502 The Rev. T. R. Robinson's Experimental Researches on the 



larities towards the extremities : there is scarcely any free magnetism, and the 

 internal molecular excitation must be of quite a different character. The state 

 of things is probably of this sort. 



Let ABCD represent the magnet, supposed continuous ; let it be excited 



by a circular current at a. The action of this on the dif- a b 



ferential slice lying in its plane will induce opposite polari- 

 ties on its surfaces, till their natural attraction, aided by the 

 coercive force of the molecules, balances that action. But 

 these polarities induce opposite ones on the adjacent surfaces 

 of the slices i and d, — less, however, in intensity; both because 

 their coercive force resists the inducing force, and because 

 it is weakened by the intermolecular space. These, again, 

 induce on c and e, which are also affected by a, though more 

 weakly ; and so on through the whole circuit. These in- c d 



duced polarities react on a itself : its + no longer tends to its — with the same 

 force, for it is drawn the other way by the combined effects of the — ones 

 b, c, &c. Therefore, the current is able to produce a higher polarization of a than 

 if it were alone, which again acts more powerfully on its neighbours ; so that a 

 far greater development of magnetic force is obtained, than would be possible 

 if the magnetic circuit were interrupted, and the magnet were a mere bar 

 covered with spires.(a) 



When the equilibrium of these forces is obtained, it is evident that the 

 molecular polarity must be greatest at a, and decrease symmetrically on each 

 side, till it becomes a minimum at a, the opposite point of the circuit. The 

 magnitude of the maximum is independent of the place of a, but must be a 

 function of c, the length of the circuit, such that it decreases as c increases, 

 but will continue finite when that is infinite. At any other point b the 

 polarization depends on the maximum, conjointly with the distance ab, and also 

 with c.{b) 



If instead of one circular current a spiral or helix be used, the action is 

 the same, though the distribution of the magnetism is somewhat altered. 



At any section A the magnetism will be the resultant of the force trans- 

 mitted by induction through the branch Aa — z, and that through the comple- 

 mentary branch aCDBA = c — z. 



