54 



Prof. G. Wiedemann's Magnetic. Researches. 



longitudinally, it suffers torsion in a direction opposite to that 

 which I found for an iron bar treated in the same way. 



Maxwell* and Chrystalf have formerly endeavoured to refer 

 this torsion, in the case of iron, to the above-mentioned ex- 

 pansion which it suffers upon being magnetized. Wires 

 magnetized at the same time longitudinally and transversely, 

 and therefore in an obliquely spiral direction, ought to expand 

 in this latter direction, and thus to suffer torsion in the direc- 

 tion determined. But with nickel, since, upon being magne- 

 tized, contraction instead of expansion takes place, the torsion 

 ought to take place in the opposite direction, as is shown by 

 experiment to be the case. 



I have myself endeavoured to refer this phenomenon, ob- 

 served in my first experiments in 1858, at least in the case 

 of iron wires, to the obliquely spiral direction which the mole- 

 cules take up in consequence of the two magnetizations at 

 right angles to each other, and which is accompanied secon- 

 darily by a displacement of the longitudinal fibres and sections 

 of the wires, of which, in the case of iron, the first is the most 

 important. 



The chief proof of my assumption which I had given was 

 the phenomenon, completely coordinate with the above results, 

 that a wire equatorially magnetized, whether temporarily or 

 permanently, by a current passed through it, becomes longi- 

 tudinally magnetized upon torsion. 



I have now extended these experiments to the case of nickel 

 wires. They were made essentially in the same way as the 

 former ones. The wire, 52 centim. long and 2 millim. thick, 

 was stretched east and west between a clamp, a, screwed tight, 

 and a clamp, b, coaxial with a, capable of rotation in an up_ 



right, and provided with a divided circle. The wire was placed 

 end on in front of a magnetic steel mirror swinging in a thick 

 copper box, with the fixed clamp at a distance of about 20 cen- 

 tim, from it. The current of a Bunsen battery was led to this 



* ' Electricity and Magnetism ' (2nd ed.), ii. p. 86. 

 t " Magnetism," Encyclop. Metropol. p. 270. 



