h-pink stain is produced
on the paper, which is turned blue on moistening with dilute alkali.

The amount of the boric acid radicle may be determined by incinerating
5-10 grammes of soap, extracting with hot dilute acid, filtering,
neutralising this solution to methyl orange, and boiling to expel carbon
dioxide. After cooling, sufficient pure neutralised glycerine is added
to form one-third of the total volume, and the liquid titrated with N/2
caustic soda solution, using phenol-phthalein as indicator. Each c.c. of
N/2 NaOH solution corresponds to 0.031 gramme crystallised boric acid,
H_{3}BO_{3} or 0.0477 gramme crystallised borax,
Na_{2}B_{4}O_{7}.10H_{2}O.

LYES.

The amounts of caustic alkali (if any), carbonated alkali, and salt
present are determined in the manner already described under Alkali and
Alkali Salts. The glycerol content is ascertained by taking 2.5 grammes,
adding lead subacetate solution, and filtering without increasing the
bulk more than is absolutely necessary; the solution is concentrated to
about 25 c.c., and the oxidation with bichromate and sulphuric acid
conducted as described in the examination of Crude Glycerine. The
solution, after oxidation, is made up to 250 c.c., and titrated against
standard ferrous ammonium sulphate solution, the formula for the
calculation being:--

{0.25 - 2.5}
Per cent. of glycerol = { ---} x 40
{ n }

where n equals the number of c.c. of oxidised lyes required to oxidise
the ferrous ammonium sulphate solution.

The estimation of actual glycerol in this is necessarily a matter of
considerable importance, and a very large number of processes, which are
constantly being added to, have been suggested for the purpose.
Hitherto, however, only two methods have been generally adopted, _viz._
the acetin and the bichromate processes. Unfortunately the results
obtained by these do not invariably agree, the latter, which includes
all oxidisable matter as glycerol, giving sometimes considerably higher
results, and it has been suggested that a determination should be made
by both methods, and the average of the two results considered the true
value. This involves a considerable amount of time and trouble, and it
will generally be found sufficient in a works laboratory to determine
the glycerol by one method only in the ordinary course, reserving the
other process for use as a ch