REPORT ON THE COMPOSITION OF OCEAN- WATER. 
39 
II.— ON THE SALINITY OF OCEAN-WATER. 
Although the composition of the material dissolved in ocean- water is substantially 
the same everywhere, the quantity in a given volume is subject to considerable 
variation ; and it is one of the great problems of oceanography to define this ratio 
numerically as a function of longitude and latitude, of depth and time. 
For the determination of the salinity of a given sample of sea- water, the following 
methods readily suggest themselves : — 
1. The determination of the specific gravity at some chosen standard temperature. 
This constant, of course, bears a fixed relation to the salinity, which relation can be 
ascertained, once for all, by standard experiments ; but even as it stands, it obviously 
“ measures ” the salinity in the sense in which the reading of a thermometer “ measures ” 
the temperature. The specific gravity of a water can easily be determined, both with 
promptitude and precision, by means of a delicate hydrometer — a method which is 
specially adapted for being carried out on board ship. Mr. Buchanan adopted it, and 
during the voyage applied it to a large number of samples. 
2. The determination in a given weight or volume of the water of the weight 
of chlorine present, which latter, on multiplication by a certain constant factor, 
yields the total solids. According to my 77 complete analyses, as recalculated in the 
chapter on Alkalinity, this factor should be = 1 ’8058. According to p. 28 in the dis- 
cussion of the results of the complete analyses, the probable uncertainty of this factor, as 
applying to any one sample, taken at random, should be about of its value, 
or= ±0'002. This method, to a chemist working in a laboratory on terra firma , would 
naturally suggest itself as the best, and I accordingly applied it to the samples of water 
collected by the Challenger, 
3. The direct determination of the total salts in a known quantity of the water. 
This at first sight would appear to be the best method of all ; but unfortunately sea- 
water cannot be evaporated to dryness, and the residue dehydrated by ignition, without 
sending off some of the chlorine of the chloride of magnesium as hydrochloric acid. On 
p. 18 of Chapter I., I have already referred to a number of abortive attempts of 
my own for preventing this decomposition, or rendering it innocuous in the determina- 
tion of the total salts, which datum I was very much in need of at the time, as enabling 
me to calculate the percentage of soda in my analyses of the salt of sea-water. The 
Norwegian chemists,* who required the datum for the reduction of their chlorine 
determinations and specific gravities, employed the following process, and found it to give 
good results : — 
* Den Norske Nordhavs-Expedition, 1876-1878 ; Cliemi, af Torn0e, p. 56. 
