^ 
ON THE IMPURITIES OF COMMERCIAL ZINC. 19 
previous contents had been proved as above described, and at the sod of three quar- 
ters of an hour a distinct deposit of arsenic was found in the reduction-tube. The 
amount of arsenious acid actually placed in the flask was two hundred-thousandths of 
a gramme (.00002 gram.). 
0.00002 gram. arsenious acid = 0.000015 gram. arsenic. . 
Ratio of the arsenic present to the zinc — 0.000015 : 200 — Laera, 
« « «oc D * — amount of liquid in the flask — about i Late 
Our apparatus was therefore competent to detect a quantity of arsenic less than one 
ten-millionth of the weight of the zinc used, or of the amount of fluid in the flask. This 
quantity of liquid necessarily varied somewhat, in consequence of the slight additions 
of acid necessary to maintain a constant current of hydrogen, but only varied to be 
increased, never diminished. Remembering the wide limits of error in many chemical 
processes, the comparative coarseness of most chemical tests, and the many unavoidable 
inaccuracies in weighing and measuring, is not the assertion perfectly safe, and in strict 
conformity with the general use of language in other qualitative or quantitative deter- ` 
minations, that a specimen of zinc*is free from arsenic, which does not show the 
slightest trace of that metal in an apparatus abundantly capable of detecting the ten- 
millionth part of arsenic? We are aware that this is not the limit of delicacy * of 
Marsh's test, but, assured of this delicacy, we rest satisfied with it as sufficient for our 
present purpose. The more delicate the test, the stronger is our present argument, 
and the greater need is there of exceeding caution in applying this test in the investi- 
gations of medico-legal or pharmaceutical chemistry. 
- . In connection with these experiments on the delicacy of the test, we ; Would call 
Se to the fact, that the sulphuretted hydrogen, which we have shown in our exam- 
inations for sulphur to be always developed from commercial zinc, does not prevent the 
exhibition of such a very small amount of arsenic as 0.000015 gram. Chevallier,t 
and more recently Blondlot + and Leroy,$ assert that the presence of the insoluble sul- 
phide of arsenic cannot be recognized by Marsh's test, and that arsenic may therefore 
be concealed by being converted into the sulphide. This is the natural and general 
. * M. Signoret (Taylor on Poisons, 2d Edition, 1859, London, p. 396) states that he has procured 
metallic deposits with only the 200,000,000th part of arsenic in the liquid; but it is not clear from such a 
statement what the exact amount of arsenic in the apparatus was which enabled him to obtain deposits, — 
a very material point. 
f Jour. de Ch. Méd., [2.] V. 380, in Berzelius's Jahres-Bericht, 1841, X X. 192. 
1 Comptes Rendus, 1857, XLIV. 1222. 
$ Ibid., 1859, XLIX. 469. 
