STUDIES ON MAGNETIC DISTRIBUTION 105 



seen mentioned elsewhere. When a bar of iron is magnetized at any 

 part and the distribution over the rest quickly measured, on being then 

 allowed to stand some time and the distribution again taken, it will have 

 changed somewhat, the magnetism having, as it were, crept down the 

 bar further. Hence in this Table time was allowed for the bar to reach 

 its permanent state. 



1 R r 

 On looking over column 6, which contains the values of -^ -^ = R'a/i 



(equation 7), we observe that as Q' decreases, the value of R'ap. first 

 increases and then decreases. Now it is not probable that R' undergoes 

 any sudden change of this sort; and so it is probably due to change in 

 the permeability of the rod. Hence by this method we arrive at the 

 same results as by a more direct and exact method. 7 But by this means 

 we are able to prove in the most unequivocal manner that magnetic 

 permeability is a function of the magnetization of the iron and not of the 

 magnetizing force. Hence it is that I have preferred, in my papers on 

 Magnetic Permeability, to consider it in this way in the formulae and 

 also in the plots, while Dr. Stoletow (in his paper, Phil. Mag., January, 

 1873) plots the magnetizing-function as a function of the magnetizing 

 force. 



When we plot the results in this Table with reference to Q' and R'a^, 

 the effect of the variation of R' is apparent; and we see, on comparing 

 the curve with those given in my paper above referred to, that R' in- 

 creases as L increases, at least between L = 2 and L = 8, which is as 

 we should suppose from the arrangement of the apparatus. For this 

 Table I happen to have data for determining Q in absolute measure; 

 and these show that the maximum value of n should be about where 

 the Table shows it to be. 



This method of finding the variation of p is analogous to that of 

 finding conductivity for heat by raising the temperature of one end 

 of a bar and noting the distribution of heat over the bar; indeed the 

 curves of distribution are nearly the same in the two cases. 



If it were thought worth while, it would be very easy to obtain a 

 curve of magnetic distribution for a rod and then enclose the whole 

 rod in a helix and determine its curve of permeability. This would 

 give data for determining R' in absolute measure at every point of the 

 rod. 



To complete the argument that the variation of r z is in great measure 

 due to that of //, I have caused the magnetizing force on a bar to vary. 



7 Phil. Mag., August, 1873. 



