SOME CONTEMPORARY ADVANCES IN PHYSICS 113 



belief, the atoms of a paramagnetic substance all have a given con- 

 stant magnetic moment, but are oriented in every possible direction 

 so that the resultant magnetic moment of any piece of the substance 

 is zero. If all the atoms could be made to point in the same direction 

 by a powerful magnetic field, the total moment of the piece would 

 be equal to the number of atoms in it multiplied by the moment 

 of each atom, which could then be determined. No attainable mag- 

 netic field is strong enough to do this; the persistent effort of the field 

 to twist the atoms into parallelism is almost completely counter- 

 balanced by the thermal agitation. The total moment of the piece 

 when all the atoms are parallel, and therefore the moment of each 

 atom, have therefore to be calculated from the trend of the magnetiza- 

 tion-versus-field strength curve in its attainable portion. In making 

 this calculation it has heretofore been assumed that all orientations 

 of the atoms are possible. Replacing this assumption by the con- 

 trasting one explained in the foregoing, we find the method of calcu- 

 lation altered; 7 the data heretofore assembled remain valid, but the 

 values of magnetic moment computed from them are replaced by an 

 entirely new set. 



The old set of values of magnetic moment, calculated for a number 

 of solid and gaseous substances and of ionized liquids, by Weiss and 

 others, were said to be integer multiples of a fundamental constant, 

 the "Weiss magneton." No one had succeeded in calculating the 

 observed value of this constant from any atomic theory, and it is not 

 compatible with the picture of the atom given above. The new set 

 of values, according to Gerlach and to Pauli, who have worked over 

 the published experimental material, is compatible with the atom- 

 model. The values for solid platinum and palladium; for nickel in 

 its high-temperature non-ferromagnetic "beta" form; and for nitric 

 oxide gas, agree with the simplest model — the electron in a one- 

 quantum orbit revolving around a non-magnetic centre. The value 

 for gaseous oxygen agrees with the model having an electron in a two- 

 quantum orbit; gamma-iron with the three-quantum, Mn 2 with 

 the 4-quantum and MnO with the 5-quantum model. Various 

 ions in solution from Cabrera's data also give values in accordance 

 with the theory. It is implied that these covfer all the reliable ob- 



7 In the latter case it is assumed that the number of atoms oriented with their 

 axes in one permissible direction Di stands to the number oriented in another per- 

 missible direction D 2 in a ratio given by exp (W/kT) where T is the temperature, 

 k is Boltzmann's constant, and W is the work required to twist an atom from direc- 

 tion Di to direction D 2 against the magnetic field. In the former case all directions 

 are regarded as permissible, and in the assumption just stated, "number of atoms 

 oriented in direction D" is replaced by "density in solid angle of atoms oriented in 

 direction D," a fundamental change. 



