WITH MAGNETISM AND ELECTRICITY 189 



copper wire overspun with silk, and 1 millim. thick, forming 

 three layers, each of which contained 294 coils ; the length 

 of the spiral was 383 millims., the interior diameter was 19, 

 and the exterior 23 millims. Sun'ounded with gutta percha 

 for the sake of better insulation, it was inclosed in the wider 

 tube of the spiral of the electro-diamagnet, or rather the latter 

 spiral was coiled round it. 



The most essential point relating to this spiral is, that its 

 length must be divided into two perfectly symmetrical and sym- 

 metrically coiled halves. The wire is not continuously coiled 

 throughout the entire length in the same direction, but the 

 spiral divides itself into two equal portions, which are coiled in 

 opposite directions. This is necessary, if through the motion 

 of a diamagnetic bar of bismuth, or a magnetic one of iron, a 

 current is to be induced in the spiral ; for when this inducing 

 bar is placed in the centre of the spiral and moved there, the 

 force of induction exerted by its north end in one half of the 

 spiral is exactly opposed to that exerted by its south end in the 

 other half, and the action of both would destroy each other if 

 the two halves of the spiral were coiled in the same direction. 

 Through the opposed coiling, the inductive forces, instead of 

 destroying each other, are caused to exert a twofold action. 



This necessary arrangement, for the purposes of induction, 

 presents an important advantage as regards the practical carrying 

 out of the experiments. It is manifest that the current of the 

 galvanic battery, as long as it is constant in the spiral of the 

 electro-diamagnet, can exert no inductive action on the induc- 

 tion-spiral; the slightest alteration of its intensity, however, 

 would be sufficient to induce in the spiral a current much 

 stronger than that diamagnetically induced, and which there- 

 fore would prevent the observation of the latter. It is, however, 

 manifest that the same arrangement of the induction-spiral by 

 which we have secured that the diamagnetic induction in both 

 halves shall be a twofold power, also effects the mutual destruc- 

 tion of the currents excited in the two halves of the induction- 

 spiral by the galvanic current ; so that if the symmetry of the 

 two halves be perfect, the greatest changes of intensity on the 

 part of the current can produce no effect. To this we may 



b 



