TIIE VOYAGE OF H.M.S. CHALLENGER. 
208 
analysis would bo of no value. From the analyses of the deposit and the water-remnant 
left, and the complete analysis previously made, I calculated that the original water should 
have contained, per 100 of chlorine, the quantities of lime and magnesia given below and 
contrasted with those present in average surface water, according to the 43 analyses 
quoted on pp. 30 and 31 — 
Lime. Magnesia. 
Water (No. 31) in its original condition by calculation, . . 3-496 11-163 
A vemge surface or small-depth water, .... 3'018 11-203 
Hence, taking 0T4G as corresponding to ordinary surface-waters, I calculate for the 
surplus alkalinity 0T779, which, together with the normal value 0 - 146, gives 0’324 per 
100 of salts for the original water. The calculations which led to these high values for 
Nos. 5 and 31, do not, it is true, rest upon a perfectly secure basis; but I believe the 
r* -suit s arc approximately eorrect, and besides the high number 0'2079, which was found 
quite directly for the normal water, No. 616 is beyond suspicion. I have no doubt that 
far higher alkalinities than even 0'33 occur locally in many parts of the ocean, wherever 
there is abundance of carbonic acid and of carbonate of lime or magnesia at the same 
time. To obtain some insight into the possible extreme limit, I took a sea- water whose 
alkalinity was 50’2 milligrams per litre, the surplus base being present substantially as 
bicarbonate, and after having saturated it with carbonic acid, I digested it, in one case 
with carbonate of lime, in another with carbonate of magnesia. The filtered liquors 
showed immense alkalinities, the increase being 
Lime. Magnesia. 
Increase of alkalinity, . . 314-2 mgrms. per litre, 1234-0 iugrms. per litre. 
A similar set of trials with the natural water gave, in the case of magnesia, an 
increase in alkalinity of 106 ; in the case of lime there was a decrease of 3’2 milligrams 
p'-r litre.* The decrease of alkalinity caused by the addition of carbonate of lime is 
difficult to explain. Perhaps it is, only the outcome of an observational error; but in 
any case my experiments show that it would be quite possible for the alkalinity to 
increase beyond the maxima that occurred in the 130 samples. It is very curious 
that in my experiments with sea-water saturated with carbonic acid, carbonate 
<*f magic -ia proved far more abundantly soluble than carbonate of lime. My explana- 
tion i- that a considerable portion of the magnesia, immediately after having been 
dissolved by the carl ionic acid, suffered double decomposition by the large mass of chloride 
•dium present, with formation of carbonate of soda and a double chloride of sodium 
and magnesium ; so that for this part of the process a very small proportion of free or 
• ■■ unbilled carbonic acid would have sufficed, as it always comes back in the reaction, 
whi< h I nup|M,*c to go on, as an equivalent quant ity of bicarbonate of soda. The tendency 
of magnesium to form such double salt- explains how those Challenger waters deposited 
Compart- page 131. 
