0. B. WARRING. 191 



any direction means a tendency for the other end of the 

 axis to go directly opposite. Hence, if p tends to come 

 towards the observer, the other end, h, tends to go from 

 him, or in the same direction that it was made to go by 

 the upper gyroscope. 



In a '"balance" made of two theoretical gyroscopes, 

 i. e., Motionless gyroscopes with massless rings and 

 axles, the result is curious. The upper gyroscope will 

 slowly drop towards a vertical, while the lower one will 

 rise as far as the form of the balance will permit. If 

 pulled down a little (not to the vertical) it will rise 

 again, and so for several times. If sufficient load is 

 attached to the lower hook it will go down of itself. 

 There will not, however, be an oscillation nor a drawing- 

 oat in proportion to the weight. If the rotation is kept 

 uniform in both, the conclusion will be as follows : The 

 upper gyrostat will become vertical, and then the lower 

 one, and at last the two axes will be in one vertical line. 



I get nearly such an instrument by counterbalancing 

 the weight of the rings and axles (a partial approxima- 

 tion is sufficient for a qualitative experiment). I attach 

 movable weights w' and w to projecting rods, as shown 

 in the diagram, fig. 46. They lie in line with the axes. 

 I move w till it counterbalances the ring and axle, and 

 then adjust the upper one in like manner. The instru- 

 ment, when adjusted and at rest, will stand nearly verti- 

 cal. Now set the wheels both going at once and in the 

 same direction. Lift up the lower gyroscope by its free 

 end so that the instrument takes near about the position 

 it has in the diagram. 



We shall have the horizontal rotation again, but in- 

 stead of the hook, p, gradually falling till the axes are 

 vertical, it will rise — not vibrate — till the wheel of the 

 one strikes the axle of the other. The rebound produces 

 a momentary fall, but the wheel rises again as before. 



The rationale is this. The load on the upper gyroscope 



175 



