in a Letter to M. Gay-Lussac. 363 



rection alone, it will move towards the centre c 9 for the cur- 

 rents produced by a marked pole (north) are exactly such as 

 by their action on the pole to impel it in that direction. 



This relation of the currents to the pole by which they are 

 generated, may be proved by experiment as easily as by cal- 

 culation. I have shown (F. 100.) that when a pole marked 

 north is above a disc revolving in the direction of the arrows R, 

 in the figures annexed either to Messrs. Nobili and Antenori's 

 memoir or to mine, the currents (indicated by the circles) are 

 as is represented in figures 3, 6, or 7. Upon arranging a 

 metal wire which would conduct the currents in this double 

 direction (fig. 8.), and placing over it a marked pole (north) 

 capable of moving only in a parallel direction to r c, at any 

 point in the line re, I found it had not any tendency to move. 

 There was also another line perpendicular to the first, and 

 which crosses it at the point of contact of the circles, in which 

 the pole had no tendency to move. If placed in any other si- 

 tuation than upon these two lines, it moved either in one direc- 

 tion or the other ; and when placed in the positions marked 

 1, 2, 3, 4<, it moved in the direction of the arrows represented 

 at those points. Now the position of the pole, with regard 

 to the currents in Arago's experiment, when the magnet and 

 the disc are arranged as in figures 5 and 7, is exactly that 

 of the point 1 in fig. 8, and hence that pole has a tendency 

 towards the centre C. 



We will now direct our attention to the result obtained if 

 we gradually move the pole from the centre towards the cir- 

 cumference. Let figure 9. represent this new condition at a 

 given time, as figure 5. represented the first state; it is evi- 

 dent that the velocity of the parts a b of the radius beneath 

 the pole, will not differ from each other so much as they did 

 previously, being only about 3 : 2 instead of 6 : 1 ; and the 

 difference will also be less with all the curves of equal inten- 

 sity comprised in this circle. This occasions the situation of 

 the pole, and the place of contact of the circles representing 

 the currents (fig. 7.) mutually to approach in the direction of 

 the line r c, and necessarily carries the pole (fig. 8) nearer 

 to the neutral line / i. If we examine the second circle c d, 

 fig. 9, of magnetic curves of equal intensity, it will be seen, 

 that as the disc does not extend to c, or even beyond a, there 

 is nothing to add to the force of the current upon that side of 

 the pole, while at d the radius, by moving through the mag- 

 netic curves, adds to the intensity of the current excited at 6, 

 and everywhere else on that side of the pole, and may easily, 

 according to the position of the pole upon the metal plate 

 (that is, according as it is nearer or further from the edge), 



