24 He C. COCKE 
sulphides of these elements. To test this theory, well-crystallized 
specimens of realgar, orpiment, and stibnite were powdered, the 
stibnite to pass an 80-mesh, the others a 4o-mesh screen. The 
stibnite was covered with a saturated (about N/30) silver sulphate 
solution, the realgar and orpiment with N/4o silver nitrate solu- 
tions, after which the mixtures were put away in the dark. The 
stibnite was analyzed after 36 days, the realgar and orpiment after 
68 days, the analyses being conducted by passing a current of dry 
chlorine over the material heated to a low red heat. The silver 
remained in the tube as silver chloride, together with earthy impuri- 
ties; the other substances volatilized as chlorides. After weighing, 
the silver chloride was dissolved with warm ammonia, and the 
weight of the impurities determined. Results were: 
Stibnite contained 1.3 per cent Ag 
Realgar contained 1.9 per cent Ag 
Orpiment contained 9.8 per cent Ag 
Since pyrargyrite contains 60 per cent and proustite 65 per cent of 
silver, the amounts of silver entering into the sulphides in these 
experiments were far below those necessary to give a true silver 
mineral. Moreover, the reaction in the case of the orpiment was 
visibly instantaneous, the powder changing from brilliant yellow 
to black at once. As the solutions used contained much more silver 
than ground-waters commonly do, it would appear improbable 
that true silver minerals are formed in this way. 
The formation of argentite can easily be explained by the action 
of hydrogen sulphide or a metallic sulphide on silver-bearing solu- 
tions. As has been shown by R. C. Wells,’ the action of dilute 
sulphuric acid on pyrrhotite, galena, sphalerite, and other natural 
sulphides produces hydrogen sulphide. This might be supposed to 
react with the silver solution and yield silver sulphide, and, as 
already mentioned, would do so in the case of chloride solutions 
carrying silver. But in the case of sulphate solutions a complica- 
tion is introduced. Hydrogen sulphide would probably precipitate 
the silver as sulphide here also, but it tends in addition to attack the 
ferric sulphate present in the solutions, and to reduce it to the fer- 
rous form with separation of sulphur. The formation of ferrous 
*R. C. Wells, unpublished manuscript. 
