i 
! 
| 
i 
i 
Genth.] 696 [Oct.17, 
cooling a small quantity of ammonium chloride crystallized out. It 
was boiled briskly for.a whole day, and did not show any bumbing 
at all; a large insoluble residue remained. It had a slightly yellow- 
ish color from a little iron, but had not the slimy gelatinous appear- 
ance which alumina would have présented, but, on the contrary, it 
was, without being erystalline, more granular and resembled beryl- 
lium hydrate precipitated on boiling from a dilute solution in, sodium 
hydrate. The filtrate was precipitated with: ammonia and gave 5.61 
per cent of ignited beryllium oxide. That portion insoluble in am- 
monium chloride dissolved readily in a minute quantity of dilute hydro- 
chloric acid. Sodium hydrate was now added and the whole evaporated 
in a platinum dish to a pasty consistency, then diluted with cold water 
and the little ferric hydrate filtered off, which was subsequently dissolved 
in hydrochloric acid, precipitated by ammonia. In washing the iron 
precipitate the filtrate became gradually more and more turbid. It was 
still more diluted with water and boiled for about one hour, when it was 
filtered off and washed; it gave 8.99 per cent.* The filtrate was now 
acidulated with hydrochloric acid and precipitated by ammonia and gave: 
0.17 per cent of alumina, which was dissolved in hydrochloric acid and 
precipitated by ammonia and was found to be insoluble in ammonium car- 
bonate. ‘ 
The different constituents were added together and gave the results 
given below. A considerable quantity of phosphoric acid was Jost proba- 
bly in the zine silicate before I was aware of the fact that this retained such 
a large portion of it, and only 41.76 per cent were obtained. 
II. 0.8608 grm. of finely powdered perfect crystals, dried over one week 
over sulphuric acid, were fused with twice their weight of silica and 
sodium carbonate, and the mass treated asin I. It was found, however, 
that for this mineral this method, even with such an excess of silica, can- 
not be used advantageously, as the phosphoric acid, notwithstanding the 
most careful operations, was contaminating almost every precipitate and 
was largely contained in the zinc silicate. After the greater portion of the 
phosphoric acid was obtained, there was still about 4.5 per cent with the 
glucina and ferric oxide, from which it was separated by ammonium 
* This unexpected behavjor of glucina with a boiling solution of ammonium 
chloride induced me to make the following experiments: A quantity of 
beryllium carbonate, prepared from beryl], from Acworth, N. H., was dissolved 
in hydrochloric acid, to the strongly acid solution ammonia was added until it 
showed a slightly alkaline reaction, It was boiled briskly for one day, then fil- 
tered, the undissolved residue was again dissolved in an excess of hydrochloric 
acid, precipitated by ammonia and boiled for another day and filtered. The two 
filtrates were mixed and precipitated by ammonia and gave 0,1158 grm. of bery1- 
lium oxide, The insoluble residue was dissolved in just enough hydrochloric 
acid precipitated and redissolved by sodium hydrate, diluted with much water 
and boiled for one hour, The precipitate weighed after ignition 0,1422 grms, 
Both were converted into anhydrous sulphates, the first furnishing 0.4742 grm, 
containing SO, = 75,58 and BeO == 24,42; the second gave 0.5912 grm. sulphate of 
beryllium with SO, =75.95 and BeO = 24.05, 
