278 On the Magnetic Polarity of Metallic Bodies. 



exciting magnet, they would reciprocally attract each other ; 

 consequently, in whatever direction the one was made to re- 

 volve, the other would follow after it. 



This, I believe, was the explanation generally advanced, 

 and it is perhaps, to a certain extent, very correct : but from 

 the peculiar mode of experimenting which I had devised and 

 pursued, I had an opportunity of observing certain phasno- 

 mena,which I considered that hypothesis inadequate to explain. 

 I was, however, for a long time perplexed with irregularities , 

 in their exhibition, and unable to reconcile them to any deter- 

 minate law. Nor was it till I was pursuing my inquiries con- 

 cerning the thermo-viagnetism of simple metals, in 1830*, that 

 I could trace the phaenomena I had observed to anything 

 like uniformity: but by pursuing a hint afforded by the dis- 

 covery of the curious distribution of the force excited by heat, 

 which actuates the needle, on the flat surfaces of simple me- 

 tals, I became enabled to trace a similar, though distinct, dis- 

 tribution of force in thin discs of copper, zinc, &c. when 

 under the influence of powerful magnetic poles. 



I was first prompted to inquire into the distribution of 

 magnetic polarity in discs of thin copper, by a strange discord- 

 ance of results, which were obtained by vibrating it within the 

 influence of magnetic poles variously arranged with regard 

 to its surface. Fig. 9. is a representation of the apparatus 

 which I employed, and which first showed the anomalies in 

 question. B B is a rectangular mahogany board, on which 

 are erected the two brass pillars 77 j>, one of which supports 

 the horse-shoe magnet M. The other carries a pair of pa- 

 rallel projecting arms, on the extremities of which the axis of 

 a copper disc c is supported. To the lower edge of the disc 

 is attached a weight or bob to give it a vibrating tendency ; 

 so that by this arrangement the disc can be made to vibrate 

 freely in its own plane between the arms of the apparatus. 

 A quadranlal screen s e, of thin brass, is attached to the front 

 arm, and consequently a quarter of the disc is hid behind it. 

 When at rest, and in the position represented by fig. 9, the 

 quarter behind the screen is divided equally by a radial line 

 drawn on its surface. 



When an experiment is to be made with this instrument, the 

 bob is first to be brought to the point of the arms, where there 

 is a contrivance for retaining it in that position, or releasing 



* Mj' paper on the thermo-magnetism of simple metals was not pub- 

 lished till July and August 1831 (in the Phil. Mag. and Annals for those 

 months): but had it not been for certain impediments which happened to 

 be thrown inthe way, it would have been published much sooner; and this 

 paper, which was then intended to follow immediately after the former, 

 would have been published in an early part of last summer. 



it 



