Prof. Faraday on some Points of Magnetic Philosophy. 87 



tion to each other*. Becquerel considers that polarity may in. 

 certain cases occur transverse to the length, and so produce 

 results which others explain by reverse polarity. The views of 

 very many parties always include the idea of the source of the 

 polar action, whether that be supposed to depend on the accu- 

 mulation of magnetic fluids at the chief poles of the dominant 

 magnet, or the action of electric currents in a determinate posi- 

 tion around its molecules ; and such views are adhered to even 

 when the polarity induced is of the reverse kind, as in bismuth, 

 &c, to that of the inducing magnet. Others, like Weber, add to 

 Ampere's hypothesis an idea of electricity, loose as regards the 

 particles, though inseparably associated with the mass of the 

 body under induction. Some, I think, make the polarity not 

 altogether dependent upon the dominant magnet, but upon the 

 neighboimng or surrounding substances ; and I propose, if the 

 physical lines of force should hereafter be justified, to make that 

 which is commonly called polarity, in distinction from the true 

 polarity (3307.), dependent upon the better or worse magneto- 

 conduction power of the substances presenting the usual polar 

 phsenomena (2818.). 



3309. The views of polar action and of magnetism itself, as 

 formerly entertained, have been powerfully agitated by the dis- 

 covery of diamagnetisni. I was soon driven from my first sup- 

 position, that the N pole of a magnet induced like or N polarity 

 in the near part of a piece of bismuth or phosphorus ; but as 

 that view has been sustained by very eminent men, who tie up 

 with it the existence of magnetic fluids or closed electric currents 

 as the source of magnetic power, it claims continued exami- 

 nation, for it will most likely be a touchstone and developer of 

 real scientific truth, whichever way the arguments may prevail. 

 To me the idea appears to involve, if not magnetic impossibi- 

 lities, at least great contradiction and much confusion, some of 

 which I proceed to state, but only with the desire of elucidating 

 the general subject. 



3310. If an ordinary magnet M, fig. 1, Fig. 1 . 

 acting upon a piece of iron or other para- i 

 magnetic matter I, renders it polar by M 103 c 



throwing its near end into the contrary or I* n \ ' ' 



S state in the manner usually understood, b 



and, acting upon a like piece of diamag- 



netic matter as bismuth B, renders it also polar, but with the near 

 end in the same state; then B and I are for the time two mag- 

 nets, and must act back upon the magnet M ; or if they could 

 be made able to retain their states after M is removed (and that 

 is the case with 1), would act as magnets upon a third piece of 

 * Cours special sur V induction, Sfc, p. 201 . 



