584 
PROFESSOR J. A. EWING ON EXPERIMENTAL 
§ 72. It is evident as a further result of hysteresis, that the relation of load to 
magnetism may, without any change either in the field or in the initial magnetic 
condition of the piece examined, have any one of the indefinite number of values 
defined by points lying between the on and off curves. For example, it is easy with 
a load of 4 kilos, to reach a value for the magnetometer reading anywhere between 
323 and 413. The value of the magnetism depends not only on the actual load, but 
on the preceding states and changes of the load, and most particularly on the 
immediately preceding changes. To illustrate this more fully the following observa¬ 
tions were made, on the same wire, after it had again been subjected to the cycle of 
loading 0—12 - 6 —0. (That is, after a load of 12‘6 kilos, had been again applied and 
removed.) 
The successive changes of load passed through in this experiment were: 
0 — 5 — 0 — 8 — 3 —12"6 — 9 —12'6 — 3 — 8 — 0, with intermediate steps, and the resulting 
magnetometer readings are given below :— 
Stretched Iron Wire, Plate 62, fig. 32. 
Load. 
Magnetometer. 
Load. 
Magnetometer. 
0 
245 
12-6 
425 
1 
255 
1P8 
427 
1 
2 
270 
Unloading . . .< 
11 
430 
(ah) 
3 
293 
(. f9 ) 
10 
434 
L 
4 
322 
L 
9 
437 
5 
352 
10 
434 
4 
349 
Loading ....*> 
11 
431 
Unloading . 
(be) 
3 
2 
338 
319 
W 
_ 
11-8 
12-6 
428 
425 
1 
l 
1 
290 
11-8 
427 
0 
249 
11 
430 
1 
256 
10 
433 
2 
271 
9 
436 
3 
292 
Unloading . . . - 
8 
439 
Loading . 
4 
321 
(hi) 
7 
441 
(cd) 
5 
351 
6 
440 
6 
384 
5 
433 
7 
413 
4 
415 
8 
430 
3 
388 
7 
430 
4 
392 
LTnloading . 
i 
■ I 
6 
5 
427 
419 
Loading . . . .< 
5 
6 
400 
410 
(de) 
L 
4 
405 
(v) 
7 
423 
3 
381 
«_ 
r 
8 
432 
4 
385 
7 
432 
5 
393 
6 
430 
6 
405 
5 
424 
7 
419 
Unloading . . 
4 
411 
Loadino- . 
■- 
8 
430 
m 
3 
390 
(ef) 
9 
438 
2 
355 
10 
440 
l 
i 
306 
11 
436 
0 
252 
11-8 
431 
After half-an-hour 
0 
252 
12-0 
425 
