ELECTRO-MAGNETISM. 



85 



direction m E M, because the current 

 in m is attracted, and that in M -repelled 

 by the current in E D. 



(275.) If the point P were inserted, 

 not, as before, in the centre of the fluid 

 above the magnet, but to one side of it, 

 the action of the currents would be 

 more complicated, some being attrac- 

 tive and others repulsive, according to 

 their situations relative to the magnet. 

 The resultant force will be one at right 

 angles to the line joining the centre of 

 radiation with the axis of the magnet, 

 and the effect of this force will be a 

 motion of translation of the whole mag- 

 net ; that is, of revolution round a line 

 parallel to its axis, and exterior to its 

 surface. 



(276.) The presence of transverse 

 currents in every part of the surfaces of 

 magnets is well illustrated by their con- 

 joined influence, when a number of 

 magnets are placed horizontally, as in 

 jig. 160, like the spokes of a wheel, with 

 their similar poles turned towards the 



Fig. 160. 



centre C. In this situation all the cur- 

 rents on the upper sides of the magnets 

 are passing in the same direction with 

 reference to the circumference of circles 

 described from the centre C. They will 

 therefore produce continued rotation in 

 a vertical conductor, whose axis passes 

 through that centre, but is terminated 

 that is, does not extend beyond that 

 side of the plane in which the magnets 

 are situated. 



(277.) The theory of Ampere would 

 lead to the conclusion that no mecha- 

 nical arrangement of the parts of an 

 electro-magnetic apparatus can give 

 rise to rotatory movements, unless fluid 

 conductors form some part of the voltaic 

 circuit ; and accordingly no attempt to 

 obtain practically such movements has 

 ever been successful. 



(278.) It is, accordingly, impossible 



to obtain the revolution of a magnet 

 round its own axis, either by the action 

 of other magnets, or by that of an 

 electric current, which traverses neither 

 the magnet, nor a body thatjis so fixed 

 to it as to move along with it. This is 

 a direct consequence of the law derived 

 from the electro-dynamic theory, that 

 the mutual action of two closed circuits 

 cannot produce in one of these circuits 

 a continued rotatory motion in one con- 

 stant direction ; for it is evident that if 

 this be true with regard to two single 

 currents, it must also be true with re- 

 gard to any assemblages of such closed 

 currents, in whatever way they may be 

 arranged. The utmost that can result 

 is a tendency in one of them, if move- 

 able, to assume a fixed position of equi- 

 librium ; if, therefore, the system be so 

 constituted that it can only revolve 

 round an axis, about which the circuits 

 composing it are symmetrically ar- 

 ranged, it will acquire no motion what- 

 ever by the action either of a single 

 closed circuit, or of an assemblage of 

 such circuits. A magnet susceptible of 

 no other motion than rotation round its 

 axis is in this condition ; and hence, if it 

 derive its magnetic properties from elec- 

 trical currents, it must be impossible to 

 produce in it such a rotation by the 

 action of other magnets. 



(279.) On the other hand, a detached 

 portion of a voltaic circuit moveable on 

 an axis that coincides with that of a mag- 

 net may be set in motion, and made to 

 revolve by the action of the closed cur- 

 rents, in the magnet itself. Thus let 

 V v, Jig. 161, represent a section of a 

 voltaic pile, with its positive and nega- 

 Fig. 161. 



tive wires, W, w, proceeding from its 

 two poles, and inserted into the cups P 

 and N respectively ; the former being 

 placed at the top of an arch of wire, of 

 which the two branches descend on each 

 side, and terminate under the surface of 



