RESEARCHES IN MAGNETISM. 
545 
reapplied. The greatest value of 33 reached was between 14,000 and 15,000, and of 
this 57 per cent, survived the removal of the inducing force. 
§ 26. The same wire was then annealed by heating to bright redness and cooling 
slowly, and observations made on it whose results are given graphically in Plate 58, 
fig. 9. The curves here have a form much more similar than those of fig. 8 to the 
curves for soft iron, but the susceptibility to moderate magnetising forces is much 
less than in iron. The induced magnetism is nearly the same as for the hard-drawn 
wire, but the residual magnetism is considerably greater for the annealed than for 
the hard-drawn wire, being now from 76 to 80 per cent. 
§ 27. In the next experiment the same piece of wire was again tested, after being 
made glass-hard by being plunged into water while at a bright red heat. Its magnetic 
behaviour in this state is shown in Plate 58, fig. 10. The highest magnetising force 
applied (over 55 c.g.s. units) did not carry the wire past the steep part of the curve of 
magnetisation, and gave only 9300 as the value of 33. On the removal of this force the 
residual value was 6360, or 68 per cent, of the induced. This must not, however, be 
taken as a measure of the percentage of residual magnetism which would have been 
found had the steel been more nearly “ saturated.” To remove the residual magnetism 
a reverse force of over 40 units was necessary. A noticeable feature in this case is 
the smallness of the hysteresis when the magnetising force was withdrawn and 
reapplied. The “off” and “on” curves for these operations are almost coincident. 
In fact, so far as this point is concerned, Maxwell’s extension of Weber’s theory 
(§ 3) would serve satisfactorily enough to explain the magnetic retentiveness of this 
material. Fig's, 8, 9 and 10 are all drawn to the same scale, which is also the scale 
of fig. 6. 
§ 28. Experiments with Pianoforte Steel .—The foregoing experiments on steel 
showed the desirability of testing that material under greater values of the mag¬ 
netising force. Accordingly another series were made in which the method used 
was the direct magnetometric method (§ 18), and the magnetising force was raised 
to over 90 units. The wire used in this series of tests was pianoforte steel 0'078 
centim. in diameter, and 30 centims. or 380 diameters long. Observations were first 
made with the wire in its normal commercial temper; their results are given below 
and also in Plate 58, fig. 11, which shows the relation of 33 to <§ while the mag¬ 
netising force was being applied, reversed, reapplied, removed, and reapplied as in 
former examples. 
