M. Scheerer's Observations on Elccolith and Nepheline. 117 



and 2 at. soda and potash. These elements can in no way be 



Na 2 f- 



more simply combined than in the formula it' 2 < Si + 2AlSi; 



or, if regard be had to the proportions of potash and soda 



(Ka 2 Si + 2 Al Si) 4- 4 (Na 2 Si + 2 Al Si). That in these 

 formulae the alkali (semi-silicate) is combined with more silica 

 than the alumina (l silicate) cannot appear objectionable ; it 

 is, indeed, plain that this must always happen, if, on the forma- 

 tion of a mineral, there is not sufficient silica present fully to 

 saturate all the bases; these will then divide the quantity of 

 silica among themselves according to their affinity. Thus, for 

 instance, feldspars, as a constituent part of granite, is always in 

 contact with superfluous silica; potash and alumina which it 

 contains are perfectly saturated with it. But the elaeoliths, 

 whether in syenite, miascite, nepheline-rock, albite, &c, do 

 not occur accompanied by free quartz. There was, therefore, 

 a want of silica at their formation, and the potash accordingly, 

 on account of its great affinity, took up a greater portion of it 

 than the alumina. 



The proportion of the oxygen in the alkali, relative to that 

 of the silica, as 2 : 3 is, moreover, not wholly new in mine- 

 ralogical formulae ; for, according to the investigations of 

 Walmstedt*, repeated by Regnaultf, prehnite should have 



the formula Ca 2 Si + Al Si -f H ; but very probably a similar 

 relation occurs in many other minerals where the analysis does 

 not agree well with the formula. I will not here mention my 

 suspicions, but only give one instance, which speaks evidently 

 in favour of my opinion. This is, the cancrinite, recently 

 analysed by Professor Gustav Rose J. Two analyses gave 



Silica . . . 



40-59 



40-26 



Alumina . . 



28*29 



28-24 



Soda . . . 



17*38 



17*66 



Potash . . . 



0-57 



0-82 



Lime . 



7-06 



6-34 



Carbonic acid 



6*38 



6*38 



100*27 99-70 



Professor G. Rose advances accordingly the following for- 

 mula, of which, however, he himself adds that it does not 



agree well : Na 3 Si + 3_A1 Si + Ca C. 



* Berzelius, Jahresbericht, No. V. p. 217. 



f Ann. ties Mines, Ser. III. T. xiv. p. 154. 



\ See the above-mentioned Memoir on the Ilmengebirge. 



