88 ON THE IMPURITIES OF COMMERCIAL ZINC. 
been thoroughly tested à blanc and found pure. At the end of the second hour, a very 
slight deposit was discernible on close inspection in the fine tube, but the result was ` 
too doubtful to warrant the assertion that arsenic was contained in the insoluble resi- 
due. We next tried the same experiment with Silesian zinc, in which Marsh's test had 
detected arsenic. The residue from 40 grammes of this zinc, submitted to the process 
just described, produced in the reduction-tube in fifteen minutes a brown deposit, 
which in half an hour became well marked, and at the end of the hour afforded suf- 
ficient evidence of the presence of arsenic in the residue, it being well understood that 
the zinc and acid used in this experiment had been previously tested for one hour and 
found pure. The only conclusion to which these experiments point is, that when a given 
sample of zinc contains arsenic, a portion of that arsenic will escape combination with 
the hydrogen generated by the solution of the zinc, and will remain in the insoluble 
residue. It was useless in this connection to examine the residue from New Jersey 
zinc, because that spelter itself contained arsenic, and its residue contained metallic ` 
copper, which would inevitably retain arsenic, as in Reinsch’s test. The Pennsylvanian 
"zine gave no residue with acids. 
The absolute necessity of obtaining a zinc free from arsenic for many chemical in- 
vestigations, renders any process for purifying zinc from arsenic a matter of consider- 
able interest and importance. We therefore tried the process of purifying zinc by 
fusing it with one fourth of its weight of saltpetre, a method fully described by 
Meillet,* but previously suggested by Orfila. 760 grammes of Silesian zinc were finely 
granulated and mixed with one fourth this weight of saltpetre, by placing the zinc 
and nitre in alternate layers in a Hessian crucible. The mixture was heated till de- 
flagration ensued, when the melted mass was poured into cold water to separate the 
slag, caustic potassa, and any arseniate of potassa which might have formed. The 
washed mass was remelted, and again granulated. ‘The loss of zinc during the process is 
very large ; we obtained only 200 grammes from the original 160 grammes. Of this zinc 
170 grammes were tested in Otto's apparatus with pure acid, and in twenty minutes there 
began to form in the reduction-tube a brownish deposit, which, at the end of three 
quarters of an hour, was a sufficient evidence of the presence of arsenic. The deposit 
was as large as that obtained from the same Silesian zinc, before it had undergone this 
processeof fusion with nitre. It has been already stated, under the appropriate head, 
that this process did not disembarrass the zinc of the sulphur which it contained, and 
* Dingler's Polyt. Jour., 1842, LXX XIII. 205 ; from Jour. de Pharmacie, 1841, p. 625. : 
i Annales d'Hygiene Publique, 1839, XXII. 427. : : 
