PLIGHT OF SEA GULLS MILLER 397 



a level with the eye, but can be fully demonstrated only by photo- 

 graphs which catch the wings at their highest and their lowest points. 

 The full sweep of the wings can be seen by a comparison of Plate 2, 

 A and B, which indicate respectively the beginning and the com- 

 pletion of a stroke. This is illustrated a little less perfectly by the 

 two birds in Plate 2, C ; and in Plate 2, D, by a happy chance, five 

 different phases of the stroke are represented, although neither the 

 full upward nor the full downward extension of the wings is shown. 

 It will be noted by studying the lowest bird in this figure that, on 

 the down stroke, the wing is sharply flexed at the wrist, the forearm 

 being nearly horizontal. 



It should be remarked, however, that while the eye tends to un- 

 derestimate the length of the stroke, the camera somewhat ex- 

 aggerates it. The wing does not actually describe an arc of nearly 

 180°, as might be thought from its extreme upward and downward 

 extensions. It is to be remembered that the body of the bird is not 

 moving on a fixed plane, but undulates with each beat of the wings, 

 rising on the downward stroke and falling a little as the wings are 

 raised. This up and down motion appears from Marey's figures 

 (1895, p. 237) to be about equal to the thickness of the body of the 

 bird. Thus when the wings move from the highest to the lowest 

 position of a beat, their tips describe a shorter arc than if the body 

 were fixed. The undulating motion of the body is usually concealed 

 from the observer for lack of a point of reference, or because it is 

 masked by the greater motion of the wings. 



From the fact that the wing stroke is as long as we have de- 

 scribed, it follows that the beat must also be more rapid than it 

 gives the impression of being. This is found to be true when we un- 

 dertake to photograph a gull in action. The seemingly leisurely 

 flapping of the wings can rarely be caught by an exposure of less 

 than one two-hundredth of a second, and often shows movement at 

 even higher speeds than this (pi. 2, E) . 



In ordinary flight a gull will average about 120 strokes per minute. 

 This involves a rather slow movement near the shoulder, but one 

 which becomes exceedingly rapid toward the tip of a long wing, 

 as we see in Plate 2, E, and in the case of the lower right-hand bird 

 in Plate 2, F, where the humeri are sharply recorded, but the more 

 rapidly moving tips are blurred. 



It is the rapidity of the wing stroke which is the secret of flight, 

 not of gulls alone, but of birds in general. The quick stroke suddenly 

 compresses the resilient air beneath the wing, and this has usually 

 been assumed by theorists to be the means by which the bird is sup- 

 ported; it rides on successive columns of compressed air. Rather, 

 however, should be emphasized the reciprocal of this ; that is, that on 

 the downward stroke a momentary vacuum is left above the wing. 



