ANIMAL MECHANISM. 



under \vliidi (he piano of the wing strikes the air, and 

 secondly, the velocity with which it is lowered. Nothing is 

 inure easy than to obtain the second data of the problem; 

 we can reduce them from the curve which represents the 

 position of the wing at each instant, a curve of which we 

 hav.- an example in fig. 108, as obtained from a pigeon. But 

 the difficulty which presents itself, is to obtain the indication 

 of the changes which take place in the plane of the wing 

 during flight. For this purpose we have had recourse to the 

 following mechanism. 



We have seen, in fig. 99, that a rod connected with a 

 Cardan universal joint, whose centre of rotation is near the 

 Bcapulo-humeral articulation, can be made to represent ac- 

 curately the circular movements of the wing. But Cardan's 

 joint, though it obeys the rotary motions of every kind which 

 are given to the rod, does not allow any movements of torsion 

 with reference to the axis of this rod. 



Viu. 10S. Theoretical figure of tin- apparatus to investigate the torsion c.f 



the wing. 



Let fig. ION lie a kind of apparatus of this sort: we can 

 give the rod / / every kind of motion in the vertical or hori- 

 zontal direction; it will follow all the impulses which it 

 receives. But if we take hold of the extremity of the rod, 

 iM-ar the lever / which is perpendicular to it, and try to give 

 the lever a movement of torsion, as if we were t lining a screw, 

 the Cardan does not allow this movement t > be made, and tho 

 rod re-i^ts the impulse brought to bear upon it. Let us 

 ihat behind the Cardan joint, and on the prolonga- 

 of the rod / /, there is another cylindrical rod, ;>, turning 

 in a tube ; thi> rod will turn under the inllm nee of the torsion 

 evicted by the hand holding the lever /, and if ihe rod j> 

 curies a le\er /', at right angles to it,' and situuti d in the 



