382 Prof. J. A. Ewing on the Magnetic Susceptibility 
In the experiments on magnetization, iron and steel wires 
were used, either welded into rings or in the form of straight 
pieces of such great length that the influence of the ends was 
negligible. Curves were obtained, in some cases by the 
ballistic method, and in others by the direct magnetometric 
method, showing the changes of magnetization which occurred 
when magnetizing force was gradually applied, withdrawn, 
reapplied, reversed, and so on. 
The results of many experiments with several specimens of 
carefully annealed soft-iron wires have shown that they possess 
in very high degree a property not generally credited to soft 
iron, — the property of remaining strongly magnetic when the 
magnetizing force is removed. 
As an example, the case may be cited of an annealed iron 
Mare which was subjected to a magnetizing force of 22*4 
C.G.S. units. This gave it a magnetic Induction amounting 
to 16,000 C.G.S. units. When the magnetizing force was 
gradually and completely removed, the Induction fell only to 
15,000 units. In other words, the Intensity of residual mag- 
netization was equal to nearly 1200 C.G.S. units. 
Here more than 93 per cent, of the whole induced magneti- 
zation survived the removal of the magnetizing force; while 
in many other cases the residual magnetism amounted to nearly 
90 per cent. The somewhat extraordinary spectacle was thus 
presented of a piece of soft iron, entirely free from magnetic 
influence, and nevertheless holding (per unit of its volume) 
an amount of magnetism far in excess of what is ever held by 
permanent magnets of the best tempered steel. 
In this condition, however, the magnetic character of the 
iron is highly unstable. The application of a reverse mag- 
netizing force quickly causes demagnetization ; and the 
slightest mechanical disturbance has a similar effect. Gentle 
tapping removes the residual magnetism almost completely. 
Variations of temperature reduce it greatly, and so does any 
application of stress. On the other hand, if the iron be 
carefully protected from disturbance, it seems that the residual 
magnetism disappears only very slowly, if at all, with the mere 
lapse of time. 
If, after magnetization, the magnetizing force be removed 
suddenly, the residual magnetism is, as might be expected, 
less than if the force be removed gradually. 
The ratio of residual to total magnetization is always small 
when the intensity of magnetization is small, and passes a 
maximum when the intensity is increased. This maximum is 
particularly distinct in wires which have been hardened by 
stretching; but it also occurs in soft annealed wires. In one 
