189^.] -^4.1 [Browu. 



reaclioa was sliarp and exact. Calculating the above to iron (after mak- 

 ing correction for CiiSO^) this gives 8.91 per cent, of ferrous iron in 

 pyrite. As the total iron Is 46.67, this corresponds to 19.09 per cent, of 

 the iron in the mineral, or almost exactly one-flfih. These experiments 

 demonstrate iu a positive manner tlie condition of the iron in the two 

 minerals, and even show the exact amounts of each condition of the 

 iron, ferrous and ferric* 



That marcasite should hence be more readily decomposed by oxidation 

 than pyrite seems fully explained by the foregoing investigations, as it 

 consists of Fe"S.^, an unsaturated compound. In this compound, sulpliur 

 must link to sulphur, or the compound have unsaturated bonds, and 

 hence any element which would attack the sulphur would break up the 

 compound. On the other hand, the iron is held to the sulphur by iis full 

 number of bonds, and any substance that has an affinity for iron could 

 not so readily attack it in tliis condition. This wo\ild be true whether 

 ferrous iron be considered here as Fe^, with a valence of four, or as Fe". 

 That marcasite is Fe"S.^ is also indicated by its oxidation in the air into 

 FeSO<t mamly. Under these same conditions it will be noted that pyrite 

 forms both ferrous and ferric compounds, as FeSO^. but much more 

 Fc^Oy (OH)g and free sulphur. ^larcasite, however, when decomposed 

 by Water under pressure (in nature) forms much limonite also, this being 

 due no doubt to the oxidation being effected under pressure. This con- 

 stitution explains also the fact that the oxidation of marcasite is continu- 

 ous and complete, as shown by the current oxidations. It will be shown 

 also that this constitution of pyrite that has been made out explains fully 

 its action with the current. That marcasite is unsaturated is also indi- 

 cated by the fact that it has not been made artificially or at any rate posi- 

 tively identified in any of the artificial FeS.^ that has thus far been made. 

 If marcasite be a persulphide, as its formula would seem to indicate for a 

 ferrous compound, none of the methods detailed above for making FeS.^ 

 would be applicable in its case, unless perhaps the method of Deville 

 might produce it. All of the other methods would probably produce 

 ferric iron, at least iu large part, and tlie resulting product would be 

 pyrite. 



The formula for pyrite derived from my investigations and expressing 

 the relation of the two conditions of the iron in the simplest way is 

 4 Fe'vS2' Fe"S.2. This formula is also borne out by what we know of the 

 formation of pyrite as given in the early part of this paper, and by such 

 experiments as I liave made on its decomposition, as well as the fact above 

 alluded to that it, in oxidizing in nature, does not form much ferrous 

 compounds but mainly terric. And it also exi>lains the fact that it is more 

 stable as regards any element attaclilng its sulphur, for it is most probable 

 that all the sulphur of the Fe"S2 in its formula is linked to iron. I would 

 propose the following structural formula, not as expressing the exact con- 

 stitution of the compound, for of that we know nothing, but as an 



*Tliese results have been confirmed l3y experiments. made during tlie past year in this 

 laboratory, and not yet publisned. 



