350 



SCIENCE 



[N. S. Vol. XLVI. No. 1189 



One of Oxley's most interesting conclu- 

 sions is that the mutual magnetic field of 

 two diamagnetie molecules in intimate con- 

 tact is of the order of 10^ gausses. 



Langevin's hypothesis, while probably 

 the most satisfactory yet advanced, leaves 

 us quite in the dark as to a mechanical ex- 

 planation of the architecture of the mole- 

 cule. 



In paramagnetic and ferromagnetic sub- 

 stances in accordance with the views of 

 Langevin the role played by the molecule 

 is not as in diamagnetie substances inde- 

 pendent of the molecule's orientation in 

 space, and it is necessary to assume that the 

 effect of an applied field is to rotate the 

 electronic orbits so that the direction of the 

 resultant external field of a molecule tends 

 toward that of the applied field. But the 

 theory tells us nothing of the mechanism 

 which will account for this orientation. 

 Resisting the orientation will be heat agi- 

 tation and perhaps inter-atomic and molec- 

 ular actions of other than magnetic type. 

 In a paramagnetic gas the resistance to 

 orientation is supposed to be entirely due to 

 heat agitation. The theory for such a gas 

 leads to an expression for the susceptibility 

 which depends upon both the impressed 

 field and the temperature, but for fields at- 

 tainable in the laboratory the susceptibil- 

 ity varies inversely with the absolute tem- 

 perature in accordance with Curie's law 

 for paramagnetism. 



With the aid of the assumption that as 

 regards rotation the molecules of a para- 

 magnetic liquid behave like those of a para- 

 magnetic gas it is possible to extend the 

 theory of the gas to include that of the 

 liquid, and such an extension is probably 

 reasonably safe for liquids not given to 

 polymerization. 



In Weiss 's theory of f erromagnetism it is 

 assumed that, so far as rotation is con- 

 cerned, the molecules of a ferromagnetic 

 substance behave like those of a paramag- 



netic gas, a somewhat questionable assump- 

 tion in this case. The effect of neighbor- 

 ing molecules upon a given molecule is as- 

 sumed to be that which would be produced 

 by a very large localized magnetic field of 

 the order of 10^ gausses. The theory based 

 on these assumptions succeeds to a remark- 

 able extent in explaining many of the facts 

 of ferromagnetism. 



The large internal fields hypothecated 

 by Weiss and by Oxley are to be regarded 

 as devices for averaging out in a measure 

 the complicated effects due to molecular 

 structure. 



Through experiment Weiss was led to 

 belief in the existence of an elementary 

 unit of magnetic moment which he called 

 the magneton. This corresponds in elec- 

 trical theory to the electron. In many in- 

 stances the magnetic moment per molecule 

 appears to be very nearly an integer num- 

 ber of magnetons. But the evidence is not 

 weighty enough to justify the acceptance 

 unreservedly of this proposed new physical 

 unit. 



The subject of magneto-chemistry is al- 

 ready a very extended one. Here the at- 

 tempt is made to establish a connection be- 

 tween the magnetic moment of a compound 

 and those of its constituents, and additive 

 relations are sometimes found. Substan- 

 tial chemical information is often found 

 through magnetic analysis. Various at- 

 tempts have been made to explain chemical 

 valency bonds through the magnetic at- 

 tractions of rotating electrons in the atoms. 

 One of these, that of Parsons, offers promise 

 of considerable success in this direction. 



The recent magnetic experiments of Bar- 

 nett and of Einstein and deHass appear to 

 prove definitely the existence of electrons 

 rotating in closed orbits within the mole- 

 cules of material substances, and thus fur- 

 nish important support to Langevin's 

 fundamental assumptions. 



From this necessarily inadequate discus- 



