the Laws of Magnetic Forces. 287 



s, but repels n ; and S attracts n, but repels * ; so that the final 

 resultant is very complicated. We may, however, imagine these 

 forces to be so circumstanced in relation to a means by which 

 their action is evinced, and by which they are measured, that at 

 some distance N*, Fig. 6. the action vanishes. Let then the line 

 CD represent the limit at which their influence, thus estimated, 

 ceases : in this case, the effect of the polarity of BN upon that 

 of An must be considered as having no assignable value, until 

 some point in An, Fig. 8., upon the other side of the magnetic 

 centre A passes the limit CD. The same may be said of the in- 

 fluence of the polarity of BS upon that of A*, so long as the 

 points in BS remain without the limit CD ; that is to say, at a 

 distance from the points in A* greater than N*. If the magnets 

 be only hardened and magnetised about their extremities, or if 

 they be small, and of weak intensity, then there may arise a case in 

 which the action is so weak in every other part except the extre- 

 mities, that the result is not sensibly deranged until the pole n 

 actually arrives at CD, Fig. 8. There are some further conside- 

 rations as to the limit CD, not necessary here, which will hereaf- 

 ter be given. 



18. Now, in the experiment under examination, the masses of 

 iron ae and cb, Fig. 5, during the time they are operated on by 

 induction, may be considered as two magnets whose intensities 

 increase at each approximation of the bar m. It is, therefore, 

 only necessary to determine the limit CD, Fig. 6, of their action, 

 when the induced magnetic force is the greatest ; and we imme- 

 diately ascertain if any disturbance arises from the influence of 

 the opposite polarities. This limit, in the present case, was 

 found not to exceed an inch and a half ; and it not being requi- 

 site to approximate the distant poles within that space, the re- 

 sult might so far be considered free from this source of error. 



