RESEARCHES IN MAGNETISM. 
G03 
nearly cyclic state, the form of curve with which we are already familiar is given, 
with a maximum at 8^ kilos, on the on branch, and at 5 kilos, on the off branch. 
The wire being now without load, the magnetising current was next reduced to the 
value just necessary to neutralise the earth’s vertical force—in other words, the field 
was reduced to zero—and the effects of stress on the residual magnetism were 
examined. The effect of the first application of the cycle 0 — 13 — 0 on the residual 
magnetism is shown in the curves marked IV«. in fig. 40. We see in it the resultant 
of two actions, viz., the usual cyclic effect and, superposed on it, a rapid reduction 
due to magnetic instability. On applying a load of 1 kilo, the fall of residual 
magnetism (resembling the fall which would be caused by tapping, or the rise which 
either tapping or loading would cause in the magnetism induced by an increasing 
field) more than counterbalances the small augmentation of magnetism which (in the 
cyclic condition) would be the effect of this load, and a curious dip at the beginning 
of the on curve is the result. Then the augmentation by stress becomes relatively 
more influential, until the maximum is passed, after which the two actions combine to 
give a rapid fall of magnetism as the load is further increased. 
The cycle 0 —13 — 0 was again applied twice, and, when the residual magnetism had 
settled into a nearly steady state, the cycle shown at IV&. in fig. 40 was taken. 
The same wire was then subjected to several other magnetising fields, and cycles of 
loading performed in them. The results, in each case after the cycle 0—13 — 0 had 
been repeated several times, are shown in fig. 40, where the curves are numbered as 
follows :— 
No. of curve. Magnetising field (c.g.s. units). 
I. 
0-34 
II. 
1-66 
III. 
2-17 
IV. 
3*33 
V. 
5-17 
The curves show the values of 3 and the total load, which can be reduced to kilos, 
per sq. mm. by multiplying it by 2’18. 
§ 87. (May 5, 1882.) In the above experiment the range of magnetising force was 
somewhat limited, and the following experiments were made to test the effects of 
stress on higher values of magnetisation. 
In this and subsequent experiments the method of demagnetising by reversals of a 
gradually diminishing current was used to bring the wire to an initially neutral state 
between successive operations. The method has been described in § 19. 
The wire was the same piece of iron as before (§§ 83-86), but before these experi¬ 
ments it was re-annealed and then stretched by a load of 15^ kilos, which stretched it 
from 30*4 to 31’8 centims. The effects of loading up to 15 kilos, were examined in 
three fields, viz., 0'34, 2*49, and 18‘65 c.g.s. units. The operations in each case were 
as follows :— 
