146 Pottox—On the Extraction of Glucinum from Beryl. 
precipitated by ammonia, and weighed as usual. If required, the glucina is 
obtained from the filtrate by evaporation in a platinum basin, decomposition 
with sulphuric acid and potassium bisulphate, then dissolved, and the solution 
precipitated by ammonia. 
Berzelius first pointed out that the double fluoride of aluminium and 
potassium, obtained by adding a solution of acid potassium fluoride to a solution 
of alumina in hydrofluoric acid, has the composition AlF;2KI, which com- 
position my analysis entirely confirms; he also investigated the precipitate 
obtained by adding the solution of alumina in hydrofluoric acid to the potassium 
acid fluoride, when a different and more gelatinous and gummy precipitate is 
obtained, having the composition AIF;3KF; there is, however, no danger of 
forming this precipitate if the analysis is conducted as described. 
If the dilution is not of the degree specified, glucinum double fluoride will 
also be precipitated, as it is not very soluble, and requires at least fifty times its 
weight of water for solution. For this reason, the method, though convenient for 
analysis, is not at all convenient for preparation of a large quantity of glucina ; 
but if the hydrate or carbonate of glucina be nearly free from alumina, it may 
be made perfectly pure by dissolving in hydrofluoric acid, and then adding a 
quantity of potassium acid fluoride, just sufficient to precipitate all the alumina 
present, or a little more, when all the alumina will come down, and some glucina 
with it. This avoids the necessity of greatly diluting the solutions, and the use 
of very large platinum vessels. If excess of potassium acid fluoride were added 
to a strong solution of glucina, most of the glucina would precipitate, as well 
as the alumina. After the removal of the precipitate containing the double 
fluoride, the solution is evaporated down to a gummy consistency, decomposed 
with sulphuric acid, heated until dense fumes come off, then dissolved, and the 
glucina separated as hydrate or basic carbonate, by ammonia or carbonate of 
ammonia. The method, though inconvenient, is the only one I am acquainted 
with, by which one can obtain chemically-pure glucina, free from every trace 
of alumina. 
VI. Guuctnum Basic CARBONATE. 
To purify the basic carbonate of glucinum, I decided to use the method of 
fractional solution in ammonium carbonate described by Weeren in his researches ; 
that is, to add a quantity of ammonium carbonate less than sufficient to dissolve 
the basic carbonate, as it is said that under these conditions only pure glucinum 
carbonate dissolves, and any alumina present remains undissolved. 
To ascertain the solubility of the glucinum basic carbonate in ammonium 
carbonate, for the purposes of the experiment, six grams of freshly precipitated 
