544 
PROFESSOR J. A. EWING ON EXPERIMENTAL 
by removing and reapplying the same force. This force was insufficient to make the 
cast-iron approach magnetic saturation. The numerical values of <§ and 93 are given 
below : the initial value of © could not of course be determined, and it is assumed to 
have been zero. [Owing to an omission in the laboratory note-book some uncertainty 
attaches to the reduction of the magnetism of this ring to absolute measure. The 
example serves, however, to show the presence of hysteresis in the magnetisation of 
this material.] 
Cast-iron Ring, Plate 58, fig. 7. 
SB. 
SB. 
SB. 
0 
0 
- 0-73 
2450 
0-67 
-2100 
0-07 
3 
- 0-95 
2420 
0-95 
-2050 
0-20 
12 
- 1-26 
2380 
1-50 
-1960 
0-38 
24 
- 1-87 
2290 
2-39 
-1790 
0-73 
48 
- 2-37 
2200 
3-27 
-1530 
1-13 
76 
- 3-27 
2010 
3-94 
-1390 
1-87 
145 
- 3-94 
1860 
4-76 
-1080 
3-27 
344 
- 4-76 
1600 
5-77 
- 620 
5 - 54 
864 
- 5-74 
1210 
7-34 
320 
7-34 
1490 
- 7-34 
360 
8-86 
1350 
9-94 
2360 
- 8-86 
- 660 
11-28 
2390 
1315 
3170 
-11-28 
-1990 
1315 
2895 
1575 
3680 
-13-15 
-2670 
15-75 
3450 
13-15 
3560 
-15-71 
-3420 
13-15 
3330 
9-81 
3390 
-13-15 
-3300 
9-94 
3150 
7-34 
3200 
- 9-94 
-3200 
7-34 
2980 
5-41 
3080 
- 7-34 
-2940 
5-41 
2830 
3-27 
2900 
- 5-41 
-2790 
3-27 
2640 
5-83 
2760 
- 3-27 
-2590 
1-83 
2500 
1-13 
2680 
- 1-87 
-2430 
113 
2410 
0-73 
2640 
- 1-26 
-2340 
0 
2280 
0 
2550 
- 0-73 
-2320 
1-13 
2340 
0 
-2200 
1-83 
2370 
3-27 
2460 
5-41 
2590 
7'34 
2710 
9-93 
2920 
13-13 
3190 
15-71 
3460 
§ 25. Magnetisation of Steel Wire in the hard-drawn, annealed, and glass-hard states. 
—The next three experiments were made on a piece of steel wire 0T37 centirn. m 
diameter, and 19 '7 centims. or 144 diameters in length. This length would have been 
insufficient to allow of a fair experiment if the material had been soft iron, but steel is 
so much less susceptible that the effect of the ends is less important, and a condition 
approaching that of endlessness is reached with a shorter length than would be admis¬ 
sible in the case of iron. In all three experiments the step-by-step ballistic method 
was used. 
The wire was first tested when in what may be called its normal temper, the hard- 
drawn state in which it was commercially supplied. Plate 58, fig. 8, shows its behaviour 
when a magnetising force of 57'5 units was applied, reversed, reapplied, removed and 
