neglected.

[Note 1: Magnesium ammonium phosphate, as noted below, is slightly
soluble under the conditions of operation. Consequently the
unavoidable errors of analysis are greater in this determination than
in those which have preceded it, and some divergence may be expected
in duplicate analyses. It is obvious that the larger the amount of
substance taken for analysis the less will be the relative loss or
gain due to unavoidable experimental errors; but, in this instance, a
check is placed upon the amount of material which may be taken both by
the bulk of the resulting precipitate of ammonium phosphomolybdate
and by the excessive amount of ammonium molybdate required to effect
complete separation of the phosphoric acid, since a liberal excess
above the theoretical quantity is demanded. Molybdic acid is one of
the more expensive reagents.]

[Note 2: Soluble silicic acid would, if present, partially separate
with the phosphomolybdate, although not in combination with
molybdenum. Its previous removal by dehydration is therefore
necessary.]

[Note 3: When washing the siliceous residue the filtrate may be tested
for calcium by adding ammonia, since that reagent neutralizes the
acid which holds the calcium phosphate in solution and causes
precipitation; but after the removal of the phosphoric acid in
combination with the molybdenum, the addition of an oxalate is
required to show the presence of calcium.]

[Note 4: An excess of nitric acid exerts a slight solvent
action, while ammonium nitrate lessens the solubility; hence the
neutralization of the former by ammonia.]

[Note 5: The precipitation of the phosphomolybdate takes place more
promptly in warm than in cold solutions, but the temperature should
not exceed 60 deg.C. during precipitation; a higher temperature tends to
separate molybdic acid from the solution. This acid is nearly white,
and its deposition in the filtrate on long standing should not be
mistaken for a second precipitation of the yellow precipitate. The
addition of 75 cc. of ammonium molybdate solution insures the presence
of a liberal excess of the reagent, but the filtrate should be tested
as in all quantitative procedures.

The precipitation is probably complete in many cases in less than
twelve hours; but it is better, when practicable, to allow the
solution to stand for this length of time. Vigorous shaking or
stirring promotes the separation of the precipitate.]

[Note 6: The composition o