[ 55 ] 
III. Electrodynamic Qualities of Metals .—Part VII. Effects of Stress on the 
Magnetization of Iron, Nickel, and Cobalt. 
By Professor Sir William Thomson, F.R.S. 
Received May 9,—Read May 23, 1878. 
[Plates 2-13.] 
§§ 198 and 199. 
198. In a preliminary notice of investigations regarding tlie effects of stress on 
inductive magnetization in soft iron, communicated to the Royal Society on the 10th 
of June, 1875, I described experiments which afforded a complete explanation of the 
seeming anomalies referred to in §§ 194 and 195,'“ which had at first been so perplexing. 
These experiments showed that the diminution of magnetism in a soft iron wire, 
which I had found to be produced by pull, while the wire was under the influence 
of a constant magnetizing force, was to be observed only when the magnetizing force 
exceeded a certain critical value, and that when the magnetizing force was below that 
critical value the effect of pull was to increase the magnetism—a result which I after¬ 
wards found had been previously obtained by ViLLARi.t The critical value of the 
magnetizing force I found to be about twenty-four times the vertical component of the 
terrestrial magnetic force at Glasgow. Hence the magnetizing force which I had used 
in my first experiment, which (§ 183) was nearly 300 times the vertical component of 
the terrestrial force, must have been about twelve times as great as the critical value. 
Further (which was most puzzling), I found the absolute amount of the effects of pull 
to be actually greater with the small magnetizing force of the earth than that of the 
opposite effects of the 300-folcl greater magnetizing force of my early experiments. 
Thus the effect of the terrestrial force was not only in the right direction, but was of 
amply sufficient amount to account for the seeming anomalies which had at first been so 
perplexing ; and in going over the details of the old observations I find all the anomalies 
quite explained. One of them, that particularly referred to in § 195, is still interesting. 
The alternate augmentation of the residual magnetism by “on” and diminution of it by 
“ off,” with the weight of 14 lbs., corresponded to the normal effect on residual magnet¬ 
ism in soft iron. The elongation of 8 per cent, produced when the 28 lbs. was hung 
on, was no doubt accompanied by a shaking out of nearly all the residual magnetism, 
and an inductive magnetization in the opposite direction by the vertical component of 
* Phil. Trans, for 1876, p. 710 (Read May 27, 1875). 
t Poggendorff’s ‘ Annalen,’ 1868 ; also Wiedemann’s ‘ Galvanismus,’ vol. ii., § 499. 
