408 Allen, Crenshaw, and Merwin — 



for the results with the Stokes test.* When marcasite which 

 had been subjected to the above treatment was heated in 

 vacuo, it gave off no more gas than marcasite which had never 

 been subjected to such treatment (Table III). The former, 

 therefore, could hardly have contained occlusions which 

 affected the Stokes test. Some change in the marcasite, prob- 

 ably some change in its surface, has doubtless taken place, but 

 when all the facts are considered it seems impossible that there 

 should be any change to pyrite. Equally puzzling is the 

 irregular increase in pyrite from the highest acid concentra- 

 tions ; some unknown condition, perhaps the greater solvent 

 power of strong acid, enables the difference in potential 

 between the two forms to assert itself. 



(d) The contemporaneous precipitation of pyrite and 

 marcasite. — The simplest explanation one can offer for the 

 relation found between the acidity and the percentage of mar- 

 casite in the synthetic disulphides formed from acid solutions, 

 is that the pyrite and marcasite crystallize simultaneously. Our 

 general knowledge of the subject of unstable forms is quite 

 limited, but many cases are known where the unstable form 

 appears first and is followed by a stabler form. In the case we 

 are considering, the stable form is found in larger quantity in 

 the products from low acid concentration and we know the 

 acid increases with the progress of precipitation ; the obvious 

 inference is, therefore, that less pyrite is precipitated in the 

 later stages of the process. It might be supposed that pyrite 

 alone is precipitated until a certain limiting acid concentration 

 is reached, when pure marcasite appears, but this hypothesis 

 is easily disproved, for if the percentage of pyrite in any 

 product is known, as well as the initial and final acid concen- 

 trations which prevailed during its formation, the limiting 

 concentration can be easily calculated and yet a different cal- 

 culated value is found for every product. 



If precipitation occurred entirely at any given temperature, 

 we could say without reserve that the percentage of marcasite 

 in the product was a linear function of the acid concentration, 

 at least within the errors of experiment, and then there would 

 apparently be no escape from the conclusion that both crystal- 

 line forms of the disulphide come down together, the relative 

 proportions at any temperature depending on the average acid 

 concentration. 



This conclusion cannot be hard and fast, however, because, 

 owing to the considerable heat capacity of the bomb and its 

 contents, several hours are required before the maximum tem- 

 perature is reached, and consequently some of the product is 

 precipitated while the temperature is rising (see fig. 2). It 



*See p. 375 of this number. 



