Brown.] ^^0 [Jlayl?, 



The important point broiiglit out in these experiments is that pyrite 

 contains a large amount of feiric iron, while in uiaroasite the iron appar- 

 ently exists in the ferrous condition. Some reducing action mig^it, how- 

 ever, have taken place, due to the sulphides formed. The condiiioa of 

 the iron in the chlorides found in the boat and tube was very carefully 

 tested by several reagents in each case, and there can be no doubt as to 

 the correctness of ihe results as stated above. 



The decomposition of the sulphides by metallic salts seemed to offer 

 some hope of being productive of results that would show in a qaantita- 

 tive way the exact amounts of ferrous or ferric iron that are present in 

 these two minerals. In this line, the action of gold chloride, silver 

 nitrate and silver sulphate were tried in a qualitative way with both min- 

 erals. Of these, the first gave a ready decomposition with both, and 

 produced both ferrous and ferric salts io each case. The silver nitrate 

 gave a similar result. Silver sulphate acted very slowly and without any 

 definite results. 



The action of copper sulphate in neutral solution and under pressure 

 was tried with very remarkable results. At the ordinary temperature 

 and pressure the solution of this salt has little effect on either mineral, 

 and the same is true of the solution at a boiling temperature, but under 

 pressure the reaction is complete. The experiment was conducted as 

 follows : 0.2 gram of the finely pulverized mineral was introduced into a 

 stout glass tube, and 50 c.c. of a 10 per cent, solution of the salt, CuSO^"- 

 SHjO- added, the air displaced with a pinch of NaCOg and a drop or two 

 of H.^SO^ (dilute), and a heavy seal made on the tube. The tubes con- 

 taining the two minerals were heated for six hours in an autoclave 

 to a temperature of about 200°. The contents of the tubes were found 

 to contain no traces of undecomposed mineral, but there was a black, 

 more or less fiocculent precipitate in its place. This proved to be copper 

 sulphide. Tlie solution had not altered appreciably in appearance. The 

 liquid contents of the tube were in each case transferred to a flask pre- 

 viously filled with C0.2 and with 10 c.c. dilute sulphuric acid in the bot- 

 tom, the tube then rinsed with water and the amount of ferrous iron 

 present titrated with freshly standardized potassium permanganate. In 

 the case of marcasile this gave 18 c c. KMnO^ solution (this was two- or 

 three-tenths of a cubic centimeter too much, on account of the difiicully 

 in catching the end reaction). To correct this for the iron in the copper 

 sulphate, a blank of 50 c.c. CuSO^ solution, the same as used above with 

 10 c.c. dilute sulphuric acid, was titrated with the permanganate, giving 

 0.5 c.c. reduction. Tiie factor of the permanganate was .0054 gm. Fe for 

 1 c.c. Making the correction for the reduction of 50 c.c. CuSO^ solution, 

 this gives 47.25 per cent, of iron in solution as against 4G.67, the theoreti- 

 cal amount in FeS.^. No doubt if the end reaction had been more exact 

 there would have been a still closer correspondence in the result. 



The tube containing the pyrite was treated in exactly the same manner, 

 and gave a reduction of permanganate of 3.8 c.c. This time the end 



