REPORT ON THE COMPOSITION OF OCEAN -WATER. 
241 
presumed to be chemically pure, nor was I absolutely sure of the purity of the chloride 
of sodium; but as the sea- water was intended merely for critical trials of my method of 
bromine determination, the preparations were accepted as sufficiently pure. A quantity 
of the sulphate of potash, representing the potash present according to my sea- water 
analyses beside 160 grams of halogen reckoned as chlorine, was weighed out, which at the 
same time gave part of the sulphuric acid. The rest of the latter was added in the form 
of normal sulphuric acid as used for alkalimetry. The correct volume of this acid, 
together with a measured volume of normal hydrochloric acid, served to dissolve 
the magnesia ; but as an excess of the latter had been employed intentionally to 
hasten on its action, this excess was subsequently compensated by the addition of 
the correct volume of a standard solution of pure caustic soda. The carbonate of 
lime was dissolved in ^ c.c. more than the calculated volume of normal hydrochloric 
acid, and the excess of acid neutralised by the caustic soda solution. The several 
instalments of chlorine thus introduced were added up and made up to 160 grams by 
addition of the calculated weight of the chloride of sodium. The magnesia, potash, and 
chloride of sodium solutions were mixed and diluted to 4 kilograms. On the other hand, 
the lime solution was made up to 42 13 '9 grams in a separate flask. The two solutions 
were then mixed to produce 8213'9 grams, equal to very nearly 8 litres, of “ sea- water.” 
This water on standing deposited a small gelatinous precipitate which settled fir m ly. 
The clear supernatant liquor was drawn off by means of a syphon and preserved in well- 
stoppered bottles. 
The chlorine in a known weight of this water was determined by my gravimetric 
modification of Yolhard’s method, and found equal to 1 9 ‘455, 19’448; mean, 19'452 
grams of chlorine per kilogram of sea-water. 
To test this water for bromine, 1 kilogram of it was precipitated with 50 c.c. of acid 
cleci-normal silver solution (see page 99), the precipitate collected and washed, by 
decantation, dried, weighed, and chlorinated. The chlorination involved a loss of 0'20 
mgrms., corresponding to 0'36 mgrms. of bromine ; but whether it really does so or is a 
mere observational error, the water may be accepted as being practically free from bromide. 
6. To pass now to the critical experiments, the first question which we sought to 
answer was whether fused chloride of silver, when heated in dry chlorine, and then allowed 
to cool in a current of dry air, as had been done in the analyses of the Challenger sea- 
waters, retains its correct weight. To enable the reader to form his own opinion in regard 
to the unavoidable uncertainties in the weighings, I give all the weighings which were 
made in the first experiment. 
Tube and boat, ...... = 36 '9073 grams. 
Do. + chloride of silver as it came out of the bottle, = 39 "7358 „ 
Chloride of silver taken, 
(PHYS. CHE.M. CHALL, EXP. PART I. 1884.) 
A 31 
2-8285 
