RESEARCHES IN MAGNETISM. 
585 
The same observations are plotted in Plate 62, fig. 32, which shows the loops formed 
in consequence ol hysteresis by the on and off curves of subordinate cycles of stress, 
as well as by those of the main cycle. At each change from loading to unloading, or 
from unloading to loading, the new branch of the curve starts off in a direction tangent, 
or nearly so, to a line parallel to the axis along which stress is plotted. The sub¬ 
ordinate cyclic operations 0 — 5 — 0, 8 —3 —8, 12-6 — 9 — 12*6, and 3 — 8 — 3 give rise to 
magnetic changes which are very nearly cyclic, and consequently produce scarcely any 
effect on the form of the main curves. 
§ 73. Influence of Vibration on the Effects of Stress .—The indications of hysteresis in 
magnetic changes caused by stress disappear almost entirely when we submit the 
piece under test to mechanical vibration either during or after the changes of stress. 
The on and off curves then become nearly coincident. The range of magnetic change 
corresponding to any change of load is considerably increased. A maximum point is 
still found in the curve of magnetism and stress, but it is nearly the same for both 
(on and off) operations, and its position lies, as regards stress, between the positions 
which the maximums occupy when the operations of loading and unloading are 
performed without vibration. These results will be seen in the following observations, 
made on the same wire as the foreg oing, and without any change of conditions, except 
that after each change of load the wire was vigorously tapped before the magnetometer 
reading was taken. 
Stretched Iron Wire with Vibration, Plate 62, fig. 33. 
Load. 
Magnetometer 
(after tapping). 
0 
181 
1 
237 
2 
290 
3 
348 
4 
384 
5 
415 
6 
434 
7 
447 
8 
451-5 
9 
451 
10 
447 
11 
442 
11-8 
436-5 
12-6 
431 
Loading . 
Unloading . 
Load. 
12-6 
11-8 
11 
10 
9 
8 
7 
6 
5 
4 
3 
2 
1 
0 
Magnetometer 
(after tapping). 
431 
434 
435 
440 
443 
445 
444 
429 
416 
389 
353 
306 
250 
192 
The same observations are shown graphically in Plate 62, fig. 33. The on and off 
curves are not quite coincident: they still exhibit some hysteresis, but far less than 
when there is no vibration. 
§ 74. An instructive method of studying the influence of vibration in destroying, or 
rather in greatly reducing hysteresis, is to partially load or unload in the ordinary 
way, and then, pausing at any stage in the process, to tap the wire. An immediate, 
