MAGNETIZATION uF XICKKL. 



287 



eyliiiik'r. the roil wns fixed between the ^"^i nnd Ixjred on a latlie l)y 

 turning it tooether with the cylinder. A small clamping screw served 

 to j)in the wire fast against tiie side nf the rod, so tliat tlie wire and 

 cylinder rotated together. 



It might at first sight appear that this arrangement might prevent 

 the longitudinal stress being applied uniforndy for various angles of 

 twist, because of the friction of the rod against the ^"s. To test this 

 point, the win' was fixed by the screw, and the longitudinal pull 

 applied by known loads iuuig on below. The upper end of the wire 

 Avas fastened to a spring balance, by means of which any > ariations 

 of stress could at once be detected. Witli a given load, the wire was 

 twisted though various angles ; but scarcely any sensible variation of 

 longitudinal stress was indicated. 



The magnetic experiments were conducted in the following 

 manner. At first, a constant current was made to pass through the 

 maunetizing coil, ni\d the magnetometer zero was determined. The 

 wire was then placed in position, and was first twisted through 180° 

 in what we call the positive direction — although it may as well be 

 stated once for all that the [lositive direction mean the jir.sl chosen 

 direction, whether that is, so t» speak, with the magnetizing I'urrent 

 or against it. Then after a complete revolution in the opposite direc- 

 tion, it was Ijrought back to its original position. This process was 

 repeated till the changes became nearly cyclic. Then at e\ery succes- 

 sive 20° of twist, the deflection of the magnetometer magnet as given 

 by the scale reading was taken and noted, the reading being (jbser\ed 

 by means of a telescope. Each complete set of experiments was 

 made in a cuiistaiii magnetic field, while the wire was subjected to 

 gradually increasing longitudinal stresses. The results tlius obtained 

 are mven in C. G. S. electromagnetic units, thouuh such reduction 

 Would have been unnecessary in experiments of this kind. The 



