388 YEARBOOK OF THE DEPARTMENT OF AGRICULTURE. 



at Lanesboro, Minn., 102 days when the wind was south, southeast, 

 or southwest, as against 96 days when the wind was north, northeast, 

 or northwest. Thus the birds migrated with the wind against them 

 just about as frequently as with the wind in their favor. Observa- 

 tions at the lighthouses of southern Florida point to the same conclu- 

 sion. The Biological Survey has the records for many years of each 

 night in spring on which birds were noted passing the lights. These 

 migrants had just reached Florida by a flight over the ocean from 

 Cuba. One might expect them to wait for a favoring wind before 

 starting to sea, but the records indicate that they paid no attention 

 to the direction of the wind. 



In fall it seems to be different, but it must be remembered that 

 most that has been published on the interrelation of bird flights and 

 the wind in fall refers to the late migrants, which have waited 

 until they are forced south by the advance of winter. The larger 

 part of fall migration occurs in late summer and early autumn, be- 

 fore the equinoctial storms set in and the temperature drops. There 

 is no reason for believing that the movements of birds at this time 

 have any more intimate relation to the direction of the wind than in 

 spring. It is true that late migrants hurry southward with a north 

 storm and halt on the advent of a south wind, but the real cause of 

 the southward journey is probably the cold that accompanies the 

 north wind. 



EQUAL FLIGHT LINES. 



Another question arises: Do the individuals of a given species 

 migrate along the Atlantic slope at the same time and at the same 

 average temperature as those in the Mississippi Valley or on the 

 Pacific coast? Few species extend their range from ocean to ocean 

 and are also so common and well known that sufficient data con- 

 cerning them have been accumulated to permit definite deductions. 

 But a study of several wide-ranging species makes it certain that 

 each one is a law unto itself, and that it is not safe to reason from one 

 species to another, even if closely related. Thus the purple martin 

 and the cliff swallow both desert the United States during winter 

 to sojourn in South America ; both return to the United States in 

 spring and breed from the Atlantic to the Pacific; but while the 

 purple martin keeps approximately the same temperature in its ad- 

 vance along the coast and the interior, the cliff swallow moves up 

 the Mississippi Valley at a much lower temperature than along the 

 Atlantic coast, and, indeed, orders its movements with less relation 

 to the progress of the season than any other bird so far studied. 



The summer warbler is so abundant and well known that volumi- 

 nous records of its migrations are on file in the Biological Survey. 

 It observes a very regular spring schedule, as is shown by the accom- 



