C. \i. WARRING. Ill 



may be attached. All things being in readiness, connect 

 with the battery. The wheel of the gyroscope at once 

 begins to revolve, quickly reaching its maximum speed. 

 Attach at either end a moderate weight, gyration will 

 instantly commence. 



In a very few seconds it will be evident that the loaded 

 end is descending. Now change the weight to the oppo- 

 site hook ; the same results will follow, varying a little in 

 the time, according as the torsion of the suspending 

 string acts with, or against, the gyration. 



In every case (save where the torsion is too great, and 

 acts with the gyration ) the loaded end falls. 



The experiment need not occupy more than ten 

 seconds, and for that time even a bichromatic battery 

 may be regarded as sensibly constant. 



I think, therefore, we may justly conclude that the 

 fall is not " due to the retarding influence of friction and 

 the resistance of the air ;" for all this influence is coun- 

 teracted by the magnet and the torsion of the string, and 

 yet the gyroscope falls. 



The absurd results that would follow if General Bar- 

 nard's statement, were correct, will become evident by a 

 little thought. 



If the descent were due only to loss of rotary velocity, 

 then, as this loss is very gradual, and, in fact, almost 

 nothing, for, say one second, the power of the gyroscope 

 to resist a downward force for that time ought to be 

 almost infinite. During the next second a similar result 

 would follow, and so on. But, in fact, a very small force 

 overcomes the instrument's staying up power. 1 



From all of which I deduce the following law : 



A gyrating body tilts in the direction of the force 

 which tends to make it revolve around a second axis ; 

 and this tilting occurs whether the rotation is uniform 



1 The nutation of the earth, as will be shown later, affords an example of a gyrating 

 body absolutely frictionless, which yields to the force pulling it. 



95 



