1884.] 



On Magnetic Polarity and Neutrality. 



407 



with high, powers the exterior soon arrives at its saturation, the 

 distant layers rise in value, and also if the bar is not homogeneous 

 there is a consequent deformation, owing to the comparative rigidity 

 of its molecules. 



In all cases, whatever the force employed, or nature of the iron or 

 steel, there are no reversals of polarity in the interior, but a constant 

 diminishing curve of penetration from the outside to the centre. 

 This changes, however, the instant the exterior polarising force ceases, 

 the different degrees of force between the external and internal react 

 upon each other, producing the following results : — 



Internal Waves of Opposite Polarity. 



All varieties of iron and steel have a high magnetic capacity 

 whilst under the influence of its inducing force, such as the electro- 

 magnetic coils, or strong permanent magnets, but this power in a 

 great measure disappears on the cessation of the inducing influence, 

 a return more or less perfect towards neutrality being the result ; 

 remaining magnetism is therefore a partial neutrality, more perfect in 

 soft iron, where the molecules are in a greater state of freedom, than 

 in comparatively rigid cast steel. Our so-called permanent magnets 

 are simply the remains of a far higher magnetic state, and it is 

 already in most cases half-way down on its road to neutrality. 



It is absolutely necessary in a theory of magnetism that we should 

 know the cause of neutrality, for it is really the starting point to 

 appreciate how polarity becomes evident. In my previous researches 

 upon neutrality I used the induction balance, but in these I have 

 employed more simple methods, which allow of repetition by the most 

 simple means. 



The first consists in forming compound bars of ordinary hoop- 

 iron, \ millim. thick, and 'SO or more centims. long, twenty or more 

 of which could be superposed, bound together by a fine copper wire 

 and forming a rod of any desired thickness ; they were magnetised 

 by drawing over magnets of various powers, and the degree of 

 approach to neutrality observed by the amount of its remaining 

 magnetism. Now, on carefully separating them, there were invariably 

 found violent curves of opposing magnetism, previously held bound 

 by the closed circuit of mutual attractions. 



The second method consists in superposing the divided concentric 

 tubes, already mentioned, bound together by a fine copper wire, and 

 magnetising them in the electro-magnetic coils of the measuring 

 balance ; by this means we could observe the charge or full magnetic 

 •capacity under the influence of an electric current, the remaining 

 magnetism upon its cessation, and after taking out the tubular core, 

 ■separate it, and observe the polarity of its successive internal layers. 

 This method is objectionable, as the slightest rubbing of one tubular 



