342 BELL SYSTEM TECHNICAL JOURNAL 



celebrated theories of magnetization rest upon speculations about the 

 nature of this agency which fights against the field. 



On considering the unsurpassably simple system composed of only two 

 elementary magnets close together, J. A. Ewing discovered that their 

 interactions are such, that they can prevent each other from aligning 

 themselves immediately along the field; one can almost say that they 

 "interlock," and they interlock in such a way, that the pair of them 

 displays a tripartite /-vs.-// curve, and the quality of hysteresis, 

 though neither separately has any such properties. Systems com- 

 prising a dozen, a score, or a multitude of such magnets, arranged in 

 chains or in a cubical array, may be devised to imitate actual initial 

 curves and actual hysteresis loops with stunning accuracy (Fig. 11). 

 Such close agreements need not be overstressed. The astonishing 

 feature of Ewing's discovery is (I think) that although each individual 

 magnet possesses neither the quality of gradual magnetization nor the 

 quality of hysteresis, a pair of them put close together possesses both. 

 So great a result is attained from so simple an apparatus, that it 

 seems very unlikely that any radically different explanation of either 

 quality will ever be put forth. Whatever may be added to Ewing's 

 model, its central idea will probably never be supplanted. 



P. Langevin, devising a theory for paramagnetism, supposed that 

 the agency which combats the aligning influence of the field is the 

 thermal agitation of the magnetic atoms. Contrary to one's first 

 impression, this theory is not easily visualized; but it establishes a 

 union between paramagnetism on the one hand, and the great general 

 principles of thermodynamics and equipartition of energy on the other. 

 In the form in which Langevin put it forth, it does not account for 

 hysteresis. 



P. Weiss supplemented Langevin 's theory by supposing that the 

 actual magnetizing field prevailing inside a magnetizable substance is 

 not that sum of the applied field He and the "demagnetizing field" Hi 

 which I defined in Section A, but a combination of this sum with 

 another term depending on the magnetization. As I stressed in 

 Section A, experience teaches us nothing about the value of the true 

 field inside a magnet; Weiss' assumption was therefore a perfectly 

 legitimate choice, to be justified (if at all) by its fruits. One of these 

 is, that it accounts for the presence of hysteresis at low temperatures 

 and its absence at high. 



Ewing's Theory 



Ewing conceived a piece of iron as an assemblage of tiny bar- 

 magnets, each endowed with a fixed and constant magnetic moment, 

 and wheeling about a pivot under the combined influence of the 



