458 
1)R. J. HOPKINSON ON THE MAGNETISATION OF IRON. 
Let the mean length of the lines of induction in the sample be l, and cr the section 
of the sample; let V be the length of lines of induction in the block, and cr' their 
section, © the intensity of induction in the sample, 53 in the block, then crSB=crb8'=I; 
let 
53 = fo§, 
and 
then 
or 
A* P 
II 
=Z^TTllC — 
fia 
§ds 
Vf_ 
cr' 
where n is the number of convolutions of the magnetising coils. Now in the instru- 
II' 
ment used cr' is large, and p/ is as large as can be obtained, hence the term is 
small comparatively. My first intention was to correct the magnetising force by 
deducting this small correction, but finally I did not do so, because in the more 
interesting results the magnetism of the block is dependent in part upon previous 
magnetising forces, the effect of which cannot be allowed for with certainty. We know 
then that in all the curves the magnetising force indicated is actually too great by a 
small but sensible amount, which does not affect the general character of the results or 
their application to any practical purpose. The magnetising force then at any point 
4r7T?iC • 
of the sample is—= —-a small correction which we deliberately neglect. There 
C fX i 
is another source of uncertainty iu the magnetising force : the length l is certainly 
greater than the space within the wrought iron block, but it is not possible to say 
precisely how much greater. If the sample bars and the block were a single piece, the 
results of Lord Rayleigh for the resistance of a wire soldered into a block would be 
fairly applicable ; but it is essential that there should be sufficient freedom for the bar 
to slide in the hole ; the minute difference between the diameters of the sample and 
the hole will increase the value which should be assigned to l. Throughout, l is 
assumed to be 32 centims., and it is not likely that this value is incorrect so much as 
half the radius of the bar, or 1 per cent. The magnetising forces ranged up to 
240 C.G.S. units when both bars were of the same material. In some cases a single 
bar only was available for experiment; the plan then was to use it as the bar -which 
enters into the exploring coil, and for the other to use a known bar of soft iron. We 
have then to deduct from Anne the magnetising force required to magnetise the bar of 
soft iron to the state observed, and to distribute the remainder over the shorter length 
of sample examined. The results obtained in this way are subject to a greater error, 
because some lines of induction undoubtedly make their way across from the end of 
