THE VOYAGE OF H.M.S. CHALLENGER. 
202 
Th< r< suits, while fairly agreeing with Forchhammer’s, were in still closer accordance 
with om* another, and thus showed that Forchhammer’s proposition may be extended 
from surface-waters to ocean-waters obtained from all depths. 
The solid matter dissolved in sea-water, though strictly speaking, and we may add 
Uoeessarilv, of a very complex composition, consists substantially of the muriates and 
sulphates of soda, magnesia, lime, and potash. Forchhammer, after having satisfied 
himself that all the other constituents taken conjointly amount to only a small fraction 
.if one percent, of the total solids, in his individual analyses limited himself to exact 
determinations of the chlorine, sulphuric acid, lime, and magnesia. The potash he deter- 
mined only in a comparatively small number of samples; and where he reports the soda 
this component is calculated by difference, on the assumption that the acids and bases 
present exactly neutralise each other. But this assumption had never been proved to 
be correct, and d priori, is improbable, because it leaves out of reckoning the carbonate 
of lime which many animals need for forming their shells. I therefore in my analyses 
made it a special point, in addition to the other bases, to determine also the soda by a 
method independent of the assumption quoted; and on calculating my first set of (21) 
Challenger water analyses, had the satisfaction of finding that they had all given a small 
Mirplus of base, amounting on an average to 8G equivalents per 10,000 equivalents of acid 
present, corresponding (if we assume the excess of base to be present as normal carbonate) 
to about OT 1 gram of carbonic acid, equivalent to 0’25 gram of carbonate of lime per 
1000 grams of sea-water analysed. While recognising the importance of this result, I 
was keenly alive to the possibility of its having been brought about by a constant positive 
error in my sum total of base determinations, and accordingly sought for an exact 
direct method for the determination of the surplus base in a given sea-water. 
One of the methods I tried was to distil a measured volume of the sea-water with 
a certain proportion of sulphate of ammonia; then in a strictly comparative manner 
t«i repeat * he experiment with an exactly neutral artificial sea-water substituted for the 
natural sea-water, and to determine the ammonia in the two distillates. There was a 
distinct .surplus of ammonia in the distillate from natural as compared with that from 
the artificial sea-water, proving the existence of surplus fixed base in the former; 
but the results were not sufficiently constant to pass for quantitative determinations 
• >f the “alkalinity. 1 he problem was subsequently solved in a surprisingly simple 
manner by the chemists of the Norwegian Expedition.* 1 lost no time in testing their 
method by synthetical trials, and finding it trustworthy, applied it to 154 samples of 
< h.tlb nger water. I hey all proved to be alkaline, but the mean value corresponded 
only 54*7 milligrams of carbonic acid (Co, present as RaCO.,) per kilogram of sea- 
water, which showed that my complete analyses had considerably over-estimated the 
lus base. I refer to my memoir in regard to the manner in which I utilised my 
* TU Norwegian North Atlantic Expedition, 1870 to 1878 ; Chemistry by Tornjle, Christiania, page 31. 
