ON THE MEASUREMENT OF MAGNETIC HYSTERESIS. 
81 
Load, 
gnus. 
Ho = 
15-0. 
Ho = 
10-0. 
Ho = 
7-0. 
Ho = 
5-0. 
Ho = 
= 4-0. 
11 
O 
3-0. 
lb- 
M'. 
B, 
tY. 
Bo- 
Bo- 
tv. 
Bo- 
tv. 
\ Bo. 
1 
t\7 
i 
0 
14680 
17470 
13210 
13780 
11200 
9970 
7400 
5180 
5100 
3140 
2280 
907 
50 
14420 
17560 
12920 
13900 
10940 
9990 
7540 
5440 
4750 
2860 
1980 
766 
100 
13730 
17080 
12220 
13440 
10310 
9760 
6910 
4920 
4110 
2400 
1520 
484 
150 
13120 
16370 
11540 
12720 
9640 
9020 
6120 
4250 
3440 
1840 
1350 
389 
200 
12570 
15610 
10800 
11930 
8820 
8180 
5550 
3760 
3100 
1630 
1220 
336 
250 
12000 
14900 
10240 
11280 
8300 
7700 
5170 
3420 
1 2760 
1380 
1140 
302 
300 
11530 
14320 
9740 
10630 
7790 
7180 
4860 
3135 
■ 2540 
1242 
1070 
276 
250 
11880 
14680 
10190 
11190 
8150 
7460 
5070 
3260 
: 2550 
1265 
1120 
286 
200 
12320 
15330 
10710 
11830 
8700 
7940 
i 5390 
3443 
2950 
1480 
1220 
329 
150 
12970 
16130 
11280 
12480 
9350 
8600 
5870 
3950 
3260 
1700 
1320 
367 
100 
13610 
16850 
11930 
13150 
9960 
9210 
6460 
4405 
3660 
1980 
1460 
443 
50 
14300 
17400 
12730 
13770 
10700 
9660 
7250 
5130 
4040 
2290 
1650 
563 
0 
11670 
17350 
13200 
13660 
11170 
9770 
7680 
5350 
4740 
2745 
2230 
911 
In each cycle of torsion l)otli the mean maximum induction and the hysteresis 
W are greater when the torsion is increasing than when it is diminishing, except for 
small parts of those curves in which want of symmetry about the line of zero 
torsion has caused the crossing-point of the two branches of the curve to lie off 
tliat line. 
For the smaller fields, both and W are very sensitive to torsion. Thus, when 
Hq = 3, a load of 100 grammes liung from the edge of the wheel diminished Bq by 
about one-third and W by about one-half of their values for zero load. 
The dynamometer was not sensitive enough to allow us to continue the observa¬ 
tions for W for fields less than Hq = 3. We could have measured Bq for much 
smaller fields, but the observations would not have been of much interest in the 
absence of the observations for W. 
In the case of the smaller magnetic fields we noticed that the hysteresis continued 
to diminish consideraldy with continned reversals of the magnetic force even after it 
had been subjected to many reversals. 
The table of § 57, and the top and Ijottom rows of the table just above, give the 
values of Bq and W for zero torsion for various values of Hq. The values of Bq and 
W in the first table do not agree very closely with those in the second talde, but it 
must be noticed that the first table was made after the wire was annealed and before 
it was strained in any way. After the observations of § 57 had been made, experi¬ 
ments were made to find the limits of elasticity of the wire, and in this process 
the wire was subjected to torsions lai'ge enough to give it permanent set; to get rid 
as far as possible of the effects of this overstraining, the wire was re-annealed before 
the tests for the second table were made. Close agreement betrveen the two tables 
is in consequence not to be expected. 
These experiments may serve to show the saving of time effected by our metliod of 
vox,. CXCVIII.—A. M 
