CONTEMPORARY ADVANCES IN PHYSICS 



237 



ordiiKito at saturation ^oes down as the temperature goes up. And yet, 

 we interpret ferromaj>;netism by what is essentially an atomic theory: 

 that is to say, we suppose that any piece of iron is an aggregate of httle 

 magnets each having a constant magnetic moment (so long as the 

 temperature is kept constant) and that magnetization of iron consists 

 in aligning these magnets. 



I think it instructive to refer to these little magnets by the name of 

 "atom," with some distinctive prefix; so, for a few minutes, I will call 

 them "super-atoms," though this is not the customary name. When a 

 piece of iron is unmagnetized or demagnetized, the super-atoms are 



0.3 0.4 0.5 0.6 0.7 



MAGNETIC FIELD STRENGTH, H 



Fig. 2 — Magnetization of a ferromagnetic material (81 permalloy, annealed two 

 minutes at 1000° C); the ordinate is I. Data by P. P. Cioffi. 



pointing in all directions at random, just like the individual atoms of a 

 paramagnetic gas which is unmagnetized. When a magnetic field is 

 applied to the unmagnetized iron, the super-atoms get more or less 

 aligned with one another. If the field is strong enough they are 

 perfectly aligned, and there exists what is usually called "saturation" 

 of the iron. Now it is worse than useless to remember about Boltz- 

 mann's theorem, or impacts, or free flights between impacts, for all 

 those concepts have no relevance. We have to look at the phenomena, 

 and see what they require. 



