36 Dr. H. T. Brown and Mr. F. Escombe. [Jan. 9, 
dioxide absorption by means of barium hydroxide when the volume of the 
solution is large and the estimation has to be deduced from the initial and 
final titration of the liquid. In such a case it is manifest that we have 
to depend on the determination of small differences between two large values, 
and that the full acid-equivalent of all errors of measurement of the 
respective solution-volumes will accumulate in the final result, as well as all 
other errors incident to methods of titration generally. We therefore 
discarded the barium hydroxide method at an early stage of the enquiry, and 
turned our attention to one which is free from the cumulative errors to which 
we have referred. Such a one is found in a method which was originally 
proposed by P. Hart in 1887,* for the estimation of the relative amounts of 
caustic soda and sodium carbonate in soda-ash. As far as we know, this has 
never been regarded from any other point of view than that of a convenient 
commercial method, whereas when certain precautions are taken it is 
capable of a high degree of precision, and really affords one of the most 
accurate means of determining small quantities of carbon dioxide short 
of actual measurement of the gas. 
It is based on double titration with two indicators, one of them phenol- 
phthaléin, sensitive to free carbon dioxide, the other, methyl-orange, only 
reacting with the excess of mineral acid used in the titration. 
If we have a solution of caustic soda which has become partially car- 
bonated, the first stage of the process is the addition of the phenol-phthaléin- 
indicator, followed by a mineral acid (hydrochloric) until the pink 
colour just disappears. This marks the point of the complete conversion of 
the neutral into the acid carbonate, or more strictly speaking, the point at 
which the liberation of the first trace of carbon dioxide takes place. Up to 
this stage it is unnecessary to take any account of the actual amount of acid 
used, since this is the starting point for the true titration-process. 
Methyl-orange is now added, and this is followed by 1/10 normal hydro- 
chloric acid until the whole of the acid carbonate is decomposed, the volume 
of the 1/10-normal acid requisite to produce this effect, being of course a 
measure of the carbon dioxide displaced. 
Tt will be noticed that this method is quite independent of the relations of 
the total acid-equivalent of the alkali before and after absorption of carbon 
dioxide, the determination of the carbon dioxide absorbed merely depending 
on the amount of dilute acid required in each case between the first reaction 
with phenol-phthaléin and the second with methyl-orange. There is also the 
great advantage that errors in the estimation of the volumes of the solutions. 
* * Journ. Soc. Chem. Ind.,’ 1887, p. 347. 
