RESEARCHES IN MAGNETISM. 
549 
subsequent magnetisation proceeds much less readily as the magnetising force is 
increased, but diminishes much more considerably as the force is withdrawn. The 
residual magnetism of the stretched wire is less than half that of the unstretched 
wire ; but in the former it is much more stable, requiring nearly three times as much 
reverse magnetising force to remove it from the stretched as from the unstretched 
metal. The hysteresis in the stretched wire, during removal and reapplication of the 
magnetising force, is greater than in the case of annealed pieces. 
Several other experiments of a similar kind have shown that the sloping curves of 
fig. 14 are thoroughly characteristic of strained iron. In fact the difference between 
them and the curves of magnetisation of an annealed specimen is so distinctly marked 
that it is easy, by applying this magnetising test, to distinguish a piece which has 
been strained beyond its limit of elasticity from a piece which has not been so strained. 
A comparison of figs. 8 and 9 shows the same kind of difference between the curves 
for a steel wire when annealed (fig. 9), and when in its commercial hard-drawn state 
(fig. 8). The operation of wire-drawing gives a strain which differs from that to 
which the iron wire was subjected in the above experiment only in having lateral 
compression combined with the longitudinal extension. 
§ 82. Interpretation of in a Cycle of Magnetisation .—The hysteresis which 
occurs in the relation of magnetisation in iron to magnetising force has been observed 
and commented on by Professor E. Warburg # in a paper with which I was not 
acquainted when the preliminary notice of my own work, referred to above (§ 1), was 
published. He has also anticipated me in pointing out the important physical value 
possessed by the area of curves representing the relation of © to ■§ or to during 
cyclic changes of magnetisation. 
The quantity — for such a cycle is, in absolute measure, the work spent in 
conducting the metal through its changes of magnetisation, per unit of volume. This 
may very readily be deduced from the general expression for the energy of the 
magnetic field given by Maxwell (El. and Mag. II., § 636) as 
1 
87r 
sm- 
* Wiecl. Ann., xiii., p. 141. The rods experimented on by Warburg were, with one exception, scarcely 
long enough to satisfy the condition of approximate endlessness mentioned in § 7, as needful in experi¬ 
ments intended to answer the question raised in § 2. Reference should also be made in this connexion 
to recent papers by F. Auerbach, Wied. Ann., xiv. (1881), and C. Erohme, Wied. Ann., iv. (1878), and 
xiii. (1881). The papers of Auerbach and Fromme deal, under the name of “ magnetische Nachwirkung,” 
with the influence which previous magnetic state has on the actual magnetisation of iron and steel; but 
their mode of treating the subject differs considerably from that of Warburg. Auerbach distinguishes 
rightly between the “ Nachwirkung ” which is static, and that which (like the “ elastische Nachwirkung ” 
of strained india-rubber, or the “residual polarisation” of a dielectric solid) depends on time. It is the 
former of the two with which we are now concerned. Neither Fromme nor Auerbach used rods long 
enough to give an approach to uniform magnetisation. 
MDCCCLXXXV. 4 B 
