192 AN EMPIRICAL STUDY OF GYRATING BODIES. 



is greater than that on the lower, hence, as the horizontal 

 movement is proportional to the load — the velocity of 

 the wheels being, we will suppose, the same — the upper 

 instrument gyrates more rapidly than the lower one 

 would if left to itself ; consequently it pushes that ahead. 

 The effect of this is, by the "first law,'' to make the 

 lower gyroscope rise. 



We can trace out the movements in this case. Suppose 

 the gyroscopes to be rotating and the whole system such 

 that for the moment it may be considered to be in equi- 

 librium. The point p will be almost at rest, and h will 

 be moving from the observer at what may be called the 

 normal rate, i. e., the rate due to the rotation of the 

 wheels and the load. Now, by some means, make it go 

 faster 1 , p still being stationary. There will result a small 

 upward motion of the mass which happens to be at o, in 

 a direction parallel to the small arrow, m, which I have 

 drawn in the margin. In a very small fraction of a sec- 

 ond the rotation carries this mass to o', 90° from o. By 

 its inertia it retains its motion, unchanged either in 

 amount or direction ; consequently there is at o' an up- 

 ward push, and this tends to tilt up the end, p. As o 

 represents each portion of the rim as it passes around, 

 the effect is a continuous one. 



Mr. George Forbes, in Nature for April, 1885, p. 602, 

 in an article on " Molecular Physics,' 1 a sort of commen- 

 tary on Sir William Thomson's paper, describes a "gy- 

 rostatic molecule." It consists of two gyrostats inside 

 of a massless shell, suspended in the ether. He says : 

 "If the shell be Motionless the ether can only give 

 translational movement to it, and the double gyrostat 

 produces a gyrostatic effect when the molecule — the 

 massless shell with its two gyrostats — is accelerated in 

 any direction except along the axis."" 



1 This is easily done by pulling a St. ing- which may be attached near the joint h, only 

 taking care that the pull be in the direction of a tangent to the horizontal movement. 



176 



