minute Quantities of Arsenic in Fluids. 591 
characteristically on diluted solutions of arsenic in mixed fluids, 
nevertheless always throws it down of some colour or other, 
even when the proportion of the poison does not exceed an 
8000th part, he proposes to employ this test, with the view of 
procuring the arsenic in a convenient form for being subjected 
to the decisive process of reduction. “ The suspected matter, 
if solid, is to be divided into minute fragments, and boiled 
briskly in two or three successive portions of pure water. The 
fluid, whether originally such, or procured by digestion from 
the solid matter, is then to be subjected in a deep, narrow 
glass, for half an hour, to a brisk stream of sulphuretted hy- 
drogen gas. In many cases, however, it will be necessary to 
premise the two following preparatory steps, before transmitting 
the gas ; and, as we can seldom know beforehand whether these 
steps are requisite or not, it may be right to resort to them in 
every case. The first precaution is to add a little acetic acid to 
the fluid. By so doing, the influence of any free alkali that 
may exist in it is counteracted ; and several organic principles, 
which might impede the subsequent separation of the precipi- 
tate, are coagulated. The second precaution is to boil the fluid 
for a few minutes ; by which means some matters are separated 
that the acetic acid could not throw down entirely ; and any 
carbonic acid existing in it is driven off. The presence of car- 
bonic acid in considerable quantity, by impeding the solution of 
the sulphuretted hydrogen, prevents its action on the arsenic, 
if the proportion of arsenic be small. The fluid is then to be 
filtered. - ” — u When the stream has been continued a sufficient 
length of time, there is either a precipitate formed, or the fluid 
acquires a yellowish milkiness, which passes to a distinct preci- 
pitate as soon as the excess of sulphuretted hydrogen is driven 
off by heat. It is always right to boil, before attempting to se- 
parate the matter thrown down ; as the precipitate then becomes 
much more distinct, and falls to the bottom more readily. When 
the filtration is finished, and the filter has been gently com- 
pressed between several folds of bibulous paper, the precipitate 
is to be scraped off with a knife, and dried on a bit of smooth 
paper, at a temperature somewhat above 912° The most advi- 
sable mode of subjecting this to the test of reduction is the fol- 
lowing : The best flux is the black flux, and the best instru- 
