THE MOLECULAR PROCESS IN MACNETIC INDUCTION. 257 



little magnets, pivoted like compass needles, so that each is free to turn 

 except for the constraint which each one suiters on account of the 

 presence of its neighbors. 



But tirst let us see more particularly wliat ha[)pens when a i)iecc of 

 iron or steel or nickel or cobalt is magnetized by means of a tield the 

 strength of which is gradually augmented from nothin g. We may make 

 the ex])eriment by i)lacing a piece of iron in a coil, and making a cur- 

 rent How in the coil with gradually increased strength, noting at each 

 stage the relation of the induced mag:netism to the strength of the field. 

 This relation is observed t(> be by no means a simple one: it may be 

 represented by a curve (Fig. 1), and an inspection of the curve will 

 show that the process is divisible, broadly, into three tolerably distinct 

 stages. In thcflist stage (<i) the magnetism isbeing nequired but slowly: 



Sq 



MAGNeric Force 



Fig. 1. 



the molecules, if we accept Weber's theory, are not responding readily — 

 they are rather hard to turn. In the second stage (b) their resistance 

 to turning has to a great extent broken down, and the piece is gaining 

 magnetism fast. In the third stage (c) the rate of increment of mag- 

 netism falls oif : we are there ai)proachiug the condition of saturation, 

 thougli the process is still a good way from being completed. 



Further, if we stoj) at any point of the process, such as P, and grad- 

 ually reduce the current in tlie coil until there is no current, and there- 

 fore no magnetic tield, we shall get a curve like the dotted line PQ, the 

 height of Q showing the amount of the residual magnetism. 



If we make tliis experiment at a point in the first stage {a), we vshall 

 find, as Lord Rayleigh has shown, little or no residual magnetism; if 

 we make it at any point in the second stage (/>), we shall find very 

 much residual magnetism; and if we make it at any point in the third 

 stage (c), we shall find only a little more residual magnetism than we 

 should have found by making the experiment at the end of stage h. 

 That part of the turning of the molecules which goes on in stage a con- 

 tributes nothing to the residual magnetism. That part which goes on 

 in stage c contributes little. But that part of the turning wliicli goes 

 on in stage b contributes very mnclj. 

 H. Mis. lU 17 



