140 AN EMPIRICAL STUDY" OF GYRATING BODIES. 



tion causes the gyroscope to rise. Hence it goes up to 

 a vertical, Q. E. D. 



If the wire hook is bent so that p is in the prolonga- 

 tion of the axis, the ring as well as the wheel of the 

 gyroscope will revolve around the axis, and the whole 

 will become simply a top subject to the laws of that in- 

 strument. 



I will anticipate so far as to say that in this case — c, 

 d and p being in a straight line — the gyroscope will not 

 rise if p is a fine, sharp point. Why not '( Simply for 

 the reason that the ring revolves in the same sense as the 

 wheel, and nearly or quite as fast. If it revolves quite 

 as fast there will be absolutely no friction at d ; and if 

 nearly as fast, there will be almost none. In the first 

 case there will be no acceleration, and in the second, too 

 little to be of any service. 



To discuss the effect of friction at p, it would be nec- 

 essary to anticipate what will be said when speaking of 

 the top. I may, however, say now that if p is confined 

 to one point on the glass plate, the friction, so far as it 

 would have any influence, would tend not to accelerate 

 the gyration but to retard it, and hence to make the 

 gyroscope fall instead of rising. The friction at p, in 

 its effect, is a force pushing backward against the rotat- 

 ing body. 



A curious experiment described by General Barnard 

 comes properly here : 



Let a gyroscope be placed as in fig. 23, taking care 

 that the pulley in the ceiling is exactly over the centre 

 of the standard. Set the wheel in motion, first lower- 

 ing c till the axis is horizontal. There will be no gyra- 

 tion, since c cannot fall at all, and since the friction on 

 the end of the axle has, in this position, no tendency to 

 cause it. 



Now raise c till the axis makes a sensible angle with 

 the horizon. The end c being supported by the cord, 



124= 



