392 Allen and Crenshaw — Stokes Method. 



(7-15) than the amount required by theory. With such an 

 excess the percentage of the surfaces remains on the average 

 nearly the same as the percentage by weight, the basis on 

 which the mixtures are made up. About 1 g. is sufficient for 

 250 cc of the standard solution. Secondly, the marcasite has a 

 characteristic tendency to flocculate and thus reduce its react- 

 ing surface. This difficulty may be avoided by shaking the 

 reaction mixture with pure quartz and beads until the lumps 

 of the powder are thoroughly disintegrated. Different speci- 

 mens of pyrite and marcasite give with the Stokes reaction 

 values which differ somewhat. The differences are due in 

 some cases, if not in all, to the presence of impurities. It is 

 unfortunate that small quantities of impurities which will 

 reduce ferric iron or give up iron to the solution exercise a 

 serious influence. It is therefore not always possible to decide 

 between a natural pyrite and a pyrite containing several per 

 cent of marcasite by the Stokes reaction alone, nor to deter- 

 mine accurately the percentage of each in a natural mixture. 

 In an investigation on the conditions of formation of pyrite 

 and marcasite, this method has been very useful. 



2. The results with the Stokes method plainly indicate that 

 each mineral behaves in a mixture of the two just as it does 

 alone ; each appears to reduce a quantity of solution which is 

 proportional to its surface ; and each appears to reduce the 

 solution at practically the same rate. The rates at which the 

 sulphides are decomposed is quite different for the two min- 

 erals, because more of marcasite than of pyrite is required to 

 reduce a given quantity of ferric iron. The ratio of these 

 rates is not far from 1 : 2*5. 



3. That ferric sulphate dissolves from pyrite a smaller 

 quantity of iron than it does from marcasite means simply 

 that more reduction is effected by sulphur in the case of pyrite, 

 in other words that more of the sulphur in pyrite is oxidized. 

 Stokes considered only the relation of p, the percentage of 

 sulphur oxidized, to y, the percentage of pyrite in the sulphide 

 mixture. We have shown that this curve is a hyperbola. 

 This characteristic behavior of pyrite and marcasite towards 

 oxidizing agents is probably general. It has been found by 

 other observers that nitric acid and hydrogen peroxide both 

 oxidize more of the sulphur in pyrite under the same con- 

 ditions. 



The authors wish to thank Prof. E. S. Dana of Yale Univer- 

 sity and Drs. Geo. P. Merrill and Edgar T. Wherry of the 

 National Museum for specimens of pyrite and marcasite used 

 in this work. 



Geophysical Laboratory, 



Carnegie Institution of Washington. 



