STARLING 201 



Intensive observatioBS were made at the roost this year (1941) during the 

 month of March. On fair evenings the birds began to appear in the trees along 

 the top of tlie gorge at about six o'clocli. They seemed to converge in small 

 flocks from all directions during the next half-hour. The smaller flocks aggre- 

 gated into larger ones, but did not descend to the roost until about dusk, probably 

 because the inhabitants of the valley in attempting to break up the roost, shot 

 into the flocks frequently. The skillful maneuvering by the close formations 

 of these large flocks as they veered sharply first in one direction and then in 

 another while flying around before settling was an Impressive sight not easily 

 forgotten by any observer. When the birds settled, the trees appeared, from 

 a few hundred yards distant, as though in full leaf. The small grove in which 

 the birds settled did not exceed four acres, but the birds were so numerous that 

 many limbs of tlie trees were broken and dung in places was four inches deep. 

 It was impossible to estimate accurately the total number of blackbirds. Local 

 ornithologists set the figure at anywhere from a half to around several mil- 

 lion. * * * Of the three species seen at the roost, Starlings {Stnmiis vul- 

 garis vulgaris) were by far preponderant, outnumbering Bronzed Grackles 

 (Quiscalus quiscula aencus) at least twenty-five to one. A number of Cowbirds 

 {Molothrus ater ater) were there during early March, but few were seen late in 

 the month. 



Kalmbach and Gabrielson (1921) say of a New Jersey roost : 



At a roost in a marsh along the Hackensack River an opportunity was afforded 

 to watch the starlings congregating. As early as 3 o'clock in the afternoon 

 flocks of a dozen or two could be found gathering in the hayflelds in the vicinity, 

 or perching on dead chestnuts, singing and preening their feathers. Most of 

 these were juveniles with the molt extending up as far as the neck. They would 

 fly alternately to the hay stubble, which was heavily infested with grasshoppers, 

 and then to the tree tops when flushed. By 4 o'clock a flock of a hundred or 

 more had gathered. * * * With the approach of evening the birds would 

 rise and perform numerous flight evolutions, in which they displayed wonderful 

 coordination of action. This was best observed when they would fly in the 

 direction of the sun, and the flashes of light coming from their glossy backs 

 appeared as coming from a single mirror instead of from several hundred bodies 

 acting independently but in perfect unison. * * * As dusk approaelied, 

 the birds had worked their way toward the Hackensack River, where they 

 gatliered in compact flocks, singing in the tree tops along the bank. A few 

 were seen feeding with a large number of red-wings on the tidal flats along 

 the edge of the marsh. When darkness finally came the starlings in the tree 

 tops sailed out over the marsh and joined their relatives, perching on the cat- 

 tail flags for the night. 



Mr. Kalmbach (1932) tells lis something about the city roostings 

 of starlings in Washington, D. C, where he and others did some ex- 

 tensive banding in winter roosts. Fully a thousand starlings roosted 

 in ventilators on top of the Post Office Department Building on 

 Pennsylvania Avenue, in December 1927, but they were not easily 

 caught there ; they abandoned the roost and promptly changed to the 

 tower of the First Presbyterian Church, where as many as 2,000 re- 

 paired nightly ; 1,241 starlings were banded there on March 2, 1928. 

 In the winter of 1928-29, starlings were banded in the tower of the 

 Metropolitan Memorial M. E. Church. 



843290—50 14 



