178 



A lien. 



Crenshaw, Johnston, and Larsen- 



senting the oxidation coefficients for all possible mixtures. 

 His curve had the form of a eutectic curve with the lowest 

 point at ten per cent, pyrite having an oxidation coefficient of 

 about 15. Fig. 4 shows the curve which we obtained for mix- 

 tures of pyrite and marcasite. Careful experiment failed to 

 reveal any mixture which gave a lower oxidation coefficient 

 than pure marcasite. The difficulty of determining accurately 

 the composition of an unknown mixture, by a measurement of 



Fig. 4. 



60 



X 50 



o 



o 

 2 40 



.2 



2 30 



O 



20 



10 



^! 



llllll 



_- 



10 20 30 40 50 60 70 80 90 100 



Marcasite. Pyrite. 



Fig. 4. Curve showing the oxidation coefficients for mixtures of pyrite 

 and marcasite. 



the oxidation coefficient, is obvious from the form of the curve ; 

 indeed, differences of two to three per cent (of the weight of 

 the mixture) were usually found in duplicate determinations. 

 Yet, though we cannot confirm the statement of Stokes regard- 

 ing the accuracy of the method, we have found it of great 

 value in studying synthetic products of iron disulphide, the 

 mineral composition of which must otherwise have remained 

 unknown, and have been thus enabled to work out the essen- 

 tial conditions which govern the formation of pyrite or marca- 



