PHENOMENA OF FLIGHT IN THE ANIMAL KINGDOM. 281 



trary, we give the mass a lateral motion, no effect is produced upon the 

 lever, and no signal is made. But it may be said that an inert mass 

 placed on an elastic membrane tends to execute vibrations peculiar to 

 itself, and that the apparatus will transmit these vibrations of the mass 

 of lead and the membrane which carries it independently of the oscilla- 

 tions of the bird. How shall we get rid of this complication ? The law of 

 vibrations teaches us that the duration of the double period of each of 

 them varies with the weight of the vibrating body, and with the elastic 

 force of the lamina which carries it. The greater the mass, and the 

 feebler the elasticity, the longer will be the jjeriod of vibration. !Now, 

 the motions which we are studying are rather frequent, some birds 

 making eight or ten strokes of the wing per second. If we arrange it so 

 that the period of oscillation of the mass of lead itself is much longer 

 than that of the bird, we shall no longer be troubled by the complica- 

 tion of these interfering motions. By emi)loying a heavier mass and a 

 less tense membrane, a good transmission of motions, which are not too 

 slow, may be obtained, for instance, such as last less than half a second. 

 It is not necessary, either, that the instrument should be applied to the 

 study of the oscillations of all species of birds. 



But to make sure of the accuracy of the apparatus it should be veri- 

 fied by the method much like that which I have used to correct all my 

 apparatus. This consists in making, directly by hand, the tracing of 

 the motion which I have imparted to the weighted drum, and observing 

 whether the registered motion was the same as the first. 



Experiments made upon different kinds of birds, ducks, harriers, hen- 

 hawks, and owls, have shown me that, in relation to the intensity of the 

 oscillations in the vertical plane, very varied types of tiight exist. 



Figure 30 shows tracings, furnished by different kinds of birds, upon 

 a cylinder turning at a uniform rate, and contrasted with a tracing pro- 

 duced by a tuning-fork making 100 vibrations per second. These 

 tracings enable us to estimate the absolute and relative duration of the 

 oscillations of flight in these different birds. It follows from these 

 figures that the frequency and amplitude of the vertical oscillations vary 

 a good deal with the kind of bird under consideration. 



To better comprehend the cause of these variations, let us register at 

 the same time the vertical oscillations of the bird and the action of the 

 muscles of its wing. If we make this double experiment upon two birds, 

 differing in their manner of flying, such as the wild duck and the har- 

 rier, the tracings represented by Fig. 31 will be obtained. 



The duck presents two energetic oscillations at each revolution of its 

 wing ; the one at b, at the moment when the wing relaxes, is easily 

 understood ; the other, at a, at the moment when the wing rises. To 

 explain the ascension of the bird, during the time of elevation of the 

 wing, it seems to me indispensable to call in the action of the boy's kite, 

 previously alluded to. The bird, moving forward with acquired velocity, 

 presents its wings to the air in an inclined position, similar to that of 

 the kite, and thus transforms its horizontal force into an ascending one. 



The flight of the harrier presents the ascension which accompanies the 

 elevation of the wing, in a smaller degree. May not the cause of this 

 difference be recognized as a smaller relative inclination of the wing 

 toward the horizon ? 



Determination of the different phases of the evolution of the wing, to tchich 

 the vertical oscillations correspond. — The interpretation of these curves 

 throws light at once upon the experiments made on the variations of the 

 transformation of velocity in the bird, at different moments, during 

 the evolution of the wing. 



