October 14, 1921] 



SCIENCE 



343 



creased. If the atom is built so that there 

 are a number of electronic orbits so oriented 

 that their resultant magnetic moment is zero 

 then there will be no tendency for the atom 

 as a whole to rotate, but on the application of 

 the magnetic field there will be a tendency 

 to alter the magnetic moment of each electron- 

 ic orbit and no matter in which direction 

 the electron is revolving the effect of the 

 magnetic field is to create a polarity opposed 

 to that of the applied field. If the magnetic 

 moment of one electronic orbit is positive 

 the effect of the external field is to decrease 

 it and if the magnetic moment of another 

 orbit is negative the external field acts to in- 

 crease it so that the total effect is the same 

 as that which we get from Weber's'' theory 

 of diamagnetism which assumes that there are 

 no revolving electrons present to begin with 

 but when a diamagnetic substance is exposed 

 to a magnetic field, currents are set up in the 

 atoms or molecules which develop magnetic 

 fields having an opposite polarity to that of 

 the inducing field. If the orbits of these 

 circuits are resistanceless the currents will be 

 maintained until the magnetic field is with- 

 drawn again. It is to be noted that in the 

 case of diamagnetic substances a finite mag- 

 netic moment is developed in the elementary 

 unit with which we are dealing and which 

 ought to have a corresponding tendency to 

 rotate in a magnetic field. This point does 

 not seem to be emphasized in the theory of 

 diamagnetic substances, but as we shall see 

 later on it is stressed in paramagnetic bodies. 

 We know that an elongated portion of a dia- 

 magnetic substance does orient itself very 

 definitely in a magnetic field. From the 

 standpoint of the theory of diamagnetism 

 just reviewed, diamagnetism must be almost 

 a universal property of matter because we 

 find the Zeeman effect in nearly all spectral 

 lines of nearly all substances. We believe 

 that the hydrogen atom has only one electron- 

 ic orbit. Its diamagnetism is difficult to ex- 

 plain by Langevin's theory. 



isDushman, Gen'l Elec. Eev., p. 20 of reprint 

 from May, Aug., Sept., Oct., and Dec. issues, 1916. 



2. Langevin's Theory of Paramagnetism. — 



We have seen that in all cases the creation of an 

 exterior magnetic field modifies the electronic orbits 

 by polarizing diamagnetically all the molecules. 



If the resultant moment is not zero, upon the 

 diamagnetic phenomena is superimposed another 

 phenomenon due to the orientation of the ele- 

 mentary magnets by the external field. The sub- 

 stance is then paramagnetic if the mutual action 

 between the elementary magnets is negligible, as 

 in the case of gases and of solutions and ferro- 

 magnetic in the case where the mutual actions play 

 the essential roles. As soon as the paramagnetism 

 appears it is, as a rule, enormous in comparison 

 with the diamagnetism and therefore completely 

 conceals it. This explains the discontinuity be- 

 tween paramagnetism and diamagnetism; para- 

 magnetism may not exist; but if it does, it hides 

 completely the diamagnetism. 



Therefore, substances whose atoms have their 

 electrons in revolution in such a way that their 

 effects are additive, are paramagnetic. The atoms 

 of such substances may be looked upon as elemen- 

 tary magnets. 



If we think of the elementary magnets at 

 ordinary temperatures as being in a state of 

 agitation then the tendency of the elementary 

 magnets to orient themselves in a magnetic 

 field will be opposed by the thermal agitation 

 of the elementary magnets and they will settle 

 down under a state of statistical equilibrium. 



3. Weiss's Theory of Ferromagnetism. — 

 Langevin has given a theory of dia- and para- 

 magnetism and largely assumes ferromagnet- 

 ism as a special case of paramagnetism. That 

 ferromagnetism is a special case of paramag- 

 netism will, I think, be conceded by all, but to 

 explain more completely the phenomena at- 

 tendant on ferromagnetism, Weiss has ex- 

 tended the theory somewhat by saying that to 

 explain the varied phenomena as we find them, 

 there must be associated with the turning of 

 the elementary magnets something which acts 

 like an extra magnetic field in addition to the 

 external field applied. After considering all 

 phases of the problem, however, and showing 

 that he can explain many of the existing 

 phenomena by means of this extra or intrinsic 

 molecular field he is forced to admit that this 

 "molecular field must be attributed to the 



