284 PROF. O. REYNOLDS ON THE SUSPENSION 



SO that the tendency of the jet to force the ball to one side 



P' 



increases from nothing to — as p moves from the bottom 



to a point at which the normal makes an angle of 45° with 

 the vertical, and then decreases to nothing as p moves to 

 the middle of the ball. 



The force Qt may be fairly assumed to increase as the 

 speed of rotation increases; and this will be as the point of 

 contact moves from the bottom to the middle of the ball. 

 In the same way the force R, which will necessarily in- 

 crease as Q, increases, will increase as p moves from the 

 bottom to the middle of the ball ; and its horizontal com- 

 ponent will follow nearly the same law as that of P. 



Considering, then, the horizontal forces only, there must 

 be some position for p in which the horizontal component 

 of Q and R will be equal to that of P ; and if a horizontal 

 circle be drawn through this point, it will limit the part 

 of the ball in which equilibrium is possible. 



For any deviation without this circle the equilibrium 

 will be stable; i.e. if the centre of the .ball gets so far from 

 the jet that the ball is struck in some point without this 

 circle, it will come back again. As to the nature of the 

 equilibrium for any deviation within this circle, I cannot 

 speak positively ; but it is probably nearly neutral all over 

 the enclosed area. This seems to agree very well with the 

 fact that the ball is in equilibrium when struck 45° below^ 

 its horizontal circle and oscillates about this position. 



The following is a description of some experiments. The 

 object was to ascertain : — first, whether or not air is the me- 

 dium by which the water acts on the ball ; secondly, how 

 far the horizontal equilibrium of the ball depends on its 

 rotation; and, thirdly, what is the exact position of the 

 point in Avhich the ball must be struck so as to be in equi 

 librium, and, moreover, what is the nature of the equi- 

 librium. 



