A POCKET GALVANOMETER. 



281 



By altering the point.-^ of support of the springs (as in the time 

 rating of a watch) we may, if it is required, adjust the instrument so 

 as to make one division correspond to some simple fraction of an 

 ampère, say .5 or .2 



Field of Force and Arrangement of Magnets. 



In studying the field of force due to four magnets arranged as in 

 Fig. 4, and a single straight conductor parallel to their axes, we have 

 only to consider the action of a single straight conductor upon one set 

 of four poles which lie in a plane perpendicular to the conductor, 

 since the other set of poles is an exact counterpart of the one con- 

 sidered. As proved in Maxwell, the electromagnetic force at an 

 external point of a straight cylindrical conductor of infinite length 

 depends only upon its distance from the center of the section, for 

 any concentric distrihution of current ; and since the action be- 

 tween a magnet and a current is mutual, if a compound cylindrical 

 magnet consists of concentric layers of uniform intensities, its 

 action upon an externally placed current running parallel to the 

 axis of the magnet, must be reducible to the action of a single 

 equivalent pole at its centre, whatever be the law of distribution 

 of magnetism from layer to layer. As we make our magnets of 

 cylindrical wire, and as most conductors are cylindrical, we may 

 safely reduce the action to the centers of the sections of the magnets 

 and the conductor, neglecting the pole-shifting effect due to the in- 

 duction of magnets on each other. 



Let 2a be the distance between like poles, and 2b that between 

 unlike ones : and put r for ^/c(P + b^. 



Let X, y be the co-ordinates of any point referred to the center 

 of the i-ectangle [2a, 2/>], measured j^arallel to a and /> re- 

 spectively. 



