456 Prof. Beetz on the Influence of Magnetization 



Wire No. 7, when magnetized, vibrated so strongly that a 

 qualitative determination of the contraction was not possible. 

 The annealed wire also vibrated strongly. The tension 20000 

 grms. was produced by a lever which was fastened at one end of 

 the wire. The results obtained are easily accounted for if we 

 assume that the stretching separated the molecules of the mag- 

 net, while the magnetization brought them together. If the 

 tension is too great, the contraction is only produced with 

 difficulty, and may not occur at all. The different contractions 

 which wire 8 undergoes, in the hard and in the soft condition, 

 cannot possibly be explained from the different ductility of the 

 wire ; it is rather to be supposed that the wire was bent, and 

 that the stretching weight, 1000 grms., could straighten the soft 

 but not the hard wire, that in this case a lengthening was opposed 

 to the contraction. It may be supposed that this was also the 

 case with the wire 9, where in the last case the covering was re- 

 moved, not by heating, but by dipping in concentrated sulphuric 

 acid. With a tension of 300 grms. the covered wire is still 

 stretched (partly, no doubt, because the coating bears a good 

 deal of the stretching weight), the uncovered but hard wire 

 remains unchanged; the soft is even more strongly stretched 

 than is necessary for its maximum contraction. 



I obtained entirely similar results on placing the apparatus of 

 observation in a vertical position. The iron wires were soldered 

 to a quadrangular piece of brass which moved up and down in a 

 brass sheath, and could be fastened by a screw. The magnetizing 

 spirals were fastened on consoles in the wall. At the end of 

 the wire was a second quadrangular piece which passed through 

 a guide. To this piece of brass was attached the thread which 

 was coiled about the axis of the mirror and supported the stretch- 

 ing weight. The distance of the scale from the mirror was 

 4207 millims. ; so that a displacement on the scale amounting to 

 one division corresponded to an increase in the length of the 

 wire of 0*00006658 millim. This vertical arrangement of the 

 apparatus enables the wire to be very accurately adjusted on the 

 spiral axis; and moreover the guiding at the lower end is a 

 guarantee against the action of the twisting of the wires. The 

 screw at the top makes it easier to find the image of the scale in 

 the mirror, and to fix its zero-point upon the cross wires. The 

 displacements of the scale found were the following : — 



