1887.] On the Magnetisation of Iron in Strong Fields. 209 



H, this would mean that in the two metals last named J had passed a 

 maximum, and the process of diamagnetisation which the Ampere- 

 Weber molecular theory of magnetism anticipates had set in. But the 

 uncertainty which attaches to the value of p prevents this conclusion 

 from being fairly drawn from these experiments. A slight excess in 

 the mean value of jj within the metal neck over the value of p in the 

 space contiguous to the neck would suffice to convert the apparent 

 decrease of |J into an increase, with increasing values of §{. So far as 

 these results can be said to bear upon the point in question, they 

 rather support the idea that the intensity of magnetism |j becomes 

 and remains a sensibly constant quantity when the magnetising force 

 is raised to very high values. This maximum of |[ appears to exceed 

 1700 in wrought iron and 1250 in cast iron, and it does not appear 

 likely that any increase of magnetising force will bring the intensity 

 of magnetism in cast iron to a value equal or nearly equal to that 

 which wrought iron is capable of acquiring. It is scarcely necessary 

 to add that our experiments give no support to the suggestion that 

 there is a maximum of the induction ^. The value of ^ capable of 

 being reached by the method we have employed depends mainly on 

 the scale of the experiments. Larger field magnets with pole-pieces 

 tapering to a narrow neck should yield values of ^ greatly in excess 

 even of those we have observed. 



The experiments will be continued and various qualities of steel will 

 be examined with the following modification in the apparatus : — The 

 pole-pieces will themselves be turned, at the ends which face each 

 other, into cones with flat ends, between which the test-piece in the 

 form of a round cylinder will be inserted. The induction will be 

 measured in the neighbourhood of a medial transverse plane only, and 

 the value of the field outside the iron will be determined in this plane 

 at various distances from the axis. Since there is no free magnetism 

 in the iron bar in the medial plane, the magnetic force within the 

 metal is continuous with the force in the surrounding space, and a 

 curve showing the relation of the magnetic force at various points 

 outside to the distance from the axis should admit of being produced 

 so as to give a good approximation to the magnetic force within the 

 metal. If this can be successfully accomplished, the value of the 

 isthmus method of examining the magnetisation of iron will be greatly 

 enhanced. 



[Dr. Hopkinson informs me that he experimented by what we have 

 called the "isthmus" method nearly three years ago, but gave it up 

 from uncertainty about the induction which took place through the 

 coil but not through the iron. In the present experiments this diffi- 

 culty has been avoided mainly by using larger bobbins with a single 

 layer of fine wire for induction coil. I am indebted to Dr. Hopkin- 

 son for the suggestion (soon to be put in practice) that the "isthmus " 



VOL. XLTI. Q 



