SEPARATION OF ARSENIC 
59 
firmation by other observers. The experiments which he 
describes, have been repeated with every possible care, and 
found completely exact. I will now give the additional 
observations which I have made on this subject : 
1 . I produced a strong extrication of hydrogen gas from 
perfectly pure materials, by placing some turnings of zinc in 
a glass tumbler, pouring upon them some acid, and inverting 
over them a funnel with a very fine extremity. The gas, 
when ignited, afforded a brilliant flame, totally colorless, 
and gave no stain to a porcelain cup held over it. Without 
extinguishing the flame, I added to the liquor a few drops of a 
very weak solution of arsenious acid in hydrochloric acid; 
immediately the flame became a light blue; on this change 
taking place, thick vapors of arsenious acid were given off, and 
a porcelain cup held over the flame w T as covered with a strong 
film of arsenic, so that it was only necessary to move the 
porcelain cup, to blacken, in the space of a few seconds, a 
space as large as the hand. If the quantity of acid is consi- 
derable, as, for example, the one thousandth of the liquor, all 
the flame is of a light blue, and if it be directed for a long time 
upon the same part of the porcelain, the film of arsenic becomes 
so thick that it curls up and detatehes itself. 
2. The liquor was then decanted; the turnings of zinc were 
washed with an abundance of water, and moistened with 
a fresh portion of pure acid, in the same apparatus. The 
inflamed gas again burnt with a blue flame, and exhibited the 
presence of a notable quantity of arsenic. This presence was 
again very apparent, even after pouring upon it once more 
some pure acid; and it results from this that, in these re- 
searches, we must employ zinc which has never before 
been used for a similar purpose. 
3. Of all the modes of detecting the arsenic in the flame, 
that of placing above it a piece of porcelain, or a tile from a 
Dutch stove, certainly deserves the preference; it is better 
than the use of glass or mica; as these, on account of their 
transparency, do not permit us to recognise very delicate 
shades, which are set off very strongly by a white ground. 
