RESEARCHES IN' MAGNETISM. 
547 
§ 30. Plate 50, fig. 13, shows the behaviour of a third piece of pianoforte steel wire cut 
from the same bundle as the two last, but rendered glass-bard by sudden cooling in water. 
This time the length was 22 centims. or nearly 300 diameters. The numerical values 
of <§, 23, and p during the first application of the force are given below :— 
Pianoforte Steel Wire, Glass-hardened, Plate 59, fig. 13. 
■s. 
ss. 
0 
85 
(initial) 
— 
10-23 
670 
65 
15-42 
1,150 
74 
20-76 
1,770 
85 
25-69 
2,500 
98 
30-83 
3,260 
106 
35-97 
4,020 . 
112 
41-11 
4,710 
115 
46-25 
5,470 
118 
51-39 
6,060 
118 
56-57 
6,690 
118 
61-97 
7,300 
117 
66-80 
7,880 
117 
72-25 
8,460 
117 
77-08 
8,880 
115 
82-22 
9,380 ' 
114 
87-36 
9,780 
112 
98-11 
10,700 
109 
§ 31. Effect of permanent strain on the magnetic susceptibility and retentiveness of 
Iron .—In the experiments on wrought-iron hitherto cited the metal used was in 
every instance in the annealed state. If after being annealed the piece be subjected 
to a, longitudinal stress sufficient to cause permanent set, and, the stress having been 
removed, the magnetic quality be then examined, a very remarkable change will be 
found to have taken place. Not only is the piece less susceptible, especially to weak 
magnetising forces, but the form of the curves connecting 33 or 3 with Sf is greatly 
changed. The residual magnetism is far less than when the material was in the soft 
annealed state. 
In the following experiment with a piece of soft iron wire 60 centims. long and 
0T58 centim. in diameter, the direct magnetometric method was used. The wire 
was annealed and put through a cycle of magnetisation as in former examples. It 
was then demagnetised by reversals (§19) and hardened by loading it with a weight 
of 60 kilos., which produced a permanent extension of nearly 6 centims. The weight 
was removed and a cycle of magnetisation again gone through. 
The observations have been reduced by calculating $ and 3 (instead of Sf and 23 as 
in former examples; 23 may be readily found from the relation 23=4 tt 3+>§)- Of 
course in calculating 3 for the wire after stretching, allowance has been made for the 
