crucible will filled with the solution.

Pour the liquid above the silver chloride slowly onto the filter,
leaving the precipitate in the beaker as far as possible. Wash the
precipitate twice by decantation with warm water; then transfer it
to the filter with the aid of a stirring-rod with a rubber tip and a
stream from the wash-bottle.

Examine the first portions of the filtrate which pass through the
filter with great care for asbestos fibers, which are most likely to
be lost at this point. Refilter the liquid if any fibers are visible.
Finally, wash the precipitate thoroughly with warm water until free
from soluble silver salts. To test the washings, disconnect the
suction at the flask and remove the funnel or filter tube from the
suction flask. Hold the end of the tube over the mouth of a small test
tube and add from a wash-bottle 2-3 cc. of water. Allow the water to
drip through into the test tube and add a drop of dilute hydrochloric
acid. No precipitate or cloud should form in the wash-water (Note 16).
Dry the filter and contents at 100-110 deg.C. until the weight is constant
within 0.0003 gram, as described for the preparation of the filter.
Deduct the weight of the dry crucible from the final weight, and from
the weight of silver chloride thus obtained calculate the percentage
of chlorine in the sample of sodium chloride.

[Note 1: The washed asbestos for this type of filter is prepared by
digesting in concentrated hydrochloric acid, long-fibered asbestos
which has been cut in pieces of about 0.5 cm. in length. After
digestion, the asbestos is filtered off on a filter plate and washed
with hot, distilled water until free from chlorides. A small portion
of the asbestos is shaken with water, forming a thin suspension, which
is bottled and kept for use.]

[Note 2: The nitric acid is added before precipitation to lessen the
tendency of the silver chloride to carry down with it other substances
which might be precipitated from a neutral solution. A large excess of
the acid would exert a slight solvent action upon the chloride.]

[Note 3: The solution should not be boiled after the addition of the
nitric acid before the presence of an excess of silver nitrate is
assured, since a slight interaction between the nitric acid and the
sodium chloride is possible, by which a loss of chlorine, either as
such or as hydrochloric acid, might ensue. The presence of an excess
of the precipitant can usually be recognized at