THE QUANTUM PHYSICS OF SOLIDS 



717 



temperatures. Another theory is that there are only 0.2 electrons in 

 the 45 band, thus leaving the remaining 7.8 electrons of iron distributed 

 5 to one 3d band and 2.8 to the other, leaving an unbalance of 2.2; this 

 theory is unsatisfactory because it would require an inexplicable dis- 

 placement upwards of the 4s band compared to the 3d in going from 



? 2 



Fig. 30 — Intrinsic magnetization versus electron concentration. 



The data for this figure were obtained from the following sources: 

 Fe-V and Fe-Cr M. Fallot Ann. de Physique 6, 305-387 (1936). 

 Fe-Co R. Forrer J. de Physique et le Radium, 1, 325-339 (1930). 

 Fe-Ni M. Peschard Comptes Rendus 180, 1836 (1925). 



Ni-Co P. Weiss, R. Forrer, and F. Birch Comptes Rendus 189, 789-791 (1929). 

 Ni-Cu and Ni-Zn V. Marian Ann. de Physique 7, 459-527 (1937). 



cobalt to iron. Another theory has been proposed by Pauling ^' ; he has 

 stated it in the "atomic" language but it can be translated into the 

 band language as follows: the 3d band is broken into two parts, an 

 upper part containing 4.88 levels per atom, 2.44 for each spin, and a 

 lower part separated from the upper by an energy gap and containing 

 5.12 levels per atom, 2.56 for each spin. A number of electrons per 

 atom varying from 0.6 for nickel to 0.7 for cobalt are in the 45 band; 

 for simplicity we shall suppose that this number has a constant value 

 of 0.65 electrons per atom. According to this simplification, one of the 

 upper parts of the 3d band has 0.65 holes for nickel. This band will 

 become empty if the electron concentration is decreased by 1.79 

 (= 2.44 - 0.65)— that is, for a concentration of 10 - 1.79 = 8.21. 



39 L. Pauling, Phys. Rev., 54, 899 (1938). 



