ELECTRO-MAGNETISM. 



a quantity of mercury, contained in the 

 vessel, also seen in section ; and the 

 latter being at the end of a wire pro- 

 ceeding from the lower part of the 

 vessel, and in contact with the mercury ; 

 while a magnet is made to float in an 

 upright position in the axis of the vessel, 

 by a weight of platina fixed to its lower 

 end. The magnet, it will be seen, being 

 unconnected with the wires, forms no 

 part of the voltaic circuit, and remains 

 unmoved by it; yet it excites move- 

 ments in the conductors which surround 

 its upper portion. For since, in the 

 parts C and D, the currents of the wires 

 are approaching those of the magnet, 

 they will be impelled (see 200) to turn 

 round it in a direction opposite to that 

 of the currents in the magnet : a reac- 

 tion is at the same time exerted by the 

 currents of the wires upon those of the 

 magnets, which therefore tend to move 

 the magnet progressivety, or in the 

 same direction as its own currents. 

 ($ 203.) But the currents which pass 

 from the lower ends of the wires through 

 the mercury to the exterior of the ves- 

 sel, recede from the magnet, and tend 

 to impress on the mercury a motion of 

 revolution in the direction of the mag- 

 netic currents; and, consequently, by 

 the reaction of this force the magnet 

 receives a tendency to revolve in the op- 

 posite direction. The two forces result- 

 ing from these contrary tendencies of 

 the descending and the receding cur- 

 rents, oppose and partly destroy each 

 other, as far as regards their effect on 

 the magnet; and when the rotatory effects 

 of the whole of the remaining part of 

 the current composing the whole cir- 

 cuit, and including that of the pile itself, 

 and its two wires, W and w, are taken 

 into account, the compensation becomes 

 complete, and the total effect reduced to 

 nothing. Hence we see that, although 

 the wires are made to revolve in one 

 direction and the mercury in another, 

 the magnet itself, being acted upon by 

 equal and contrary rotatory forces, and 

 unattached to any part of the circuit, 

 remains perfectly unmoved. 



(280.) But the case is altered if the 

 magnet be so connected with the appa- 

 ratus of the wires as to form a part of 

 the circuit, even for a portion only of 

 the current; for that portion of the 

 current which thus passes through the 

 magnet no longer exerts upon it any 

 rotatory tendency, and may, therefore, 

 be considered as suppressed : and since 

 the action of this portion exactly coun- 



terbalanced the equal and opposite action 

 of the remainder of the circuit, that 

 equilibrium can no longer subsist, when 

 this portion is removed, and the remain- 

 der of the current becoming effective, 

 will produce a rotation of the magnet on 

 its own axis. The direction of this rota- 

 tion will be the same as that of the 

 descending wires; hence the magnet 

 may be connected with these wires, 

 without altering the nature of the action, 

 as in the experiment of Mr. Faraday, 

 described in 76. 



(281.) It follows, also, from the prin- 

 ciples of Ampere's theory, that when 

 the moveable portion of the circuit 

 which is attached to the magnet has 

 both its extremities in the axis, no mo- 

 tion of this kind will take place ; because 

 no action can result between a system 

 of closed currents and another current 

 terminating at both extremities in the 

 axis of the system. 



(282.) The theory of Ampere implies 

 a perfect identity in the mode of action 

 of a magnet and of an electro-dynamic 

 cylinder. A remarkable difference, 

 however, has been observed between 

 them. In the electro- dynamic cylinder, 

 the poles are situated at the very extre- 

 mities of the cylinder ; whereas, in ordi- 

 nary magnets", they are always found 

 to be nearer to the centre than the 

 ends ; the distance varying in different 

 magnets. This circumstance was long 

 considered as invalidating the truth of 

 the theory *. It may, however, be ex- 

 plained consistently with the hypothesis, 

 in two ways ; either by supposing that 

 the intensities of the currents gradually 

 diminish from the middle to the extre- 

 mities; or else by assuming that they 

 acquire a degree of obliquity when at 

 a distance from the centre of the magnet ; 

 that is, that they move in planes which 

 are not exactly perpendicular to the 

 axis of the magnet, but differently in- 

 clined in different parts. These effects 

 are, indeed, not only quite consistent 

 with Ampere's hypothesis, but follow as 

 the natural consequences of the esta- 

 blished laws of electro- dynamic action t- 

 These positions of the different cur- 

 Tents, according to their positions rela- 

 tive to the axis, will be best understood 

 h-Qmftg. 162, which represents a longi- 

 tudinal section of a magnet by a plane 

 passing through 1he axis ; the directions 

 of the currents being marked by short 



* This was urged as an objection by Mr. Faraday, 

 in the Quaiterly Journal of iiciencp, vol. xii. p. 76. 

 f Ampere, Recueil, &c. pp. 257 and 310. 



