582 
PROFESSOR J. A. EWING ON EXPERIMENTAL 
formed the magnetising force.'” More usually magnetising force was applied by means 
of a solenoid which completely enclosed the wire : the strength of the current in the 
solenoid was measured by a mirror galvanometer, and was generally varied by means 
of the liquid slide described in § 18. 
§ 70. The study of the effects of stress on the magnetic quality of a material consists 
in observing relations between the intensity of magnetisation and the two independent 
variables, magnetising field and stress. The effect of stress on residual magnetism is 
a special case, namely, the case of field =0. We may in all cases use one or other of 
two methods of inquiry : we may keep the field constant and vary the stress, or we 
may keep the stress constant and vary the field. The former method has been used 
by Sir William Thomson in a series of researches to which our present knowledge of 
this subject is principally due.t I have employed both methods, and where I have 
used the former plan I have observed not merely the resultant change of magnetism 
caused by the application or removal of a considerable amount of load, but the inter¬ 
mediate values of the magnetism when the total change of load is divided into a large 
number of steps. 
The presence of hysteresis in all changes of magnetism which are caused either by 
change of field under constant load, or change of load under constant field, makes the 
phenomena to be examined exceedingly complex, and it is only when the results 
obtained by both methods are compared that we can obtain a moderately clear view of 
the subject. 
Effects of Stress on Magnetism induced by the Earth’s Vertical Force. 
§71. An iron wire (of the same quality as that tested in §21), 0‘079 centim. in 
diameter and 29 centims. long, was hung in position and annealed there by the flame 
of a spirit lamp. Load was applied by stringing on discs weighing 1 kilo, each, and the 
magnetic changes (the inducing field being the earth’s vertical force) were observed. 
The first application of load caused magnetic effects of a very complex character; and 
I shall reserve the account of them. But once the wire was somewhat stretched by 
a load which exceeded its limit of elasticity, the subsequent behaviour, when the load 
was removed and reapplied, was dearly marked and perfectly regular. Earlier and 
later experiments agree in showing that this behaviour, which will now be described, 
is thoroughly characteristic of stretched iron. 
A total weight of 12'6 kilos, was applied, under which the wire stretched about 
07 centim. Its section was then 0'48 sq. mm., and consequently every kilogramme 
of load caused a stress whose intensity was 2 - 08 kilos, per square millim. The load 
of 12'6 kilos, was applied and removed four times, during which the changes of 
magnetism gradually became nearly cyclic. Then, during the fifth application and 
* The value of this in Tokio, Japan, where the experiments were made, is 0’34 c.g.s. units, 
t “ Electrodynamic Qualities of Metals,” Phil. Trans., 1856, 1875, 1879. 
