294 Proceedings of Boyal Society of Edinburgh. [sess. 
magnetised) into a magnetically “ neutral ” condition by the process 
of reversals. 
Eowland gives 17'6 as the maximum value for A;, the true suscep- 
tibility. The apparent maximum susceptibilities for the four nickel 
bars were found as follows : — 
A;' = 6-6, 5-4, 5*2, 5T . 
Assuming the relation It yl*P) , we get 
and find for the four corresponding values of P the numbers 
P=‘095, -128, T35, T39 . 
The areas of section of the metal are in the ratios of the numbers 
‘095, -127, T45, T49 . 
To satisfy the law that the demagnetising factor is proportional to 
the cross-section, these corresponding numbers should be in the 
same ratio, each to each. The first two pairs are in accord ; and 
so are the third and fourth pairs. But otherwise there is an 
obvious discrepancy. Nevertheless, when we consider the numerous 
probable sources of discrepancy between theory (admittedly approxi- 
mate) and experiment, it is matter rather of surprise that there 
should be any concordance. 
