560 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.53. 



experiments by Simon on amethyst, the color, if due to manganese — 

 and we believe that it is — seems to be independent of the processes 

 of oxidation and reduction, since he found it was destroyed by 

 heating the mineral in either an oxidizing or a reducing atmosphere, 

 and likewise could be restored on exposure to radium rays in an 

 atmosphere of either oxygen or hydrogen. 



From the investigations thus far made into the cause of color of 

 amethyst the following conclusions seem to be warranted: (1) That 

 since the color can be destroyed in an atmosphere of either oxygen 

 or hydrogen and restored in an atmosphere of either, it is inde- 

 pendent of the processes of oxidation and reduction, which definitely 

 eliminate organic substances as a cause of the color; (2) that it is 

 due to some inorganic substance whose nature and constitution are 

 yet to be definitely determined; (3) that of the metaUic oxides 

 shown to be present in amethyst by chemical analysis, manganese 

 seems to be the most likely one to produce the amethystine color; 

 (4) that the analyses of the mineral show a sufficient quantity of 

 manganese oxide present to produce the color; and (5) that colloidal 

 particles of submicroscopic size, probably manganese oxide, will be 

 found upon ultramicroscopic study to be the cause of the color. 



Eose quartz. — A larger number of thin sections of rose quartz from 

 different localities were studied microscopically than of amethyst, 

 but with practically the same results. Irregularly distributed inclu- 

 sions of the same natiire and approximate quantity were observed 

 as in the amethyst, and are not regarded by the writers as having 

 any effect in producing the color. 



Eight partial analyses of rose quartz from as many different 

 localities were made, with the results showTi in table on page 554. 

 The same metallic oxides (TiOz, MnO, FczOg, and CoO) were found 

 as in the amethyst, with somewhat similar variations in each con- 

 stituent noted when individual analyses are compared. (See p. 559 

 of this paper determination of manganous oxide in rose quartz from 

 Creede, Colorado, by George Steiger.) Titanium dioxide (TiOj) 

 averages higher, while manganous oxide (MnO) and ferric oxide 

 (FcjOa) average lower for rose quartz than for amethyst. The color 

 of the borax bead imparted by the very slight amount of cobalt 

 oxide (CoO) present ranged from distinct blue to very pale blue; 

 but on the whole the bead tests indicated probably a smaller content 

 of cobalt oxide (CoO) than for amethyst. 



In every case the mineral was completely decolorized on heating, 

 which was to be expected, since it is well known that the color ol 

 rose quartz fades on exposure to daylight. The wri ars did not 

 carry out any tests to determine the restoration of color of the 

 decolorized specimens, but Doelter^ points out the lac'^ of restoration 



> Das Radium und die Farben, 1910, pp. ?I and 78. 



