524 
PROFESSOR J. A. EWING ON EXPERIMENTAL 
The effect of stress on thermoelectric quality was the first of these subjects to 
engage my attention.* In testing the changes of thermoelectric quality which a 
stretched iron wire underwent when successively loaded and unloaded so as to suffer 
alternate application and removal of tensile stress, I found that during increment and 
decrement of the load equal values of load were associated with widely different values 
of thermoelectric quality; t the difference being mainly of this character, that the 
changes of thermoelectric quality lagged behind the changes of stress. This lagging 
is, however, a static phenomenon, for it is sensibly unaffected by the speed at which the 
load is changed ; and again, when any state of load is maintained constant the thermo¬ 
electric quality does not change with lapse of time. I then proceeded to search for 
other instances of the same kind of lagging, in other physical effects of stress, and also 
in effects caused by the change of other physical conditions besides stress. Magnetic 
phenomena present many instances of a similar action—some of which will be described 
below. Thus, when a magnetised piece of iron is alternately subjected to pull and 
relaxation of pull sufficiently often to make the magnetic changes cyclic, these lag 
behind the changes of stress in much the same way as the changes of thermoelectric 
quality do. And more generally, if a piece of iron, whether magnetised or not, be 
subjected to cyclic stress variations, it can be shown that certain of its physical 
qualities, while varying cyclically in consequence of the changes of stress, exhibit this 
same lagging. I found it convenient to have a name for this peculiar action, and 
accordingly called it Hysteresis (from vcrrepeoj, to lag behind).| Thus, when there are 
two qualities M and N such that cyclic variations of N cause cyclic variations of M, 
then if the changes of M lag behind those of N, we may say that there is hysteresis 
in the relation of M to N. The value of M at any point of the operation depends 
not only on the actual value of N, but on all the preceding changes (and particularly 
on the immediately preceding changes) of N, and by properly manipulating those 
changes, any value of M within more or less wide limits may be found associated with 
a given value of N. 
§ 2. The effects of stress on magnetism and on thermoelectric quality will be 
recurred to at length further on. Meanwhile I shall describe experiments which were 
begun with the view of searching for hysteresis in the relation of magnetisation to 
magnetising force, without reference to stress.§ The presence of this peculiar action 
in the several parts of these researches forms a connecting link between them, but 
other lines of inquiry which suggested themselves in the progress of the experiments 
have also been followed up at considerable length. 
* 1 Proceedings of the Royal Society,’ No. 214, 1881, p. 399. 
t This observation had, I afterwards learnt, been previously made by E. Cohn (Wied. Ann., ^ I., 
p. 385). 
X ‘ Proceedings of the Royal Society,’ No. 216, 1881, p. 22, and No. 228, 1883, p. 123. 
§ The results to be described below were all obtained before I became acquainted with the recent work 
of Warburg on the same subject (Wied. Ann., XIII., p. 141), or with the less closely related observa¬ 
tions of Frojime and Auerbach, to which fuller reference will be made later (§ 32 below). 
