REPORT ON THE COMPOSITION OF OCEAN-WATER. 
41 
Summary. 
(Total amount of Salts per kilogramme of water, carbon dioxide being deducted. 
Method adopted. 
No. 332. 
No. 347. 
I. By full analysis, ...... 
II. By determination of chlorine and my factor, x deducting C0 2 , . 
III. By the Norwegian method, ..... 
IY. By determination of chlorine and Norwegian factor, 
Difference between III. and I. 
„ II. „ I 
„ IV. „ HI 
„ „ IY. „ I 
34-797 
34-751 
34-992 
34-866 
+ 0-195 
- 0-046 
-0-126 
+ 0-069 
34-582 
34-526 
34-755 
34-640 
+ 0-173 
- 0-056 
-0-115 
+ 0-058 
Taking the result obtained by the first method as our standard, it appears that the 
Norwegian method slightly over-estimates the total salts, and that, starting from the 
chlorine, my factor gives a deficit in'tlie total salts of about (P051, the Norwegian factor 
an excess of about 0'063, or less than the Norwegian method of determining the solids 
by about 0T20. But in my opinion method II. is the most exact, and consequently my 
two full analyses slightly overstate the total salts, while the Norwegian method gives 
somewhat high results, and their factor, as applied to my chlorine determinations, is too 
high. But the Norwegians used a method of chlorine-determination of their own — which 
I think very well adapted for its special purpose — and I have no doubt that their 
determinations of total solids, as deduced from their own amounts of chlorine, and for 
the purpose of comparison, are more exact than would appear from the above table. I 
now proceed to state the results of my chlorine-determinations in the following Table I.: — 
Column I. gives the number assigned to the sample by Mr. Buchanan. 
Columns II.-V. are explained in the heading of the table. 
Column YI.— The depth “ D ” of the ocean at the place, given in Column IV., where 
the sample was collected. 
Column VII. — The mean “y” °f 3 in general, two determinations of the chlorine, in 
grammes per kilogramme, by means of my refinement on Volhard’s method, as described 
in Chap. I., p. 4. 
Column VIII. — The mean deviation “ Ay” of the mean y from the individual results ; 
hence, in most cases, half the difference of the two analyses made. A blank in this 
column indicates that the respective y was transcribed from the report on a complete 
(PHYS. CHEM. CHALL. EXP, — PART I. — 1884.) A 6 
