BANK SWALLOW 419 



autumn. Because such concentrations are so impressive they have 

 been noted and reported by various observers in all parts of the country. 

 Abby F. C. Bates (1895) presents a very interestino; and lucid account 

 of a swallow roost located at the junction of the Mussalonskee stream 

 and the Kennebec River near Waterville, Maine. Extracts of her 

 report are as follows : 



The willow trees along the banks of this stream, particularly a close row 

 some five or six hundred feet in length, form the roosting place of vast numbers 

 of Swallows. During the forenoon and early afternoon very few swallows are 

 to be seen in the sky, — indeed they are conspicuous by their absence, — but a 

 little before sunset the birds begin to arrive in the vicinity, flying, sailing, chasing 

 each other around in the upiwr air, everywhere within eye's reach. * * * 

 Shortly after sunset they gather more nearly in the region directly above the 

 trees, incomers from every point of the horizon still joining them and toward 

 the last exhibiting great hurry and intentness, as if fearful of being "late to 

 meeting." 



Then begin the movements that are the most interesting feature of this gath- 

 ering. At intervals clouds of Swallows will evolve something like order out of 

 their numbers and perform en masse some of the most fantastic curves, spirals, 

 counter-marches, snake-like twists and turns, with the sky as a background, that 

 ever a company of genus homo executed on a finely polished floor. For instance, 

 one evening they separated into two parts, one going to the right, the other to 

 the left, each division making a grand circle outward, then joining again for a 

 forward movement. There were some stragglers, but the figure was distinct 

 and was twice performed, with other evolutions interspersed. Then a long 

 snake-like movement from the upper air down very slightly inclined from the 

 vertical, with two twists in it, a loop around a tall tree farther down the stream, 

 and back, brought them into the tree-tops for roosting. * * * Occasionally 

 they drop down into the trees like pieces of paper, but oftener the final alighting 

 is a combined movement, sometimes in the shape of an inverted cone, — usually 

 in a grand sweep after their most elaborate evolution. Frequently they swoop 

 out of the trees company after company, several times before the last settling, 

 their wings not only making a tremendous whirring, but a perceptible movement 

 of the air. Their chattering keeps np from half to three quarters of an hour 

 after they settle in the trees and their dark little bodies against the sunset sky 

 look as numerous as the leaves. Often they weigh down a branch and then a 

 great chattering, scolding and re-adjustment ensues. * * * The noise which 

 they make is suggestive of the whirring of looms in a cotton mill, heard through 

 the open windows, — or of some kinds of waterfalls. 



They leave the trees in the morning a little before sunrise. August 26 we 

 watched them go out. At 4 :15 there were sounds as if of awakening and grad- 

 ually the noise increased. At 4 :25 they began to arise in companies at intervals 

 of two or three minutes. They did not remain long in the locality and by five 

 o'clock not one was to be seen. 



The above unusual colony consisted of bank and barn swallows with 

 a large number of martins. The concentrations began about the mid- 

 dle of July and reached their maximum during the latter part of 

 August. By September 9 the swallows had departed and only martins 

 were observed at the roost. On September 26 practically all the birds 

 were gone, there being only about 40 martins at that time. The roost 



