CONTEMPORARY ADVANCES IN PHYSICS 343 



impressed magnetic field and the magnetic attractions and repulsions 

 of its neighbors. 



Imagine a chain of long slender bar-magnets end to end, the positive 

 pole of each almost touching the negative pole of the next — ^that is 

 the equilibrium position which they would naturally assume, so long 

 as no external field affects them. By preference, build such a chain 

 out of pivoted magnets; for Ewing's model enjoys the singular merit, 

 that it can be made out of actual magnets and exhibited to the eye. 

 Now there is a remarkable feature of this chain : if a magnetic field is 

 applied to it in some oblique direction, then so long as the fieldstrength 

 is quite small the individual magnets incline themselves toward it 

 slightly, each setting itself at the same angle to the direction of the 

 chain which was originally the common direction of them all ; and when 

 the fieldstrength is gradually increased the angle increases gradually, 

 but only up to a certain point — for suddenly, at a critical moment, all 

 the bar-magnets very suddenly capsize, and set themselves in nearly 

 the direction of the field. I use the word capsize to invoke the too- 

 familiar analogy of the upsetting boat. As weights are piled upon one 

 side of a boat, it responds at first by tilting gradually sidewise and 

 downward; to each slight increment of the load it accommodates itself 

 by finding an equilibrium-slant a little farther over; but eventually 

 there comes a moment when balance and compromise are no longer 

 possible; the boat cannot find a position of equilibrium except by 

 overturning, and this it does, suddenly and irrevocably. Such is the 

 behavior of a chain of bar magnets; and this is the property which 

 adapts it for representing the general shape of an initial magnetization- 

 curve such as I showed in Fig. 1, with its first slowly rising arc followed 

 by the rapid uprush and the final slow adjustment to saturation. 



The overturned boat will not right itself even when the load which 

 upset it is removed; will the chain of bar-magnets be equally unfor- 

 giving? The analogy is not perfect, except in one very particular 

 case: if the angle between the direction of the chain (defined as the 

 direction in which the north poles of all the magnets originally pointed) 

 and the direction of the field is 180°, the capsizing will result in a 

 right-about-face of each magnet and a reversal of the so-defined 

 direction of the chain, and this reversal will persist after the field 

 is annulled. 



Suppose however that there is a multitude of chains oriented at 

 random, so that half of them are inclined at less than 90° and half 

 at more than 90° to the direction of any strong field which we choose 

 to apply. The field will cause all the bar-magnets to capsize (except 

 those belonging to the few chains to which it is almost parallel) ; 



