614 
PROFESSOR J. A. EWING ON EXPERIMENTAL 
Galvanometer 
readings. 
Magnetometer readings. 
T. 
After 
demagnetisation 
with no load. 
II. 
After the cycle 
0 — 18J — 0. 
III. 
After the cycle 
0 US 1-0 
and then vibration. 
0 
0 
0 
0 
0 
20 
115 
5 
8 
5 
35 
2-01 
11 
19 
10 
50 
2-87 
19 
40 
17 
75 
4-31 
44 
73 
35 
100 
5'75 
78'5 
110 
70 
150 
8-62 
149 
176 
150 
200 
11-50 
212-5 
230 
214 
250 
14-37 
267 
278 
268 
300 
17-25 
3l4 - 5 
321 
314 
350 
20-12 
355 
3585 
354 
400 
23-00 
390 
394 
388 
450 
25-87 
420 
420 
422 
576 
3312 
472 
472 
471 
400 
23-00 
445 
441 
443 
300 
17-25 
412 
407 
409 
200 
11-50 
355 
351 
352-5 
100 
5-75 
278 
275’5 
276 
50 
2-87 
234 
232 
232 
0 
0 
183-5 
183 
182 
The same experiment is shown in the curves of Plate 65, tig. 49, where the dotted 
curve is No. III., taken after applying the cycle 0 —18'5 — 0 to the previously de¬ 
magnetised wire, and then vibrating it before magnetising, the full line (I.) is the 
curve taken after demagnetisation with no load, and the broken line (II.) is the curve 
taken after the cycle of loads 0 —18^ — 0. No vibration occurred during the taking 
of the curve. 
This experiment (which was confirmed by many others like it) shows that the appli¬ 
cation and removal of load, after the wire is demagnetised, increases its susceptibility 
to subsequent magnetisation, but that this increase of susceptibility is removed, and 
the original condition approximately restored, by vibrating the wire after the applica¬ 
tion and removal of load, and before beginning to magnetise. Curve No. III., nearly 
coincides with No. I., proving that the residual effect of the previous loading and 
unloading, which had raised the susceptibility to the level shown by Curve II., was 
neutralised in great part by mechanical disturbance. 
§ 98. Now it seems from this, that if we apply and remove stress in a wire whose 
magnetic state is entirely neutral, we cause some kind of molecular displacement in the 
relation of which to the applied stress there is hysteresis. This molecular displacement, 
whose precise character has yet to be determined, is the cause of the changes of 
magnetic susceptibility which accompany changes of stress. 
When a load is applied and removed the molecular displacement lags behind the 
changes of stress, and we therefore find on subsequently magnetising the wire that 
the susceptibility has been changed by the process—that its value is more nearly 
