Ferguson: oxidation of lava 545 



material to dissolve during the decomposition of the sample 

 with hydrofluoric and sulphuric acids, and if the formation of this 

 mineral were at the expense of the oxygen in the remainder of 

 the lava, the analytical results would probably not indicate this 

 but would rather show merely less ferrous iron. The mechanism 

 of such a reaction within the lava with the formation of magnetite 

 can only be speculated upon. Sosman^^ has suggested the possi- 

 biUty of the presence of iron (Fe) in solution in the glass or min- 

 erals after this reaction. Another possible reaction would be 

 the formation of a lower oxide of titanium than the Ti02 usually 

 reported. 



The fact that both the fine and coarse material, although 

 of shghtly dififerent initial ferrous iron content, contain prac- 

 tically the same percentages after treatment is probably trace- 

 able to the larger percentage of glass and smaller percentage of 

 Fe304 in the fine material. 



Whatever the cause of the small decrease in ferrous iron, and 

 of the variations therein, the experiments leave no doubt that 

 considerable ferrous iron, when in silicate combinations, can 

 exist in the presence of water vapor at high temperatures. 



The vacuum experiment (No. 20) with the oxidized material 

 shows clearly the ease with which oxygen can be removed from 

 such materials and indicates that due regard must be given to 

 this phenomenon not only in the study of the chemistry of lavas, 

 but also in the interpretation of the results obtained in experi- 

 ments in which gases are pumped from rocks at high tempera- 

 tures. 



The data given in this article are of a qualitative rather than 

 a quantitative character, and are not to be regarded as final data 

 on the reactions of water vapor with iron-bearing silicates. 

 The complete story cannot be obtained by experiments on so 

 complex a material as a natural lava, but must be learned by 

 experiments with simple siHcates of known composition. These 

 results are presented here, however, on account of their current 

 interest and in the absence of any fundamental data on the re- 

 actions concerned. 



IB Private communication; see also this Journal. 7: 58. 1917 



