CHIMNEY SWIFT 289 



at 4.30 p. M., when most of the flock was found to have grouped itself in close 

 formation, as shown in the rough sketch. * * * 



The birds seemed two or three deep, and several of us estimated well over a 

 hundred of them. They were snuggled together, seemingly to keep warm, and 

 the heads all concealed beneath the wings of those above. This patch of birds 

 was of irregular shape, nearly 5 feet high and 7 to 8 inches wide at the widest 

 part. It was constantly changing, as some birds seemed to lose their grip and 

 fly off and return, so that a dozen or two were on the wing and seeking a 

 place to work into the group. We saw some alight at the edge and work up 

 close, while others lit in the middle of the group and must have reached through 

 with claws to grasp bird or bark, those failing falling back and taking wing. 

 All had their heads concealed but the few upper ones. Toward dusk the birds, 

 matching the moist bark, were invisible, but we examined them again by flash- 

 light after dark, and all was quiet. 



Next morning, to our surprise, they were still there, in broad daylight, 

 and some remained through to the afternoon. 



Cottam (1932) describes in detail some remarkable gatherings of 

 swifts "at niglit circling the great dome of our national capitol, feed- 

 ing on the small insects attracted there by the powerful flood lights." 

 He observed the birds on many evenings, both in spring and fall, once 

 in a flock of "approximately 2,000," circling "the dome — the area of 

 greatest light concentration — where they remained until the lights 

 were turned off shortly after midnight." Of the bird's evolutions, he 

 says: 



On the nights when flocking occurred at the capitol, the birds began to arrive 

 in small groups from all directions about sundown, and by the time they nor- 

 mally would have been going to roost they had formed into one great swarm. 

 For the first fifteen or twenty minutes after sundown the birds foraged over 

 the tree tops and flew in all directions without any apparent system to their 

 movements, except that they remained in a rather restricted area. Gradually, 

 as it grew darker, a greater number were seen to fly more or less in the same 

 general circular direction ; in other words, there was a distinct impression of 

 group movement. About the time the lights came on or shortly thereafter, all 

 Avere following a definite course. Each time flocks of incoming birds disrupted 

 the rhythm and unison of the concentric flight there was a momentary disband- 

 ing. When tiiey reformed, however, all seemed instinctively to fly in the same 

 direction. Most often the flight was uniformly circular, but occasionally it took 

 the form of a conical cloud somewhat resembling a cyclone funnel. On one 

 occasion it was seen to form a great figure "8" with one loop at a lower elevation 

 than the other. 



Frederic H. Kennard (MS.) makes this note of an unusual roosting 

 place at Duck, Maine: "In the evening [August 8, 1924] I was treated 

 to a performance of flocking, roosting chimney swifts, which at sun- 

 down flocked up and, flying in circles about one end of the smaller of 

 Charlie Boyce's barns, gradually dribbled into a little window up 

 under the ridgepole of the gable end and there clinging by the hun- 

 dreds in an almost solid sheet against the gable end of the barn. We 

 climbed up into the hayloft and flashed a light onto them, and they 

 gradually flew out until only perhaps 75 to 100 were left. There 



178223—40 25 



