CONTEMPORARY ADVANCES IN PHYSICS 361 



netized by weak fields. This means that the thermal agitation is too 

 strong an antagonist to the applied field. Weiss therefore provided 

 the latter with a powerful ally, in the form of an intense molecular 

 field parallel to it and proportional to the magnetization. The 

 applied field and its ally together are able to overpower the thermal 

 agitation and bring about saturation in cold iron. Now to say 

 "molecular field" is merely to use a different phrase for "influence of 

 the atoms on one another." In the theory of Weiss, this influence of 

 the atoms on one another helps the field to align them; " in Ewing's 

 theory, it hinders the field. How do away with this arrant contra- 

 diction? 



Perhaps a partial union may be effected, in this wise. According to 

 Langevin and Weiss, a piece of cold iron consists of a multitude of 

 small zones or regions of atom-groups, each magnetized to a high 

 degree, their directions of magnetization dispersed at random; an 

 applied field acts primarily by wheeling these magnetizations into line. 

 According to Ewing, a piece of cold iron consists of a multitude of 

 chains or pairs of systems of elementary magnets, which an applied 

 field upsets, perhaps only to re-weld them anew into more favourably 

 oriented chains. Weiss deals with the state of affairs inside the atom- 

 groups; Ewing deals with the effect of the applied field in breaking up 

 and rebuilding the atom-groups. Might one say that Weiss explains 

 the conditions, under which the elementary magnets form themselves 

 into groups or chains such as Ewing preassumed? that Ewing describes 

 the action of the external field upon these groups, an action which 

 Weiss left imprecise? so that the two theories, when properly revised, 

 will complement each other? It seems possible. At all events, each 

 of the theories has so many successes to its credit, that there can be no 

 thought of discarding either for the sake of the other. Those who are 

 weary of trying to reconcile waves and quanta might refresh them- 

 selves by reflecting on this problem. 



McKeehans Theory 

 In the theory of McKeehan, magnetostriction is promoted to the 

 dominant role. The distortion which a metal undergoes when it is 

 magnetized is held responsible for hysteresis, and for the fact that the 

 rise of the /-vs.-// curve is gradual, not sudden. This view was 

 suggested by the fact which I have mentioned already: that, in the 

 series of the permalloys, the permeability reaches a surprisingly high 

 maximum value and the hysteresis a surprisingly low minimum value, 

 just at that alloy of which the magnetostriction is indetectably small 

 until saturation is nearly attained — the alloy intermediate between 



