CHANGES IN THE PLANE OF THE BIRd's WING. 259 



same plane as Z, we shall see that these levers correspond 

 with each other, and that every change of plane undergone by 

 the first will be transmitted to the second. 



Under these conditions, if we cause the lever I to signal the 

 changes of plane which the wing undergoes in the various 

 phases of its revolution, these changes will be communicated 

 to the lever V, which can in its turn act on an experimental 

 apparatus, and transmit the signal under the form of a 

 tracing. This is precisely the method which we have em- 

 ployed in our experiments. The lever I was placed upon the 

 wing of the bird, and was held in a horizontal position. 

 The lever V, also horizontal, was fastened by a wire to the 

 lever of an experimental drum placed above it, and arranged 

 in the same manner as in the experiments described in the 

 former chapter. 



When we caused the plane ^f the wing to oscillate, so as 

 to turn its upper surface more or less backwards, the registered 

 curve was depressed ; it rose, on the contrary, when we turned 

 the wing so as to carry its upper surface forwards. 



Still a difficulty presented itself. It was not possible to fix 

 the lever I at one point of the rod t t ; and, at the same time, 

 to render it immovable at a single point in the bird's wing. 

 In fact, the Cardan joint, not having the same centre of motion 

 as the articulation of the wing, it followed that in the vertical 

 movements the rod slipped upon the wing. It was necessary, 

 therefore, for the lever I, while fixed to the feathers of the 

 bird, to glide freely on the rod in the direction of its length, 

 and yet that it should cause it to receive, under the form of 

 torsion, all the changes of inclination that are transmitted to 

 it by the wings of the bird. We see in fig. 109 how this 

 result has been obtained. 



Let tthe the rod which is to follow all the circular move- 

 ments executed by the bird. This rod has in it deep 

 longitudinal grooves, which give its section the appearance of 

 a star ; it glides freely in a tube which is applied to its 

 external surface. But at one of the extremities of the tube is 

 a metallic sliding casting, the interior part of which is grooved 

 like a star, through which passes the rod whose grooves slide 

 in those of the star-shaped opening. Then the lever I is 



