107 



round it in the opposite direction to that in wiiich it was 

 spinning. 



It might be that the electricity in the inducing body- 

 would produce an opposite magnetic effect on the top ; but 

 even if this were so (and I do not think it has been experi- 

 mentally shown that it would be so), its effect, owing to its 

 distance, would be much less than that of the electricity on 

 the very surface of the top. If we take no account of the 

 effect of the inducing body the current round the top would 

 be of such strength that it would carry all the electricity 

 induced in the top once round every revolution. And if the 

 top were spinning from west to east by south it would be 

 rendered magnetic with the positive pole uppermost, that 

 is, the pole corresponding to the north pole in the earth or 

 the south pole of the needle. 



In order to show that such a current might be produced, 

 a glass cylinder, twelve inches long and four across, was 

 covered with strips of tinfoil, parallel to the axis, with 

 very small intervals between them. These strips were about 

 six inches long and one half inch wide, and the intervals 

 between them the two-hundreth of an inch. In one place 

 there was a wider interval, and from the strips adjacent to 

 this wires were connected by means of a commutator with 

 the wii'es of a very delicate galvometer. This cylinder 

 was mounted so that it could be turned twelve hundred 

 revolutions in a minute, and brought near the conductor 

 of an electrical machine. This apparatus, after it had been 

 thoroughly tested, was found to give very decided results. 

 As much as 20^ deflection was obtained in the needle, and 

 the direction of this deflection depended on the direction in 

 which the cylinder was turned, and on the nature of the 

 charge in the conductor. When this was negative the 

 current was in the opposite direction to that of rotation. 

 It may be objected that the measurement was not actually 

 made on the cylinder. It must, however, be remembered 



