194 Mr. Howard's Experiments and Observations 
immediately after the same alkali, pure, had thrown down what 
I took wholly for oxide of iron. I had now obtained every thing 
in the subject of my analysis, except magnesia and nickel. The 
former, and a trace of the latter, were held by carbonic acid in 
the liquor, from which the ferruginous precipitate was, in the 
first instance, thrown down by carbonate of potash ; and the 
latter was found in the last named muriate of ammonia. I dis- 
engaged the magnesia, by the assistance of potash, and by 
evaporating to dryness. The oxide of nickel was precipitated 
by hydrogenized sulphuret of ammonia. 
Under all circumstances, I am induced to state the proportions 
of constituent parts thus : 
* 
Silica - 
5° 
Magnesia 
1 5 
Oxide of iron 
34 
Oxide of nickel 
Qj 
lOlf. 
The excess of weight, instead of the usual loss, is owing to 
the difference of oxidizement of the iron, in the stone and in 
the result of the analysis ; which will be found to be the case 
in all analyses of these substances ; indeed it is always necessary 
to reduce the oxide to the red state, as being the only one to be 
depended upon. To avoid future repetition, I shall also observe, 
first, that by preliminary experiments, I could not detect any 
other substance than those mentioned. Secondly, that the earth 
obtained as alumina, appeared to me to be mostly, if not 
entirely, siliceous ; because, after it had been ignited, and again 
treated with potash and muriatic acid, I found it was very 
nearly all precipitated by evaporation. Thirdly, I examined, 
and judged of, the silica collected from the oxide of iron, in the 
