65 
wJiicli depend on W ':cde Affiriities. 
■^liich the neutral sulphate of alumina bears to the subsulphate. 
It is evident, from the very operation by which this compound is 
produced, that it must have an excess of its bases. If, on the 
other hand, one portion of caustic potash be made to saturate it- 
self with pure alumina, and another portion with silica ; and if 
the former of these solutions be poured, drop by drop, into the 
latter, without however precipitating all the silica j we shall ob- 
tain an abundant precipitate, which contains alumina and po- 
tash in the same mutual relation as the preceding compound, 
but united with a double portion of silica ; in other words, the 
silica contains twice the oxygen which 'exists in the bases. And 
this is exactly the composition of vesuvian. I have not yet at- 
tempted to produce felspar by artificial means; but I have 
reason to believe, that it could be accomplished by digesting the 
precipitate just mentioned in a solution of silica by water It 
is very evident that our knowledge of minerals would gain a 
much higher degree of certainty, if by synthesis w'^e could con- 
firm our analytical results. 
In several double siliciates, we find one of the two component 
siliciates saturated w^itli silica, in a higher degree than the other. 
Such, for example, is the composition of emerald, in which the 
glucina contains a quantity of silica double of that combined 
with the alumina = Be Si'’^ + 2 Al Si^. In apophyllite, the po- 
tash is combined with a dose of silica double of that with whicli 
the lime is combined = K Si^ -f 8 Ca Si^.f It would be a simi- 
* This expression requires explanation. I have shewn {Afh. i Fys. Kern. ^c. 
tom. V. p. 499) that, if crystallised boracic acid be made to absorb a quantity of 
fluo- siliceous acid, and if these acids be then extracted by means of caustic ammo- 
nia in great excess, the silica which remains is visibly soluble in water, and 
plainly decreases by lixiviation. No fluoric or boric acid can be discovered in it ; 
and the water which holds it in solution, does not act upon the vegetable colours^ 
It does not lose its solubility by boiling ; but, when dried, it is soluble no longer. 
If, in a vessel of water saturated with silica, two small cups, the one containing 
ai'tificial vesuvian, and the other silica in a state of solubility, were placed conti- 
guous, it may be presumed that the former would be converted into felspar; 
since the silica, which it might extract from the water, would be continually re- 
placed by the latter. 
+ M. Berzelius is generally known to have demonstrated, that, (with the 
exception of those substances which have azote or phosphorus for their base) 
VOL. I. NO. I. JUNE 1819. ^ 
