﻿214 Clarke and Diller — Turquoisfrom J^ew Mexico. 



In the Calif or nian turquois the analytical results fit these 

 formula? quite sharply, and give the ratio between the two 

 compounds as approximate!}' four to one. The first formula 

 may be regarded as that of normal turquois, and may be 

 written in rational form, halved, as 



Al a HP0 4 (OH) 4 . 



The copper salt, to which the mineral owes its color, is to be 

 considered merely as an impurity : a view which is emphasized 

 by the analysis of the dark-green turquois, "C." In the 

 latter case the same ratios apply, modified by the presence of 

 silica, which is nearly sufficient to form with, the copper a 

 normal metasilicate, similar to if not identical with chrysocolla. 

 This silicate, with whatever blue tinge of color it might have, 

 affected by the yellow or brown of the iron present, probably 

 gives the turquois its green hue. It is exceedingly probable 

 that the purity of tint in gem turquois is due to the copper 

 salt alone : and that degradations of the color towards green 

 are ascribable to admixture of salts of iron. It is noteworthy 

 that, of the three turquoises analyzed, the bluest contains the 

 lowest percentage of copper. This could hardly be the case 

 were not the colors of the other samples modified by some 

 impurities : and compounds of iron would naturally produce 

 an effect in the observed direction. 



Sections of the three varieties of turquois were studied 

 under the microscope, and found to be of essentially the same 

 character. Although deeply colored in the hand specimens, 

 the thin sections appeared almost clear and transparent. 

 Between crossed nicols the deep blue and green forms were 

 seen to be composed of minute grains, or short, thick fibers; 

 but in the paler varieties the fibrous structure was more pro- 

 nounced. The optical properties of both grains and fibers are 

 the same throughout. They are all weakly doubly refracting, 

 but have a rather high refractive index. The finely granular 

 portions have a pale bluish aggregate polarization, less intense 

 than that of chlorite : but when the mineral is distinctly fibrous 

 it polarizes like some forms of serpentine, with light colors of 

 the first order. The fibers are generally somewhat bent and 

 interwoven, but lie approximately in the same direction. Each 

 fiber becomes dark when parallel to the principal section of 

 either of the crossed nicols, indicating that they must crys- 

 tallize according to either the quadratic, hexagonal, or rhombic 

 system, instead of in one of the inclined systems as was the 

 case with the fibers studied by Bucking* in the turquois of 

 Fresno Count}', California. 



A section was prepared of a distinct vein of pale-green tur~ 



* See paper by v. Zepharovich and Moore, previously cited. 



