158 AN EMPIRICAL STUDY OF GYRATING BODIES. 



different in its application. If a particle on the wheel at cj 

 be pushed horizontally towards the observer it will retain 

 its momentum when it arrives at s. At the same time h 

 will, relatively to the centre of the wheel, move from the 

 observer and, retaining its motion when it gets to r, the 

 joint effect of the two particles will be to tilt the wheel 

 from the observer, or in other words, towards a vertical. 



Another and very curious experiment is described by 

 Sir William Thomson, page 389, Thomson & Tait's Nat. 

 Phil., where is an excellent drawing of his form of the 

 instrument. He calls it a gyrostat on stilts, but says 

 very little by way of explanation. 



Fig. 34 will give a correct idea of my machine. At b a 

 stout wire is soldered, ending in an acute point, which 

 rests loosely in the socket a. At d and e are similar 

 sockets, in which a double-pointed wire rests with great 

 freedom to move in any direction. 



Set the wheel in rapid motion. In a moment or two 

 it will begin to sway towards and away from the ob- 

 server (who is supposed to stand here as in all the ex- 

 periments with right hand opposite b, and left opposite 

 d). These swayings slowly increase in extent till the 

 instrument falls. The fall is not due to loss of rotary 

 velocity, for the process can be gone through several 

 times with once setting the wheel going. 



If, by sliding a sleeve from the wire to the frame 

 (or by any other means), the joint d becomes rigid, the 

 swaying will not occur. On the contrary, the instru- 

 ment will fall at once. This last is merely an illustra- 

 tion of the principle before pointed out, viz., that the 

 knife-edge must have freedom to move laterally. 



Take off the sleeve so that d is free again, then re- 

 move the socket e. If the supporting surface is so smooth 

 and hard that the point e moves over it almost without 

 friction the instrument will at once fall. If e cannot 

 slide, the instrument will stand. 



142 



