made by placing a piece of the bolting cloth over the top of a small
beaker in which the ground mineral is placed, holding the cloth in
place by means of a rubber band below the lip of the beaker. By
inverting the beaker over clean paper and gently tapping it, the fine
particles pass through the sieve, leaving the coarser particles within
the beaker. These must be returned to the mortar and ground, and the
process of sifting and grinding repeated until the entire sample
passes through the sieve.

[Note 1: If the sample of feldspar for analysis is in the massive or
crystalline form, it should be crushed in an iron mortar until the
pieces are about half the size of a pea, and then transferred to a
steel mortar, in which they are reduced to a coarse powder. A wooden
mallet should always be used to strike the pestle of the steel mortar,
and the blows should not be sharp.

It is plain that final grinding in an agate mortar must be continued
until the whole of the portion of the mineral originally taken has
been ground so that it will pass the bolting cloth, otherwise the
sifted portion does not represent an average sample, the softer
ingredients, if foreign matter is present, being first reduced to
powder. For this reason it is best to start with not more than the
quantity of the feldspar needed for analysis. The mineral must be
thoroughly mixed after the grinding.]

FUSION AND SOLUTION

PROCEDURE.--Weigh into platinum crucibles two portions of the ground
feldspar of about 0.8 gram each. Weigh on rough balances two portions
of anhydrous sodium carbonate, each amounting to about six times the
weight of the feldspar taken for analysis (Note 1). Pour about three
fourths of the sodium carbonate into the crucible, place the latter on
a piece of clean, glazed paper, and thoroughly mix the substance and
the flux by carefully stirring for several minutes with a dry glass
rod, the end of which has been recently heated and rounded in a flame
and slowly cooled. The rod may be wiped off with a small fragment of
filter paper, which may be placed in the crucible. Place the remaining
fourth of the carbonate on the top of the mixture. Cover the crucible,
heat it to dull redness for five minutes, and then gradually increase
the heat to the full capacity of a Bunsen or Tirrill burner for
twenty minutes, or until a quiet, liquid fusion is obtained (Note 2).
Finally, heat the sides and cover strongly until any material which
may have co