66 
SIR WILLIAM THOMSON ON THE 
The results in these experiments were obtained by beginning with a negative (i.e., 
opposed to magnetizing force of earth) magnetizing current of about 700 divisions 
of the battery galvanometer scale, which was gradually diminished to zero and then 
increased until a positive current of 700 (20 cells) divisions was reached. This process 
was then exactly reversed. The results may be examined in the order in which they 
were obtained, by beginning at the left hand ends of the curves A and B, and passing 
to the right along the thinner lines, returning to the left along the thicker lines. 
218. The magnetometer was found in this position to be at too great a distance 
from the barrel to show the residual magnetism with accuracy, and accordingly its 
distance from the barrel was reduced to one metre, and the effects of residual magnetism 
alone observed. These are shown by curves marked B' of Diagram YIII'. The method 
and order of experimenting were here the same as described in § 217 ; and the 
explanation of Diagram YIII., curves P», applies also to Diagram YIII ., except that 
in the latter the directions of the ordinates are reversed from the former; thus in 
YIII'. positive ordinates indicate positive magnetization, negative ordinates negative 
magnetization (see § 205). 
219. Beginning at the extreme left of curves marked B' on Diagram YIII'., and 
following the arrows, it will be seen that the residual magnetism remains nearly 
constant in amount until the magnetizing current has been diminished to about 300 
divisions, when it begins to take a greater negative value, and continues to do so until 
the current is brought down to 50 divisions, when it begins slowly to diminish. After 
the reversal, of the current the full residual magnetism diminishes with great rapidity, 
passing through zero at about 15 divisions of positive current. It then becomes positive, 
preserving nearly the same rapidity of change for some distance beyond zero. After 
a positive current of 150 divisions is reached, the full residual magnetism increases 
very slowly, and the curve becomes asymptotic towards a value of about 440. The 
curve showing the residual magnetism after ten “ ons ” and “ offs ” has similar charac¬ 
teristics to those for the full residual magnetism ; the increase on the left side of zero 
at about 300 divisions of negative current is, however, more decided. 
The general character of the return curves is similar to that just described. It is 
to be remarked, however, that the zero of current in each of them is much further 
passed before the zero of magnetization is reached. This difference between the going 
and returning curves would be done away with, and the curves from left to right and 
right to left in the diagram would be perfectly symmetrical about the zero of magne¬ 
tizing current, if the influence of the earth’s magnetic force were eliminated. I intend 
to return to this subject with a modification of the experimental arrangements, to 
allow the residual magnetism to be observed unaffected by any influence due to 
magnetizing force in the direction of the length of the tube ; instead of as here with 
the tube always under the magnetizing action of the vertical component of the earth’s 
magnetic force, when the electro-magnetic current is not flowing. 
The curves, except in the positions corresponding to zero of current, show that the 
