58 sosman: problems of the oxides of iron 



our experience, powdered iron heated in oxygen at low pressures 

 always yields a mixture of a black magnetic oxide (perhaps 

 a solution of FeO in Fe 3 4 ) and metallic iron. The facts are 

 so few, however, that this explanation is not to be considered 

 as more than a suggestion. 



There is one interesting geological application of this relation, 

 if it should be found to exist. Many basic rocks, such as dia- 

 base and gabbro, contain metallic iron. The best known occur- 

 rence is probably that in Greenland, but other examples exist 

 in all parts of the world. If our supposed oxygen-pressure 

 relationship still holds when the oxides are dissolved in a silicate 

 magma, the following reaction would occur: 



4FeSi0 3 (in complex silicate^FesCX + Fe + Si0 2 (in complex 



silicate) 



At constant temperature, increasing pressure would drive 

 this reaction in the direction of diminishing volume. The 

 formation of FeO from Fe and Fe 3 4 is accompanied by a con- 

 traction of volume, according to Hilpert's data, and it appears 

 not unlikely, therefore, that release of pressure in a rising column 

 of molten diabase might alone account for the occurrence of 

 metallic iron in the resulting rock, without the need of recourse 

 to reducing agents such as entrapped organic deposits or dis- 

 solved reducing gases. This problem calls for the measurement 

 of dissociation pressures and specific volumes of the oxides and 

 silicates concerned, in the region 900°-1300? 



The same reaction might conceivably be made a basis for a 

 commercial method of making pure iron, free from the usual 

 impurities which come from the use of coke as a reducing agent. 



Continuing our survey in the direction of increasing content 

 of oxygen, we come to the compound ferrosoferric oxide, magne- 

 tite (Fe 3 4 ). Its melting point is 1580°, and it melts sharply 

 to a mobile liquid which crystallizes in octahedra on cooling. 

 Its dissociation pressure is extremely low, being less than 0.04 

 mm. of mercury at 1200°, and less than 0.005 mm. at 1100? 



Further addition of oxygen brings us into the region between 

 Fe 3 4 and Fe 2 3 , a region which we have recently investigated 



