PRESENT STATUS OF FERROMAGNETIC THEORY 77 



agrees with experiment in that it predicts ''*' the direction and amount 

 of the deviation of B from II for any given value of B. 



The two ways of changing magnetization that have been described 

 for single crystals, namely sudden changes to new directions of easy 

 magnetization, and continuous rotation of domains, apply equally 

 well to ordinary polycrystalline material, the properties of the latter 

 being those of the former averaged for all orientations. One result 

 of this averaging, of course, is that the specimen is now isotropic and B 

 is parallel to H. 



These last remarks must be qualified, for the magnetic materials 

 used by engineers are not always isotropic, that is, the crystal axes are 

 not always distributed equally in all directions. It has been known 

 for many years that when a metal sheet is rolled, the crystals composing 

 it tend to be oriented in special ways with respect to the direction of 

 rolling and to the rolling plane. Even after the sheet has been an- 

 nealed and recrystallized, these special orientations exist, in some 

 metals all the way up to the melting point. Since the magnetic 

 properties depend on the crystal direction in a single crystal, it follows 

 that sheets composed of crystals having special orientations will not 

 have the same magnetic properties in all directions. This was ob- 

 served some years ago in iron, nickel, and iron-nickel alloys.'^ More 

 recently, there has appeared on the market a silicon-iron ^^ alloy for 

 which the permeabilities in different directions are markedly dififerent. 

 Measured parallel to the direction of rolling this material has a per- 

 meability in high fields {B = 15,000 gausses) of 4,000, while measured 

 at right angles to the direction of rolling the permeability is only 400. 

 X-ray analysis shows ^^ that the crystals in this material are aligned 

 so that most of them have a cubic axis lying within a few degrees of the 

 direction of rolling. Thus the direction of rolling coincides with the 

 direction of easy magnetization. 



In considering the properties of single crystals, the properties in 

 very low fields have not been considered, chiefly because precise data 

 for single crystals are very difficult to obtain. The process that occurs 

 in this region in single crystals and polycrystals must be different from 

 either of the two so far considered, because in ordinary polycrystalline 

 material there are no discontinuities in magnetization, i.e., no Bark- 

 hausen effect, and also the fields are not strong enough to rotate the 

 domains to any significant extent against the crystal forces, out of a 

 direction of easy magnetization. Knowing the relation between 

 magnetic force and angular displacement in high fields, it is calculated 

 that if this same mechanism applied to changes in magnetization in 

 very low fields the highest value of initial permeability in iron would 



