32 KEPOKT— 1891. 



will at any stage be determined by summing up the different 

 additions caused by the successive increases to the primary 

 current. Let us now by this last method perform the following 

 series of experiments, using an iron ring : First, increase the 

 magnetizing force H by steps until it reaches a certain value, 

 A ; then diminish it by steps until it reaches an equal negative 

 value, — A ; then increase again by steps until H has once 

 more attained the value A, — noting, of course, at each step by 

 the galvanometer swing the increase or decrease of the induc- 

 tion B caused by that step. Plotting as before the results, we 

 shall find that the history of the magnetic state of the iron of 

 the ring while undergoing this cycle of operations will be given 

 by a curve of the form represented in Fig. III. (PI. II.), which 

 is taken from Professor Swing's work. In this the ascending 

 curve OS represents the way the induction iiicreases with the 

 magnetizing force when we begin with an unmagnetized ring, 

 and it is similar to the curves in Fig. I. (PI. I.), as it ought 

 to be. The descending curve SC'S' tells how the induction 

 diminishes as the magnetizing force is decreased by steps from 

 15 C.G.S. units to —15 C.G.S. units. The ascending curve 

 S'CS tells how the induction B increases as H is increased 

 from —15 units to +15 units. 



This is a very remarkable figure, and tells us many 

 interesting things about the behaviour of iron when it is 

 being subjected to a steadily - varying cycle of magnetizing 

 forces. 



The first fact you will notice is that the descending curve 

 SC'S' is not the same as the original ascending one OS, nor as 

 the ascending one S'CS of the cycle. This shows that the iron 

 has a tendency to remain in its previous magnetic condition — a 

 property which has been called hysteresis by Professor Ewing, 

 to whom a great many of the recent advances in this subject 

 are due. 



You W'ill see from this figure also that when the magnet- 

 izing force has been actually reduced to zero the induction has 

 still the large value 11,000. This is the residual induction or 

 retentiveness, which in some kinds of iron is as much as 90 

 per cent, of the maximum induction. 



Following the curve downwards, you see that a demagnet- 

 izing force, represented on the diagram by OC, is required to 

 reduce the induction to zero. To this definite quantity OC 

 Dr. Hopkinson has applied the old and vaguely-used term 

 coercive force. The upward curve S'CS again exhibits the 

 hysteresis of the iron, the value of the retentiveness and of 

 the coercive force. 



Another yevy important piece of information can be got from 

 this curve. It has been shown by Maxwell that, when a mag- 

 netizing force H produces an induction B in any medium, the 



