210 
THE VOYAGE OF H.M.S. CHALLENGER. 
•list illation nearly to dryness decomposes it. Mr. Buchanan adopted the views of this 
hiudi authority ; hut, as a series of experiments of his own had led him to the conclusion 
that thr sulphates in the water are amongst the substances to which the carbonic acid 
adlnTi s, he subsequently, in his numerous carbonic acid determinations during the cruise, 
always expelled this component by distillation nearly to dryness with chloride of barium* 
When some years ago my attention was directed to the matter, I adopted Buchanan’s 
views as probably correct, and continued to hold them until the results of my first 
instalment of complete sea-water analyses (vide supra) led me to take a different view of 
the matter. 
These analyses, as already stated (on page 202), brought out a very appreciable propor- 
tion of base combined with carbonic acid, and although they did so, as I subsequently 
found, chiefly through an over-estimation of the soda, the fact, in a qualitative sense, 
was established beyond doubt by an independent method. But all this does not detract 
from the great merit of Torndc in having been the first to prove, by a long series of com- 
bined determinations of alkalinity and carbonic acid in sea- waters, that -the carbonic acid 
in those irate rs at least was present as bicarbonate, in general more or less incompletely 
saturated. After this discovery Jacobsen’s and Buchanan’s observations are easily 
explained. The fact that bicarbonates, being real compounds, are slow in giving up their 
surplus carbonic acid is evident, and if all their carbonic acid is given off on distillation 
nearly to dryness, this, I submit, is the effect of a slowly progressing decomposition of the 
hydrated chloride of magnesium into magnesia and hydrochloric acid, which latter 
• xpels all the carbonic acid. And if Buchanan found that in the distillation of a sea- 
water addition of chloride of barium accelerates the liberation of the carbonic acid, this 
obviously arises from the fact that this reagent decomposes the normal carbonate in the 
bicarbonate into carbonate of baryta and chlorides of the respective basilous metals. But 
why did Buchanan always obtain far less carbonic acid per litre than Jacob en had 
obtained ? Why did not the decomposition of the chloride of magnesium bring about the 
decomposition of his carbonate of baryta as it had disposed of Jacobsen’s original sea- 
wute r carbonates ? This would be a difficult question to answer if we did not know from 
experiments of Rose’s that a mixed precipitate of sulphate and carbonate of baryta is only 
tv>ry xlmrhj attacked by dilute hydrochloric acid. 
In the course of one of my critical experiments on Buchanan’s method I once, by 
mi st ake , added a quantity of chloride of barium insufficient to bring down all the 
-ulphuric acid. The carbonic acid eliminated on distillation increased beyond what was 
obtained in experiments with excess of reagent. The chloride of barium in this case had 
b* • n -ubstantially decomposed by the sulphates, and thus been prevented to a great extent 
fr .m acting »>n the carbonates <»f the water, which thus (to this extent) were exposed to 
the influence of the chloride of magnesium. 
* Sro note liy Sir Duchnnun at the cn<l of this Report, page 245 . 
