68 
SIR WILLIAM THOMSON ON THE 
chiefly for the purpose of finding the influence of torsion upon the longitudinal 
magnetization of soft iron wire subjected to different amounts of pulling force. These 
experiments, in which the magnetizing influence was simply the vertical component of 
the earth’s magnetic force, were carried out by Mr. Donald Macfarlane. The mode 
of experimenting and the results obtained are described in the following report. 
§§ 223-229. Experiments on the effect of torsion and stretching in altering the induced 
Magnetism of a very soft Iron Wire, subjected to various amounts of constant 
pud. 
223.— Description of Apparatus. —In the Diagram IX. (facing § 201 above) AB is a 
soft iron wire, gauge No. 22, 81 centims. long, to the ends of which were soldered pieces 
of No. 16 copper wire; the upper piece, AD, about 5 metres in length, was attached 
to the ceiling of the room with an arrangement for raising or lowering it through a small 
space, the lower piece, BC, about 50 centims., had attached to its lower end a scale-pan 
for holding the stretching weight. 
E is a small mirror magnetometer, the mirror being 1 centim. diameter, carrying a 
magnet 8 millims. in length and suspended at N by a single silk fibre 10 centims. 
long; and I a lens close to the mirror ; F is a paraffin lamp; and GH a scale 
(bent into a circular arc of which E is the centre), on which is formed the image of 
a fine wire placed in front of the lamp flame, at E. 
L and M are two edges at right angles to one another, fixed to the stand carrying 
the magnetometer, and just in contact with the wire : their use is to make sure 
that the wire is maintained in the same position relatively to the centre of the 
magnet. 
K is an arm soldered to the copper wire, BC, for applying torsion to the wire, and 
immediately below it is a circle divided at intervals of 45° 20', with small holes at 
each division for inserting pegs to keep the arm twisted at any angle while readings 
are being taken. 
A similar arm was soldered to the copper wire AD at 0, the two ends of which were 
in contact with two vertical guides, thus confining the twist to the portion of wire 
between K and O. 
The distance of the wire from the centre of tire magnetometer needle was 8’2 
centims. ; distance of scale from mirror, 157 centims.; one division of scale 0’5 millim. 
The experiments represented in the Diagrams No. X. to No. XXIX. were made in 
this way :— 
Having removed the wire to a distance from the magnetometer, the zero reading of 
the latter was taken ; the wire was placed in position, a stretching weight put in the 
scale-pan, and the reading on the torsion-circle noted when the wire was free from 
torsive stress; torsion was then applied by turning the lower end of the wire 20° at a 
time up to 320', and the reading on the magnetometer scale at each step taken. 
