290

QUANTITATIVE AGRICULTURAL ANALYSIS

One of the writers has had some experience with this method
for determining potassium in fertilizers of various kinds, and it
has been found possible to check reasonably well with the results
obtained by the chlorplatinate method. Some results obtained
by the two methods are given in the following table:

TABLE XXV.—COMPARISON OP THE PER CENT OF POTASSIUM OXIDE IN
FERTILIZERS BY CENTEIFUGAL AND CHLORPLATINATE METHOLS

Sample No.
Chlorplatinate method*
Centrifugal method t

1581
3.12
3.2

1604
1.74
1.8

1669
3.67
3.8

1823
8.04
8.2

1949
4.26
4.4

2176
. 9.38
9.1

2224
50.25
50.1

1979
4.83
4.9

* By the Indiana State Chemist,
f By one of the authors.
Special bottles have been described by Sherrill for this deter-
mination. These are of the form shown in Fig. 60. Or the
older Goetz bulbs, as used for rapid determinations of phosphorus
in steel, will be found convenient. The precipitate is collected
and measured in the narrow, graduated portion of the tube.
If the potassium solution contains ammonia or ammonium
salts, these must be expelled by evaporating a measured portion
to a small volume with enough sodium hydroxide to render the
solution decidedly basic, or by evaporating to dryness and ignit-
ing at dull redness. The solution is then acidified with acetic
acid and diluted to the original volume.
Determination of Potassium: Centrifugal Method.—Prepare solutions as
follows:
(a) Standard Potassium Chloride Solution.—Dissolve 15.83 gm of pure
potassium chloride in distilled water, add ten drops of glacial acetic acid and
dilute to 1000 cc. This makes a solution containing 1 per cent of potassium
oxide.
(6) Sodium Cdbaltic Nitrite Solution.—Dissolve 225 gm of sodium nitrite in
400 cc of clistilled water. Also dissolve 125 gm of cobalt acetate crystals