sosman: problems of the oxides of iron 65 



Two different views have been held concerning the relation of 

 these two classes of polymorphic inversion, although in our 

 present ignorance of atomic structure neither can be very pre- 

 cisely stated. It has been supposed, on the one hand, that the 

 slow changes were the fundamental ones, involving a rearrange- 

 ment of the atoms in the molecule or a change in the molecular 

 weight, the expression "molecular weight" being considered to 

 have a meaning in the solid as in the gaseous state. The rapid 

 reversible inversions, then, were held to be merely rearrange- 

 ments among the molecules, involving slight changes in the 

 crystalline form or symmetry. 



But on the other hand we may consider the rapid reversible 

 inversions as the fundamental ones, possibly occurring even with- 

 in the atomic nucleus itself, and thereby necessarily producing 

 some corresponding change in the spacing of the atoms. The 

 slow and sometimes irreversible changes would then be the less 

 fundamental, involving only a re-grouping of atoms or atom 

 complexes. 



The second of these two views seems to me to be supported 

 by the phenomena of polymorphism of the oxides of iron. It is 

 obvious that the change can not be one that occurs in the iron 

 atom independently of its chemical environment, for the mag- 

 netic inversion occurs at different temperatures in the different 

 oxides and carbides. Furthermore, corresponding inversions 

 are yet to be found in various other compounds of iron. But 

 there is one oxide of iron which I have not yet touched upon, 

 whose properties shed considerable additional light on the prob- 

 lem. I refer to the magnetic ferric oxide which was discovered 

 by Robbins 19 in 1859, was almost forgotten until recently studied 

 by Hilpert, co and on which we are now engaged in further ex- 

 periments. 



19 Robbixs, J. Chem. News, 1: 11-12. 1859. Its discovery has usually 

 been ascribed to Malaguti, F. Compt. rend., 55: 350-352. 1862. 



20 Hilpert, S. Ber. deu. phys. Ges., 11: 293-299. 1909. Hilpert's rather dis- 

 appointing conclusion from his excellent experiments is that "Fc4).-s functioning 

 as an acid is the carrier of magnetism." But what is "Fei0 3 functioning as an 

 acid" in a compound whose total composition is expressed by the formula Fe 2 3 ? 

 Furthermore, as Hilpert himself recognizes, not all "ferrites," in which Fe 2 3 

 is assumed to be the acidic oxide, are ferromagnetic. 



