NO. 2220. COLORED VARIETIES OF QUARTZ— WATSON cf- REARD. 559 



However, in light of some recent quantitative work on amethyst 

 carried out by George Steigcr in the chemical laboratories of tho 

 United States Geological Survey, there is proof that the amount 

 of manganous oxide reported in tho analysis below is sufficient to 

 produce tho depth of color shown in the mineral which he analyzed. 

 The mode of combination of the manganese is not necessarily tho 

 same. The amount of manganous oxide found by Steiger is less 

 than that determined by the writers in the four specimens tabulated 

 on page 554. 



The results of tho investigation by Steiger as prepared by him are 

 published below with his permission. 



A. Rose quartz from Creede, Colorado. 



B. Clear, fairly dark colored crystals of amethyst, locality unknown. \ 



MnO. 



0. 0002 per cent. 0. 00012 per cent. 



The minerals were finely powdered , 30 gms. cf each were decomposed with sulphuric 

 and hydrofluoric acids and evaporated, llie heat being continued until the fluorine 

 had been completely expelled, diluted with water, and filtered. 



Silver nitrate was added to the filtrate, manganese oxidized with ammonium 

 persulphate, and determined by comparison of the color of this solution with a standard 

 solution of manganese. 



In preparing the first samples a steel block was used in powdering. It was thought 

 manganese might have been introduced through small fragments cf steel, and accord- 

 ingly a second set of ground samples were prepared, this time using an agate mortar 

 in their preparation. Determinations in the second set checked with the first as 

 closely as could be expected. 



The following experiment is interesting in sho'wing that such colors as are observed 

 in these minerals can easily be accounted for by the quantities of manganese present. 



To 12 cc. cf water (this amount roughly representing the bulk of 30 gms. of quartz) 

 was added one-tenth cc. cf permanganate solution containing O.OOOG gms. cf MnO 

 per cubic centimeter. This solution contained an equivalent amount of manganese 

 per unit volume as the rose quartz; the color produced was of a much darker shade 

 than that of the mineral. A similar experiment was carried out, using an equivalent 

 amount of manganese to that found in the amethyst, and in this case the color of the 

 solution was about that of the mineral. 



It is probable that the percentage amount of either of the three 

 metallic oxides found on analysis (table of analyses, p. 554) is suf- 

 ficient to produce disperse color in quartz. From the known 

 chemical behavior of these substances when small portions are 

 taken up on heating by other bodies, such as borax, salt of phos- 

 phorus, glass, etc., manganese produces more exactly the color of 

 amethyst than either the oxide of iron or titanium. In the bead 

 tests with borax and salt of phosphorus, manganese oxide yields 

 amethystine color only when heated in the oxidizing flame, becoming 

 colorless when heated in the reducing flame. The usual explanation 

 for these reactions is that in the former case manganese is converted 

 into a higher form of oxidation and in the latter it is changed by 

 reduction to a lower form — MnO. From the carefully conducted 



