506 BULLETIN 17 9, UNITED STATES NATIONAL MUSEUM 



stragglers being left after the first week in August." At Cape May, 

 N. J., that great funnel point of migration, Stone (1937) has noted 

 the last individuals from September 3 to 15 over a long period. 

 About Charleston, S. C, the last martins are seen approximately at 

 the same time, though a marked reduction in the numbers of the 

 species occurs from mid- July on, when many begin to migrate. 



Proceeding southward, the last ones leaving Florida are recorded 

 by Howell (1932) as from late September to October 2 about Pensa- 

 cola and Tallahassee, while down in the Keys, at the jumping-off 

 place, they were noted at Sombrero Key on October 6. A rather 

 late flight was recorded at St. Marks (south of Tallahassee on the 

 Gulf) on November 10, 1912. Again, in another November flight, 

 Howell states that Mrs, Hiram Byrd saw "great flocks of swallows 

 over the glades" on the 11th, of which "some were Martins." A 

 flight seen by this observer on November 4 was described by Howell 

 as being "between Royal Palm Hammock and Homestead." Here 

 again Howell confuses Royal Palm Hammock with Royal Palm Parh^ 

 the latter being Paradise Key in Dade County, while the former lies 

 across the Tamiami Trail in Collier County, a hundred miles or so 

 to the westward. Though in this case he says "hammock," he means 

 "park" for he quotes Mrs. Byrd as saying that the martins were 

 "circling over a glade at the edge of the Park." She estimated this 

 flock to cover 9 miles of territory and to contain "anywhere from a 

 hundred thousand to a million or more." 



H. W. Ballantine noted three birds at Orlando on December 18, 

 1915, which Howell says "may be considered wintering birds." This 

 is a parallel case to his statement that gray kingbirds must some- 

 time winter in Florida because of a single specimen seen in Royal 

 Palm Park in December. Rather would it appear that, like the 

 kingbird, these three martins at Orlando were belated migrants, for 

 in six years of constant field work in southern Florida I have never 

 seen a single specimen, not even in the keys. It is my confirmed be- 

 lief that the birds leave the country entirely, and there is nothing to 

 offset this except the single exception of the December Orlando 

 birds above, if that can be considered an exception. Oberholser 

 (1938) gives the departure date (latest) from Louisiana as October 

 22. Definite knowledge of its winter home in Brazil has resulted 

 from banding, and Lincoln (1939) gives an example of a specimen 

 "found" in December 1936 (the same month in which the Orlando 

 "wintering" birds were seen) near Ttaituba, Para, Brazil, which had 

 been banded at Winona, Minn., on May 30, 1934. 



Thus, it will be seen that the autumnal migration of the martin is 

 almost as leisurely a matter as is the sprmg movement. It covers 

 a rather long period. 



