448 PROCEEDINGS OF THE AMERICAN ACADEMY 



Having thus shown the different results of these two methods, I 

 purpose now to show that Jarain's results, instead of requiring a modi- 

 fication of Ampere's theory, are a direct consequence of that theory. 



Let us, in approaching this subject, first see what would be the con- 

 dition of the Amperian currents in a bar placed at right angles across 

 the core of an electro-magnet. Suppose the electro-magnet to be 

 placed vertically, and the bar horizontally on top of it. Suppose fur- 

 ther that the current passes, in the part of the inducing coil nearest 

 the observer, from right to left. Then, since the Amperian currents, 

 in the core of the magnet, would tend, at the angles made by the core 

 and the cross-bar, to induce currents in the cross-bar parallel to those 

 in the core, we should have the Amperian currents, in the face of the 

 bar towards the observer, flowing from below upwards in the part of 

 the bar to the right of the core, and from above downwards in the 

 other half; i.e., a current in one direction about the core induces cur- 

 rents in opposite directions in the two halves of the cross-bar. 



I have proved this experimentall}' by placing the coil of fine wire 

 connected with the galvanometer, before spoken of as the secondary 

 coil, at different points on this cross-bar. When the coil was placed to 

 the right of the inducing coil, the deflection of the galvanometer was 

 in one direction ; when the coil was placed on the other side, the 

 deflection was in the opposite direction : showing that, in the two halves 

 of the cross-bar, opposite Amperian currents do actually exist. That 

 this effect was not due to the direct action of the principal coil on the 

 secondary, was shown by substituting a glass rod for the cross-bar. 

 Let us now, keeping the two bars in the same relative position, make 

 the cross-bar the core of two coils, and let the bar which we have just 

 used as a core represent the armature used in Jamin's experiments. 

 When opposite currents are sent through the primary coils, opposite 

 Amperian currents will be induced in the two halves of the bar; and, 

 as the converse of the preceding experiment, parallel currents will be 

 induced in the armature, and these, strengthening each other, will 

 increase the attraction between the bar and the armature. If, on the 

 contrary, parallel currents be sent through the coils, parallel Amperian 

 currents will be induced in the bar, and opposite currents in the arma- 

 ture ; and, if the armature be placed at the middle of the bar, these 

 currents should neutralize each other and the attraction oujiht to be 

 nothinfj. 



To prove these conditions of the Amperian currents experimentally, 

 I have fixed tlie secondary coil on the armature at some considerable 

 distance from the bar, and so investigated the conditions of the currents 



