190 Dr. Harrison on Variation with Temperature of 
previously employed in the thermoelectric experiments, and 
to measurements of the variation of w with temperature for 
fields not less than 10 C.G.S. In most cases, therefore, it 
will be seen by fig. 7 that the magnetic force has not been 
sufficiently reduced to exhibit the bend in the wH curve 
which corresponds to maximum permeability. 
Fig. 8. 
eet Eas Tal 
ge ie 
ce eat ae He a 
‘LER ReaRES 
Permeability. 
200 400 600 
Temperature Centigrade. 
Estimated Accuracy of the Measurements. 
Reading the scale to ; mm., the maximum error in 
temperature at 500° is 0°°5 C. And for fields of 40 C.G.8 
units an accuracy of 1 partin 500 is attained on the intensity 
of magnetization. 
Discussion of Results. 
NICKEL. 
If we consider IH curves for nickel at any two temperature 
below the critical point, ‘‘ the curve for the lower temperature 
lies at first below, and afterwards above the other” *. 
In the series of results given in Tables I. & IT. (pp. 197-200), 
and represented graphically in fig. 4, the effect of temperature 
on permeability is well brought out. 
* Ewing, ‘Magnetic Induction in Iron, and other Metals.’ Electrician 
Series. 
