition. 



Distribution. 



Ammodramus caudacutus is restricted in the breeding season to 

 the salt marshes of the Atlantic coast from Virginia to Massachu- 

 setts. North of the latter named State, in the limited marshes 

 of the New Hampshire and Maine coasts, it is probable that 

 subvirgatus would be found. In fact a few stragglers have been 

 secured that, singularly enough, approach more nearly to nelsoni 

 than to caudacutus as would naturally be expected. One speci- 

 men is from Cambridge, Mass., May 31, another from Revere, 

 Mass., June 7, and a third from North Madison, Conn., June 9. 

 These birds may have been late migrants but the probability is 

 they were breeding. It is obvious therefore that breeding speci- 

 mens from the Maine coast are greatly to be desired. A few 

 caudacutus Linger through the winter as far north as New Jersey 

 (Stone, Birds E. Pa. and N. J., 1894, 114), the bulk passing to 

 the South Atlantic States and even reaching Tarpon Springs, 

 Florida, on the Gulf coast (Scott, Auk, VI, 1889, 322). 



A. c. nelsoni has been sparingly found during the breeding sea- 

 son in Northern Illinois, in Wisconsin, in Minnesota, in Kansas, 

 in the Dakotas, and in Manitoba. It seems to be a compara- 

 tively rare species and spring records are few and far between. 

 Large numbers of migrants are found, however, at many points 

 on the Atlantic coast from Massachusetts to South Carolina, they 

 have been taken at Corpus Christi, Texas (Chapman, Bull. Am. 

 Mus. N. H., Ill, No. 2, 223), and a straggler to the vicinity of 

 San Francisco, California, was described as a new race, to which 

 reference has already been made. It is probable that the birds 

 found at Galveston (Nehrling, Bull. N. O. C, VII, 1882, 12) 

 were of this race and not caudacutus as recorded, and the same 

 maybe true of the record of caudacutus for Ottawa, Ontario (E. E. 

 Thompson, Auk, VI, 1889, 204). 



