58 6 
PROFESSOR J. A. EWING ON EXPERIMENTAL 
often very large, change of magnetisation is the result. If the operation of loading 
or unloading be then resumed, hysteresis again becomes conspicuous. 
The following observations form a good example of this method of procedure. They 
are plotted in Plate 62, fig. 34, where the full lines refer to the parts of the operation 
during which there was no vibration, and the lines drawn thus-show the 
changes of magnetism produced by tapping. The wire was the same as that of 
§§ 71-73, and in the same conditions. A cycle of loading without tapping was gone 
through first. This brought the magnetometer reading, with no load, to 252. On 
tapping the wire this fell to 180. Loading was then begun, and carried on up to 5 
kilos, without tapping. The wire was then tapped, when the magnetometer reading 
changed from 337 to 405. Loading was then resumed and carried on to 8 kilos, 
and the wire again tapped—and so on, as will be clear by inspection of the table:— 
Stretched Iron Wire, with and without Vibration, Plate 62, fig. 34. 
Load. 
Magnetometer. 
Load. 
Magnetometer. 
0 
252 
12-6 
423 after tapping 
0 
180 
after tapping 
11-8 
425'") 
1 
19E') 
11 
429 | 
2 
213 
10 
433 without tapping 
3 
251 
f without tapping 
9 
437 | 
4 
294 
8 
441J 
5 
337J 
8 
450 after tapping 
5 
405 
after tapping 
7 
452 'l 
6 
4110 
6 
451 ^without tapping 
7 
425 
> with out tapping 
5 
443J 
8 
440 J 
5 
425 after tapping 
8 
453 
after tapping 
4 
419^1 
9 
4520 
3 
400 | 
10 
451 
2 
361 ^-without tapping 
11 
447 
without tapping 
1 
322 
11-8 
442 
0 
265J 
12-6 
436^ 
0 
181 after tapping 
The agreement of points on the on and off curves, with tapping, is scarcely so good 
as in the experiment described in the last paragraph, probably because the vibration 
was scarcely vigorous enough. It is very interesting to notice how hysteresis re¬ 
appears at each continuation of the loading or unloading, after the vibration has 
ceased, manifesting itself by the way in which each curve corresponding to such a 
continuation starts off tangent, or nearly tangent, to a line parallel to the axis of 
loads. Finally, in the descending limb of the off curve, the accumulated effect of 
hysteresis during the reduction of the load from 5 kilos, to 0 leaves the magneto¬ 
meter reading higher by more than 80 divisions than the value which is reached 
by vibration at the zero of load. Nothing could illustrate better than this the manner 
in which hysteresis causes a diminution of the total range through which magnetism 
changes under varying stress, as compared with the range which we find when the 
variations of stress are accompanied by vibration. 
