580 
PROFESSOR J. A. EWING OR EXPERIMENTAL 
Specimen. 
Maximum 
r- 
Maximum 
K. 
Y alue of 43, giving 
maximum g. 
Moderately soft iron ring (of fig. 1, § 10). 
1740 
138 
51 
Soft iron ring (of fig. 2, § 11). 
2300 
183 
2-6 
Very soft annealed iron wire, 300 diameters long (fig. 3, § 15) 
3500 
279 
26 
Soft annealed iron wire, 400 diameters long (fig. 4, § 20) 
2670 
212 
2-8 
Annealed iron wire, 400 diameters long (fig. 5, § 21) .... 
2370 
189 
4-5 
Another similar piece (fig. 6, § 22). 
2040 
162 
4-1 
Steel wire, hard-drawn (fig. 8, § 25). 
320 
25 
30 
ditto annealed (fig. 9, § 26). 
470 
37 
18 
Pianoforte steel wire, normal temper (fig. 11, § 28) .... 
273 
22 
39 
ditto annealed (fig. 12, § 29). 
295 
23 
28 
ditto glass-hard (fig. 13, § 30). 
118 
9'3 
55 
Soft annealed iron wire (fig. 14, § 31). 
3080 
245 
2-7 
The same, after being hardened by stretching (fig. 14, § 31) 
670 
53 
107 
Annealed iron wire (fig. 17, § 37). 
2470 
196 
4‘5 
Iron wire hardened by stretching (fig. 19, § 42). 
450 
36 
17 
Pianoforte steel (bis), normal temper (fig. 20, § 44) .... 
258 
20-5 
38 
ditto „ annealed (fig. 21, § 45). 
415 
33 
20 
ditto „ glass-hard (fig. 22, § 46). 
126 
10 
55 
Very soft annealed iron wire (fig. 24, § 50). 
The same wire, with vibration during the application of magnet- 
3500 
279 
27 
ising force (fig. 24, § 50). 
20,000 
(about) 
1600 
(about) 
From 0 to 0'2 
Effects of Stress on the Magnetic Qualities of Iron. 
§ 69. Method of Experiment. —In all the experiments already described, in which 
an iron or steel wire formed the subject of test, the wire hung vertically under the 
action of no other stress than that due to its own weight. I shall now describe a 
series of experiments which were designed to study the variations of magnetic 
quality caused by pull applied in the direction of the wire’s length, which was also 
the direction of magnetisation. 
The wire to be tested was hung vertically with its upper end on a level with and 
due east of a mirror magnetometer. In my early experiments the desired longitudinal 
stress was applied by running mercury into a tank which hung from the lower end of 
the wire, but this plan was soon abandoned in favour of the much simpler one of 
stringing discs of lead on a copper wire which formed a continuation of the iron wire 
under test. Both in accuracy and convenience solid lead weights are preferable to a 
tank containing fluid, and I found it practicable to apply and remove the discs of lead 
with much less mechanical disturbance than was produced by running mercury into 
and out of a tank. This is an important point, for, as will be shown later, the 
influence of stress on magnetism is much affected by even the slightest vibration. 
Lead weights formed the load in all the experiments given below. 
To enable the magnetometer zero to be tested from time to time it was desirable to 
have an easy means of removing the wire and replacing it in exactly the same position. 
The method, showm in the sketch, by which the wire was fixed, admitted of this. 
