212 
THE VOYAGE OF H.M.S. CHALLENGER. 
If it ho permissible to accept my artificial, as a sufficient substitute for natural, sea- 
wator, my synthetical experiments may claim a certain degree of oceanographic significance. 
Weording to the current notions on “dissociation,” the liberation of carbonic acid from 
bicarbonates in sea-water by repeated shaking with air should come to a stop as soon as 
the partial tension of the carbonic acid in the air-residue has come up to a limit 
value, which is a function of temperature alone, as long as there is any bicarbonate left in 
the water undecomposed. Hence, supposing at a certain constant temperature we 
shako 1 volume of a sea-water with say 5 volumes of air, then remove the air- 
residuo and shake with 5 volumes of renewed air, and so on, although the quantity 
of carbonic acid in the water gets less and less at each stage, the proportion of carbonic 
m id in the successive air-residues removed should retain a constant value. I take “5 
volumes” of air as an example, because this is the exact proportion of air which 
1 consistently applied in the one series of experiments in which I have most confidence, 
and which is reported on page 115 of the memoir. What I really did in this, as in my 
ot her synthetical experiments, was to determine the number of milligrams of carbonic 
acid contained — firstly, in the original water ; secondly, in the residue after the first 
shaking; thirdly, in the residue left after two successive shakings with 5 volumes of 
air, of a fresh volume of the original water, &c. But this comes to the same result as 
if the same volume of water had been operated upon successively, and from the weight 
of carbonic acid found we can calculate the successive portions of the gas which were 
carried away. I have done so and found that the carbonic acid tensions in the three 
successive volumes of air removed were 57, 5’5, and 3'9 ten-thousandths of an atmo- 
sphere. These three numbers, considering the difficulties involved, are sufficiently near 
to their mean of 5 of the above units, and this value, accordingly, may be provisionally 
adopted as an approximation to the dissociation tension of sea- water bicarbonates at 
the temperature at which I operated, and which, quoting from memory, I may say lay 
l>etween 18° and 21° C. 
1 am aware that this part of my work lacks the degree of precision which would be 
desirable for my present train of reasoning. But I had not the time to embark in the far 
more clal»orate investigation which would have been desirable. 1 have, however, quite 
lately resumed the matter on a new basis, and hope before long to be able to formulate 
the exact conditions of stability in sea-water bi carbonates as they exist when dissolved in 
real .-ea- water, and amongst others t<> decide the question whether in this process they 
quite directly tend to become normal, and do not perhaps more directly gravitate towards 
the state of se-qui-carbonate. In the meantime we must reason on what data we have. 
tonsidering that at a temperature* of 18 to 21° C. the dissociation tension of the 
bicarbonate- in sea-water is 5 ten-thousandths of an atmosphere, at temperatures not 
differing by more than one or two degrees from 0° C., such as prevail in the arctic and 
antarctic circles, it is far more likely than not to fall below 3 ten-thousandths, which 
i" al*»ut the partial tension of the carbonic acid in the atmosphere. Admitting this, 
