CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY. 
117 
( b ) The Effect of Variation of Temperature on Xa an< l the Influence of Thermal 
Expansion over the Wide Range of Temperature used in these Experiments. 
Since 
I (xJM.) = £ (x'„V„), 
and 
where the subscript c, l refers to the crystalline or liquid state, we have, dashing the 
deflections d cJ and d l for the higher or lower temperature, 
d' e ,t-d' i 
d c , l -d 1 
Xc.l 
.M- Xa M a +{S-X) 
O Cl ( X c, l • V„ I ) X a 
ca 
, tt a' 
-s—y 
as- 
X« 
as 
Xc.i • M— Xa M a 
which is the ratio of the deflections due to thermal expansions of the substance and 
air, and the variation of the susceptibility of the substance and air with temperature 
over an interval S—S 0 . The necessary data for calculating the magnitude of this 
correction are not known for most of the substances which have been used in these 
experiments, but the largest factor involved is that due to the variation of density 
and susceptibility of the contained air. (The variation of the susceptibility of the air 
external to the phial is eliminated since the difference between d' c l and d\ is taken.) 
For a temperature interval of 150° C., this factor amounts to 2 per cent.* (approxi¬ 
mately) of the deflection, d Ctl — d u due to the substance. This error is, however, 
distributed over a range of 150° C. When the data of thermal expansion of the 
liquid and crystalline states are known, there will be two corrections, one for each 
state, which will be distributed throughout the range of temperature for which that 
state exists. It is to be observed that in this and the foregoing correction the 
thermal expansion of the glass and the variation of its susceptibility witli temperature 
disappear, since they are equally involved in d' cj and d\, the difference of which 
is taken. 
(c) The effect of displacement of the substa?ice {liquid), as a whole, during cooling 
was tested in two ways :— 
(i) By actually displacing the liquid and repeating the observations ; 
(ii) By using, in some of the later experiments, a phial in which the displacement 
was prevented. 
* 
* 
This factor is 
- 0-25 x IQ' 7 x 0-00365 x 150+I‘Ox 0-25 x lQ- 7 /150 
7-25 x 10- 7 
