ettled, and decant the clear liquid for use as required. It is better to
add this by degrees than to run the risk of overdoing. If too much is
added, the bath is not of necessity spoiled, but it takes a great deal of
working to bring it in order again.

About eight ounces of the double sulphate to each gallon of distilled or
rain water is a good proportion to use when making up a bath. There is a
slight excess with this. It is a mistake to add the salt afterward, when
the bath is in good condition. The chloride and cyanide are said to give
good results. I can only say that the use of either of these salts has not
led to promising results in my hands.

In preparing the double sulphate, English grain nickel is decidedly the
best form of metal to use. In practice, old anodes are generally used.

The metal is dissolved in a mixture of nitric and dilute sulphuric acid,
with the application of a gentle heat. When sufficient metal has been
dissolved, and the unused nitric acid expelled, the salt may be
precipitated by a strong solution sulphate of ammonia, or, if much free
acid is present, carbonate of ammonia is better to use.

Tin, lead, and portion of the iron, if present, are removed by this method.
The silica, carbon, and portions of copper are left behind with the
undissolved fragments of metals.

The precipitated salt, after slight washing, is dissolved in water and
strong solution ammonia added. A clean iron plate is immersed in the
solution to remove any trace of copper. This plate must be cleaned
occasionally so as to remove any reduced copper, which will impede its
action. As soon as the liquid is free from copper, it is left alkaline and
well stirred so as to facilitate peroxidation and removal of iron, which
forms a film on the bath. When this ceases, the liquid is rendered neutral
by addition of sulphuric acid, and filtered or decanted. The solution, when
properly diluted, has sp. gr. about 1.06 at 60 deg. F. It is best to work the
bath with a weak current for a short time until the liquid yields a fine
white deposit. Too strong a current must be avoided.

If the copper has not been removed, it will deposit on the anodes when the
bath is at rest. It should then be removed by scouring.

Copper produces a reddish tinge, which is by no means unpleasant compared
with the dazzling whiteness of the nickel deposit. If this is desired, it
is far better to use a separate bath, using anodes of suitable composition.