90 Prof. Faraday on some Points of Magnetic Philosophy. 



tube, will behave like bismuth in air but like iron in water. 

 Now these are precisely the actions which have been attributed 

 to polarity, and by which the assumed reverse polarities of para- 

 magnetic and diamagnetic bodies have been considered as esta- 

 blished ; but when examined, how will ideas of polarity apply to 

 these cases, or they to it ? The solution / points and acts like 

 bismuth in air and like iron in water ; are we then to conclude 

 that it has reverse polarity in these cases ? and if so, what are 

 the reasons and causes for such a singular contrast in that which 

 must be considered as dependent upon its internal or molecular 

 state ? 



3314. In the first place, no w r ant of magnetic continuity of 

 parts can have anything to do with the inversion of the pheno- 

 mena ; for it has been shown sufficiently by former experiments*, 

 that such solutions are as magnetically continuous in character 

 as iron itself. 



3315. In the next place, I think it is impossible to say that 

 the medium interposed between the magnet and the suspended 

 cylinder of fluid can cut off, or in any way affect the direct force 

 of the former on the latter, so as to change the direction of its 

 internal polarity. Let the tube be filled with the solution m, 

 then if it be surrounded by the solution /, it will point as iron; 

 if the stronger solution n surround it, it will point as bismuth; 

 and with sufficient care a succession of these fluids may be ar- 

 ranged as indicated in figs. 2, 3, where the outlines between 

 the poles represent the forms of thin glass troughs, and the let- 

 ters the solutions in them. In fig. 2 we see 

 that the action on m is the same as that on 

 m', and the pointing of the two portions is the 

 same, i. e. equatorial; neither has the action 

 on m been altered by the power of the poles 

 having to traverse n, m 1 and n' ; and in fig. 3 

 we see, that, under like circumstances of the 

 power, m' points as bismuth and m as iron, 

 though they are the same solution with each 

 other and with the former m ml solutions. No 

 cutting off of power by the media could cause 

 these changes; — repetitions of position in the 

 first case, and inversions in the second. All 

 that could be expected from any such intercep- 

 tions would be perhaps diminutions of action, 

 but not inversions of polarity ; and every con- 

 sideration indicates that all the portions of these 

 solutions in the field at once have like polarity, 

 i.e. like direction of force thro ugh them,and likeinternal condition; 



* Phil. Mag. 1846, vol. xxix. p. 254. 





