PEESIDENTIAL ADDRESS — SECTION A. 31 



represent magnetizing forces H, and the ordinates the induc- 

 tion, B, corresponding ; and, if rings of different material be 

 used, two most striking facts will be at once apparent : — 



(1.) If the material of the ring be of iron, we may get B as 

 much as 2,000 H ; if of cobalt, as much as 120 H ; and, if of 

 nickel, 170 H ; while, if of any other substance (unless a few 

 compounds of these three), B will be found to all intents and 

 purposes simply equal to H. 



This is a very wonderful fact : that, of almost all known 

 substances, only three are magnetic, and these enormously so, 

 while all others are non-rnagnetic. 



(2.) While, in the case of non-magnetic substances, no 

 matter how much H may be increased, B is always equal to 

 it, we find that in the case of the magnetic metals B is not 

 even proportional to H, but increases in a vei*y peculiar manner 

 as the magnetizing force H is increased. 



Fig. I., taken from Eowland's work, shows this clearly. For 

 small values of H, B increases slowly as H increases ; then, 

 when the magnetizing force has reached a certain value, B 

 increases with enormous rapidity as H increases — just as if at 

 this stage the internal magnetic structure of the iron was in a 

 state of unstable equilibrium ; then, as H is still further in- 

 creased, the rate of increase of B becomes small again, and 

 gradually decreases as H is still further increased — thus in- 

 dicating a saturation limit. Fig. II. (PL I.), which is deduced 

 from Fig. I., shows how the permeability fx — that is, the ratio 

 B/H — varies as the magnetizing force is increased. 



Diamagnetic bodies, when the above method of classifica- 

 tion is applied to them, may be considered non-magnetic, as 

 the following table (Hopkinson) will show : — 



If the ring be of iron, B may = 2,000 H. 

 If the ring be of brass, wood, &c., B = H. 

 If of bismuth (the strongest diamagnetic body known), 

 B = 0-999H. 



There are many other most interesting pi-operties peculiar 

 to these three magnetic bodies, the most important of which I 

 will now describe ; but I must first explain another method of 

 conducting the experiments necessary to obtain the curve that 

 represents the relation between B and H. 



With the same apparatus we begin by starting a small cur- 

 rent in the primary, and noting the deflection caused in the 

 secondary ; then, after the needle of the ballistic galvanometer 

 has come to rest at zero, increase the current in the primary 

 suddenly from its last value to a higher one, and note again the 

 deflection caused in the secondary, and so on, increasing H by 

 steps, and calculating from the deflections the corresponding 

 increases in B. Thus the whole induction through the rinjj 



