106 



wherry: phosphate and silicate minerals 



seemed probable that the silica might be due to lamellae which 

 are present but unrecognizable because of their thinness, an 

 attempt was made to determine the composition of the white 

 lamellar mineral. It proved impracticable to separate the 

 lamellae from the green ground-mass with any. degree of com- 

 pleteness, but a very small sample, containing perhaps one-third 

 of the latter, was analyzed by the writer with the following results : 

 CaO+CuO 9.0, MgO 0.5, Al 2 3 +Fe 2 3 23.3, P 2 5 12.1, Si0 2 

 30.0, H 2 below 100° 10.4, above 100° 14.8, sum 100.1. 



Although amorphous colloidal minerals like these do not 

 necessarily possess definite formulas, it seemed worth while to 

 attempt to determine at least their approximate nature. The 

 known aluminium phosphate minerals fall into four divisions 

 with reference to the ratios of the A1 2 3 to P 2 5 , as shown in the 

 following table: 



The ratio Al 2 3 (+Fe 2 3 ) : P2O5 shown in analysis 1 is 3.1: 2, 

 so that this mineral evidently belongs to the second of these 

 divisions. It is also very high in water, and accordingly lies 

 toward the bottom of the table. The possibility of its identity with 

 vashegyite must therefore be considered. The latter mineral was 



