Experiments on Allanite. 283 
the filter, and mixed with a solution of sal ammoniac, by 
means of which a white powder precipitated fromit. This 
white matter being heated to redness, weighed 7-9 grains. 
It was digested in sulphuric acid, but 3°76 grains refused 
to dissolve. This portion possessed the properties of silica. 
The dissolved portion being mixed with a few drops of 
sulphate of potash, shot into crystals of alam. It was 
therefore alumina, and amounted to 4:14 grains. 
8. The yellow matter which refused to dissolve in the 
potash-ley was mixed with nitric acid. An effervescence 
took place, but the liquid remained muddy, till it was ex- 
posed to heat, when a clear reddish-brown solution was ef- 
fected. This solution was evaporated to dryness, and kept 
for a few minutes in the temperature of about 400°, to 
peroxidize the iron, and render it insoluble. A sufficient 
quantity of water was then poured on it, and digested on it 
tor half-an-hour, on the sand-bath. The whole was then 
tLrown upon a filter. The dark red matter which remained 
on the filter was drenched in oil, and heated to redness 
in a covered crucible. It was then black, and attracted by 
the magnet; but had not exactly the appearance of oxide 
of iron. It weighed 49:4 grains. 
4. The liquid which passed through the filter had not 
the sweet taste which I expected, but a slightly bitter one, 
similar to a weak solution of nitrate of lime. Hence it was 
clear that no yitria was present, as there cught to have 
been, had the mineral contained that earth. This liquid 
being mixed with carbonate of ammonia, a white powder 
precipitated, which, after being dried in a red heat, weighed 
17 grains. It dissolved in acids with effervescence: the 
solution was precipitated white by oxalate of ammonia, but 
not by pure ammonia. When dissolved in sulphuric acid, 
and evaporated to dryness, a light white matter remained, 
tasteless, and hardly soluble in water. These properties in- 
dicate carbonate of lime. Now, 17 grains of carbonate of 
Jime are equivalent to about 9°23 grains of lime. 
_ 5. From the preceding analysis, supposing it accurate, it 
followed that the mineral was composed of 
OWRCH Wie Geta de ec tie we le pot er nee 
EME ete ieee se eens ee 
Po i a ei mame LEY. | 
Oxide or roe st ees ee 8 t VERO 
Wolatie wittter™ TP eh? Bios 
96°91 
RD aes acts st 4 a's ork easels 3°09 
100°00 But 
