1883.] Theory of Magnetism based on New Researches. 193 



the available total amount of rotation. This is so evident from a 

 series of experiments I have made on this subject, that I can safely 

 predict that if an iron wire could be held charged to magnetic satura- 

 tion, we should obtain no conduction whatever, and its action as 

 regards the production of induced currents would be similar to copper ; 

 or, if the rotation were already complete and rigid, we should have no 

 magnetic conductivity. We can observe this partial rotation in an 

 iron rod, whose conductivity has been reduced by magnetising, as we 

 have only to vibrate the wire, its molecules being rendered free 

 instantly return to zero, and we have its previous full conducting 

 powers. 



The slightest torsion also reduces its conductivity,, the greatest 

 effect being on the first few degrees of torsion ; thus torsion not only 

 rotates the molecules during its elastic stage, but holds them im- 

 prisoned or fixed as rigidly as in hard iron or steel. 



This form of induction balance allows us to demonstrate that 

 polarity apart from the molecules does not exist. For if we balance 

 one pole of a long magnetised iron or steel bar, we find the same con- 

 ductivity for either pole, and the induced currents obtained are all in 

 one direction, no matter if the coils act upon either pole or the 

 apparent neutral centre-; thus it is impossible with) this form of 

 balance to perceive any difference between north or south polarity. 

 All that we observe is the degree of rotation of the inherent polarised 

 molecules, and that this is symmetrical throughout is shown by the 

 equal conducting power of all parts of a magnetised^ rod. 



This form of induction balance also shows that the conducting 

 powers of a bundle of fine iron wires and thin flat strips are equal, 

 the thin band or strip of iron being superior to the fine wires if they 

 are closely pressed together, as the maximum of induced rotative 

 effect in all magnetic bodies is at or near the surface, consequently, a 

 tube of iron will give greater power for a feeble force than a solid bar. 



The time of discharge during which the molecules rotate or return 

 to zero is comparatively short in bundles of loose iron wires, flat 

 strips, and thin tubes, but exceedingly slow in solid bars. This con- 

 firms the results obtained by other methods, and which have already 

 received practical application. 



Visible Effects of Magnetism. 



I have been able to repeat the greater portion of my researches 

 with the induction balance, by simply observing the effects produced 

 by magnetising iron and steel upon a magnetic direction needle 

 according to the indications already given by the balance ; by this 

 means we are enabled to render visible most of the effects, but cannot 

 so well determine the cause ; in fact, without the aid of the induction 

 balance all researches upon magnetism must remain incomplete. We 



vol. xxxv. 



