PHENOMENA OF FLIGHT IN THE ANIMAL KINGDOM. 283 



But, before going further, we may remark that the precediug experi 

 ment furnishes a very precious lesson in the theory of flight. In foct, if 

 the bird executes a series of ascents and descents, the duration of the 

 descending i^eriod will approximately inform us of the amount of the 

 positive work which the bird must perform to rise again to the height 

 from which it fell, and we see that the duck, which makes nine vibra- 

 tions of the wing per second, executes two vertical oscillations during 



Fig. 31. 



In tlie upper half is seen superposed the muscular tracing, and that of the vertical 

 oscillations iu a wild duck. Below the undulation a, which indicates the elevation of 

 ihe wing, is seen a vertical oscillation; and another, below b, which indicates the low- 

 ering of the wing. In the lower portion are the same tracings obtained from a harrier; 

 here the oscillation at a, which corresponds to the elevation of the wing, is less marked 

 than in the duck. 



each vibration, or eighteen in a second. Each oscillation is composed 

 of a rise and fall, so that each descent of the bird cannot last more than 

 one thirty-sixth of a second. Now, if we substract the effect [)roduced 

 (as in a parachute) by the outspread wings of the bird, we find that a 

 body which falls during one thirty-sixth of a second traverses onlylifty- 

 two millimeters. This fall repeated eighteen times a second constitutes 

 a total rise of 9.3G centimeters, necessary to maintain the bird iu the 

 same horizontal plane during one second. 



In the tracing of the harrier, the descents are less than in the wild 

 duck, probably on account of the large surface of the wings of this bird. 



Determination of the variations of the rapidity of flight. — Tlie second 

 question to be solved relates to the determination of the various phases 

 of rapidity of flight. The solution can be found iu the following man- 

 ner : If the weighted drum be placed upon the bird's back in a A'ertical 

 plane perpendicular to the direction of flight, it will be insensible to 

 vertical oscillations, and will only indicate those of forward and back- 

 ward; also, by turning the membrane of the drum forward it is clear 

 that if the advance of the bird is accelerated, the retardation of the 

 weight on the translation of the annaratus will produce a crowding of 

 the air in the second drum, and a, c' ele\4t^ion of tUft registering lever, 

 while a relaxation of the etibrt of tL'e bird will bring about a descent of 



