318 BELL SYSTEM TECHNICAL JOURNAL 



pernicious; one per cent of this element mixed with iron brings the 

 maximum permeability down to 350. A few per cent of manganese 

 mixed into iron reduces //if to the nearly-constant value of .03. 

 Tempering, cold-working, forging, and drawing all tend to reduce the 

 permeability. Since these processes render the metal harder in the 

 literal sense of the word, the change which they imprint upon the 

 magnetization-curve is called by association of ideas a "magnetic 

 hardening." As ^max. is reduced by any of these processes, the 

 contrast between the three segments of the I-vs.-H curve diminishes, 

 and in some cases there is scarcely more than the point of inflection 

 left to mark the passage from the initial to the final range of the curve. 

 The final approach to saturation conforms to the law 



7 = /, 



max. 



O-f/) 



The value of the constant c is large for magnetically-hard materials, 

 and small for the well-annealed samples for which the tripartite 

 division of the I-vs.-H curve is obvious. In iron (I quote Weiss) 

 saturation is approached within one promille at a field strength of 

 5500, in nickel at H = 10000. In permalloy it must be approached 

 as closely with a field of a few dozens of gauss. 



The saturation-intensity of magnetization, or saturation for short, 

 is much more nearly independent of the present hardness and the 

 past mechanical and thermal treatments of the material than the 

 other features of the initial curve — much more nearly, therefore, a 

 function of the chemical composition exclusively, than is any other 

 single nameable magnetic quality. For this reason it is possible to 

 present such a Table as the accompanying one with comparatively 

 few qualifications. The first column of figures contains values of 

 /max. obtained near room-temperature; the second, values measured 

 at the temperature of boiling hydrogen, inserted here for future 

 reference.* 



* These values may be described as the "saturation magnetization of a cubic 

 centimetre" of the materials in question. Dividing each by the density p of the 

 material, we obtain the "saturation magnetization per gramme." Multiplying this 

 by A, the atomic weight of the element or molecular weight of the compound (if the 

 material is of either sort) we get the "saturation magnetization per gramme-mole- 

 cule." This last is the quantity most often tabulated, being sometimes expressed in 

 "magnetons" (units equal to 1126 C. G. S. units; cf. page 353). It may be advisable 

 to recall that an isolated cube containing one cubic centimetre, or one gramme, or 

 one gramme-molecule of material would not acquire the magnetization in question 

 at any finite field, since it could not be magnetized uniformly. 



I 



