RESEARCHES IN MAGNETISM. 
633 
100° C. are nearly proportional to the changes of temperature,* and exhibit no 
hysteresis with respect to them. The whole range of change, from 17,415 at 14° C. 
to 17,262 at 100° C., is only 0'9 per cent, of the total magnetism. 
§ 117. The actual amount of magnetic change, and even its sign, depends on the 
degree of magnetisation. This follows from the known fact that there is a critical 
value of the magnetising force, or (more probably) of the magnetisation, below which 
heating increases susceptibility, and above which heating reduces susceptibility. 
(Baur, Wied. Ann., xi., 1880.) My own experiments have illustrated this in two 
ways. In some of them the effects have been studied of repeated heatings (up to 
100° C.) and coolings, on various degrees of induced or residual magnetism in an iron 
wire : in others, curves of the relation of 2> to have been drawn when the same 
iron wire is magnetised, (l) at ordinary temperature, (2) at 100° C. Both methods of 
experiment have been applied to samples of iron wire both in the soft annealed and 
hard-drawn states. One or two representative experiments may be quoted. 
§ 118. (Jan. 28, 1883.) An iron wire, 1*2 mm. in diameter and 3 5’7 centims. long, 
was annealed, and then stretched until its length was 40'5 centims. and its diameter 
T16 mm. In this state it was demagnetised by reversals, and then hung (without 
mechanical disturbance) under the influence of the earth’s vertical force, and very 
near a magnetometer whose sensibility was raised to a high degree. The wire was 
surrounded by a tube, round which a magnetising solenoid was wound, and steam at 
100° C. and cold water at 6° C. were alternately passed through the tube while the 
magnetic changes were noted. From time to time, at points which will be indicated 
below, the total magnetisation of the wire was raised a step, by momentarily setting 
up in the solenoid a suitable current, which was then interrupted, so that the 
magnetism dealt with was that part which was retained after the inducing field was 
reduced to the value 0'34, namely, the earth’s vertical force. The following are the 
values of (expressed as usual in absolute c.g.s. units) which were observed through¬ 
out the process. 
3- 
Original condition of the wire . 
. . . T52 
While 
steam was 
passing 
. . . 2-05" 
Here, in addition to a 
33 
water 
33 ... 
. . . 1'98 
progressive shaking 
33 
steam 
33 ... 
. . . 2T1 
out and then shak¬ 
33 
water 
33 ... 
. . . 2-05 
ing in of magnetism, 
33 
steam 
33 ... 
. . . 2T6 
^ there is augmenta¬ 
tion of magnetism 
3 3 
water 
33 ... 
. . . 2-09 
33 
steam 
33 ... 
. . . 2-20 
with steam, arid di¬ 
33 
water 
33 ... 
. . . 2T2 
minution with water. 
33 
steam 
33 ... 
. . . 2-23 
3 is below the critical 
33 
water 
33 ... 
. . . 2 • 16 - 
point. 
* The ratio of change of magnetism to change of temperature increases slightly as the temperature 
rises. 
