RAPPORT 1903—04 ANNEXE A SO 
VII. Gas Analysis: 
It is believed that the C. L. has now successfully elucidated the theory 
according to which apparatus for the extraction of the dissolved gases should be 
constructed, and as a result it is now able to recommend an apparatus for the 
complete extraction and estimation of the three atmospheric gases dissolved in 
sea-water, in one operation. When a series of samples is being analysed, each 
one is found to require not more than about 40 minutes for the extraction and 
estimation of the CO2, Oz and Ne. The important point brought out by these 
investigations is that it is absolutely necessary, if the gases are to be completely. 
expelled, to remove them as evolved from the liquid, so as to keep the pressure 
very low. Raising the temperature is a much less satisfactory means of expulsion, 
because the coefficient of solubility of the asmospheric gases decreases only very 
slowly with the temperature. 
For this reason the apparatus as now recommended by the C. L. consists of 
(1) a flask of about 1000 cc. (fitted with condenser) into which the water sample 
is eventually brought, (2) a collecting vessel of about 500 ee. fitted with mercury 
reservoir; is is connected with a two way tap with the flask (1) and with the 
atmosphere or (3) the burette in which the analyses are carried out as soon as 
all the gases shall have been transferred to it. The flask is exhausted by boiling 
some water and acid contained in it and connecting to a water pump; the final 
traces of air are removed by washing out with a current of hydrogen drawn at 
will from a gas pipette arranged on the principle of Kıpp’s apparatus and charged 
with pure zine and acid, and in this last operation the collecting vessel (2) is used 
as a mercury pump. 
The water sample is then brought into the flask — without any external 
air getting in at the same time — and about 90° of the gas can immediately be 
pumped off and transferred to the burette. Boilmg and washing out with hydrogen 
in a few minutes completely remove the whole of the gases and the mixture of 
COz2, O2, Ne, and He, is ready for analysis. CO>2 is absorbed by potash; O2 is 
exploded with the hydrogen which must be in excess; and finally the residual He 
in the mixture of Ne and He now remaining, is determined by again exploding 
with an excess of pure oxygen (sufficient for many analyses is easily obtained and 
can be stored up in a side tube, by heating pure potassium permanganate) and 
measuring the contraction. 
Experiments seem to indicate that each dissolved gas may under certain 
circumstances be liable to change as a result of life processes, and it is therefore 
necessary — if the water is not to be instantly analysed — that the sample should 
be sterilised. Of late years the samples have usually been collected in sealed-up 
tubes previously evacuated by making the tube the head of a barometric column 
