I was in a broad grassy swale, studded here and there with scrub spruces and bordered by taller timber, when my attention was attracted by a curious far-off song which puzzled me for some time. Finally I descried the producer, a Wilson's snipe, so far overhead as to be scarcely discernible against the clear sky. It was flying slowly in a broad circle with a diameter of perhaps 600 yards, so that the direction of the sound was ever shifting, thus confusing me until I caught sight of its author. This lofty flight was not continuously on the same level, but consisted of a series of lengthy undulations or swoops. At the end of each swoop the bird would mount up to its former level. The drop at the beginning of the downward dive was with partly closed, quivering wings, but the succeeding rise was accomplished by a succession of rapid wing beats. The peculiar resonant song was a rolling series of syllables uttered during the downward swoop, and just before this drop merged into the following rise a rumbling and whirring sound became audible, accompanying the latter part of the song and finishing it. This curious song flight was kept up for 15 minutes, ending with a downward dash. But before the bird reached the ground and was yet some 20 yards above it there was apparently a complete collapse. The bird dropped as if shot for several feet, but abruptly recovered itself to fly a short distance farther and repeat this new maneuver. By a succession of these collapses, falls, recoveries, and short flights the acrobatically inclined bird finally reached the ground, alighting in the grass near me. All of the early American writers, and many others since then, supposed that the winnowing sound was made by the bird's wings, although many European observers long ago argued that it was made by the two pairs of outer tail feathers, which are widely spread and held downward at right angles to the axis of the body during the downward swoops and vibrate as the air rushes through them. W. L. Dawson (1923) says that-- the body of the sound is produced by the impact of the air upon the sharp lateral feathers of the tail, held stiffly, while the pulsations of sound are produced by the wings. At least it is certain that the pulsations of sound are synchronous with the wing beats. The sound begins gradually, as while the tail is expanding, and closes with a smooth diminuendo as the tail is closing and while the wings are sailing. N. S. Goss (1891) gives a different account of the courtship, as follows: