60 sosman: problems of the oxides of ieon 



Several interesting problems are raised by the consideration 

 of this type of solid solution. In the first place how are we to 

 picture to ourselves its internal structure? The X-ray analysis 

 of crystal structure is so radically altering our conceptions of the 

 make-up of solid substances that our ideas of solid solutions have 

 not yet become adjusted to the new facts. When it was dis- 

 covered that certain properties of a solute in dilute aqueous 

 solution made the solute seem quite analogous to a gas, and 

 when it was discovered that solid solutions existed which seemed 

 quite analogous to liquid solutions, we felt secure for a time in 

 this extension of molecular theory from gases over into liquids 

 and solids. The facts now need re-interpreting, but the greatest 

 need is for more facts on the crystalline structure of solid 

 solutions. 



Another problem raised by a consideration of the hematite- 

 magnetite series is that of the continuous transition from one 

 crystal class into another. According to the original conception 

 of isomorphism, two compounds could enter into solid solution 

 only if they crystallized in the same system. Now hematite 

 is hexagonal* while magnetite is isometric; is a continuous series 

 from one of these systems to the other possible? To say that 

 there is a hexagonal form of magnetite, which is the form that 

 dissolves in hematite, is merely to dodge the issue; such a state- 

 ment becomes a meaningless form of words if the experimental 

 consequences remain the same whether the supposed second 

 form exists or not. 



A consideration of the point systems from which the crystal 

 classes can be made up shows that we can in reality get a con- 

 tinuous transition from cubic to hexagonal. Suppose a cubical 

 portion of some cubic lattice to be standing on one of its corners ; 

 then if it be compressed along the vertical diagonal axis it 

 changes into a rhombohedron which becomes flatter with increas- 

 ing compression, and the rhombohedron is a hexagonal form. This 

 transition requires of course that the lattice of the isometric and 

 hexagonal forms be thus transformable, and this can be de- 

 termined for hematite and magnetite only by X-ray studies 

 of their crystals. If it should then be proved that the struc- 



