50 JV. L. Bowen — The Composition of Nephelite. 



and corundum (A1 2 3 ). But the loss of soda gave not only 

 excess alumina but also excess silica, which, however, did not 

 appear in any form, and therefore must have been held in 

 solid solution in the nephelite. By mere chance, then, the 

 explanation of the excess silica content of nephelite was hit 

 upon. It is held in solid solution and is therefore variable, 

 but no potash content is necessary. 



This fact suggests that in natural nephelites potash and 

 excess silica may be quite independent of each other. An 

 examination of analyses proves that it is indeed so. This 

 examination is readily accomplished with the aid of the follow- 

 ing diagram, where composition is plotted on triangular coordi- 

 nates with Si0 2 ,K 2 O.Al 2 3 and Na 2 0, A1 2 3 as apices. Only the 

 significant segment of the plot is shown. CaO has been 

 deducted from the analyses together with sufficient A1 2 3 and 

 Si0 2 to form CaA! 2 Si 2 8 . ^ 



Each small numbered circle represents the composition of a 

 nephelite. There is no apparent arrangement to the points. 

 Obviously nephelites cannot be considered as a ' mixture' of 

 soda nephelite with any potash compound, for all the points 

 would then lie on the straight line joining these two composi- 

 tions. Potash and silica vary quite independently. 



An experimental study of the composition of nephelite was 

 begun by the writer by making up mixtures in which there 

 was excess silica (deliberate) but no excess alumina. The com- 

 position Na 2 0,Al 2 3 ,2.2Si0 2 (A of diagram) was crystallized 

 below 1200°, the temperature being kept low to prevent the 

 formation of the high temperature form of nephelite (carne- 

 gieite). The product was a completely crystalline perfectly 

 homogeneous mass with properties essentially those of the 

 orthosilicate (Na 2 0, Al 2 3 ,2Si0 2 ). The excess silica had simply 

 disappeared in solid solution, as had been accidentally found 

 before. 



Mixtures with still greater excess of silica were experimented 

 with in the attempt to find the limit of solubility and the 

 excess phase after solubility was exceeded, but the increasing 

 viscosity, and the necessity of crystallizing at a low tempera- 

 ture, prevented the formation of a determinable crystalline 

 product. We have apparently a scanty amount of experi- 

 mental results on which to base any conclusion, and yet when 

 used in conjunction with facts known to be true of rocks, cer- 

 tain conclusions may be drawn. 



It is more than probable that, had the attempts above men- 

 tioned been attended with success, albite (Na 2 O.Al 2 3 .6Si0 2 ) 

 would have been found as the excess phase. By analogy, then, 

 with an ordinary aqueous solution of salt we would speak of 

 our solution as one of albite in nephelite. The excess phase 



