esult. A small quantity only of
silicon separates in the uncombined form, the greater quantity separating
in the form of silica, SiO_{2}, the amorphous silicon so obtained
apparently being more prone to oxidation than the boron so obtained.

Ferrous sulphide was next similarly treated, and gave, after the lapse of a
few hours, a copious blackish precipitation of sulphur, and possessing
properties similar to the sulphur obtained by dissolving sulphides such as
cupric sulphide in dilute nitric acid, in all other respects resembling
common sulphur.

Phosphides of iron, zinc, etc., were next introduced, and gave, besides
carbon and other impurities, a residue containing a large percentage of
phosphorus, which differed from ordinary phosphorus with respect to its
insolubility in carbon disulphide, and which resembled the reaction in the
case with silicon-eisen rather than that of the boron compound, insomuch
that a large quantity of the phosphorus had passed into solution.

A rod of impure copper, containing arsenic, iron, zinc, and other
impurities, was next substituted, using hydrochloric acid as a solvent in
place of sulphuric acid. In the course of a day the copper had entirely
dissolved and precipitated itself on the negative electrode, the impurities
remaining in solution. The copper, after having been washed, dried, and
weighed, gave identical results with regard to percentage with a careful
gravimetric estimation. I have lately used this method, and obtained
excellent results with respect to the analysis of commercial copper,
especially in the estimation of small quantities of arsenic, thus enabling
the experimenter to perform his investigation on a much larger quantity
than when precipitation is resorted to, at the same time avoiding the
precipitated copper carrying down with it the arsenic. I have in this
manner detected arsenic in commercial copper when all other methods have
totally failed. I have also found the above method especially applicable
with respect to the analysis of brass.

With respect to ammoniacal dissolution, which I will briefly mention, a rod
composed of an alloy of copper and silver was experimented upon, the copper
becoming entirely dissolved and precipitating itself on the platinum
electrode, the whole of the silver remaining suspended to the positive
electrode in an aborescent form. Arsenide of zinc was similarly treated,
the arsenic becoming precipitated in like manner on the platinum ele