repeat the boiling (Note 1). If the color does not then reappear, add
alkali until it does, and a !drop or two! of acid in excess and boil
again for one minute (Note 2). If no color reappears during this time,
complete the titration in the hot solution. The end-point should be
the faintest visible shade of color (or its disappearance), as the
same difficulty would exist here as with methyl orange if an attempt
were made to match shades of pink.

From the corrected volume of alkali required to react with the
oxalic acid, calculate the percentage of the crystallized acid
(H_{2}C_{2}O_{4}.2H_{2}O) in the sample (Note 3).

[Note 1: All commercial caustic soda such as that from which the
standard solution was made contains some sodium carbonate. This reacts
with the oxalic acid, setting free carbonic acid, which, in turn,
forms sodium bicarbonate with the remaining carbonate:

H_{2}CO_{3} + Na_{2}CO_{3} --> 2HNaCO_{3}.

This compound does not hydrolyze sufficiently to furnish enough OH^{-}
ions to cause phenolphthalein to remain pink; hence, the color of
the indicator is discharged in cold solutions at the point at which
bicarbonate is formed. If, however, the solution is heated to boiling,
the bicarbonate loses carbon dioxide and water, and reverts to sodium
carbonate, which causes the indicator to become again pink:

2HNaCO_{3} --> H_{2}O + CO_{2} + Na_{2}CO_{3}.

By adding successive portions of hydrochloric acid and boiling, the
carbonate is ultimately all brought into reaction.