110 QUICKSILVEU DEPOSITS OF THE PACIFIC SLOPE. 



No serpentine which remains dark between crossed nicols is found in 

 the slides from the Coast Ranges, and excellent cleavage pieces mounted 

 in balsam give biaxial interference figures and show that the bisectrix and 

 the j)lane of the optical axes are perpendicular to the cleavage. The angle 

 of the optical axes in the cases thus examined is not ver}- small. In cross- 

 section the extinction always takes place exactly parallel to the traces of 

 the cleavage. 



The colors of polarization vary greatly. They are sometimes confined 

 to different shades of gray; often, however, i)ortions of slides in which gray 

 is the prevalent color show dull, yellow tints, and in rather exceptional 

 cases reddish tints are visible. This variation is frequent in slides of irre- 

 proachable thinness and uniformity and does not depend on a mere dilfer- 

 ence in the thickness of the rock section. In grinding slides, however, it 

 is found, as would be supposed, that thick masses give more l^rilliant colors 

 than thin ones. The serpentine usually shows the merest trace of dichroism, 

 but occasionally a very perceptible change of color may be observed. Only 

 one case is known (No. 21, New xVlmaden) where a remarkably vivid -green 

 serpentine is strongly dichroitic and might readily be confounded with 

 chlorite unless examined between crossed nicols. A particularly pure-look- 

 ing, light-green, marmolitic serpentine (No. 1 1 0, New Idria) was selected 

 for investigation. A complete analysis w^as made, with the following- re- 

 sult : 



Silica, SiO- 41.o40 



Magupsia, MgO 40.420 



Water, H-O 14.175 



A 1 u m i u a , Al- 0» ii . 4S0 



Ferrous osi'le, FcO 1.370 



Nickel oxide, NiO 0. 040 



100. 035 



In pure serpentine 40.42 per cent, of magnesia corresponds to 41.52 

 per cent, of silica. It appears, therefore, that this mineral is in fact a 

 serpentine comparatively free from impurities and certainly containing no 

 talc. A slide was also cut across the lines of structure at a point where 

 the specimen appeared extremely uniform. Wlien reduced to the proper 

 thinness it was fi)und that the material was far from homogeneou.s. A por- 



