REPORT ON THE COMPOSITION OF OCEAN-WATER. 
oo 
oo 
From the united filtrates the lime is thrown down by addition of 20 c.c. of ammonia 
and 40 c.c. of oxalate of ammonia, and the mixture, which amounts in all to about 300 c.c., 
allowed to stand cold over night. In the morning it is heated over a water-bath, the 
oxalate of lime collected on a filter, ignited repeatedly (finally over the blowpipe), and 
the weight determined ; in the final weighings by first putting the approximately correct 
amount of weights on the balance, then placing the crucible, as it comes out of the 
desiccator, on the other pan, and lastly, without loss of time, determining the existing 
small overweight by the method of vibration. This operation of re-igniting and re-weigh- 
ing was continued until a series of at least six consecutive weighings gave practically the 
same result. We found it impossible to establish absolute constancy of weight ; but 
the sum total of the results gives me the conviction that the mean values ultimately 
adopted are right to within about ± O'l, or at most ± 0*2 mgrm. # 
The ignitions were effected in platinum crucibles of 20 c.c. capacity. The empty 
crucible, as it came out of the desiccator, required about one minute’s rest in the 
balance-case to assume a constant weight, which was then put down as “ the tare.” This 
tare, according to 10 experiments (5 with each of two crucibles), exceeds the actual 
weight, as it is when the lime is on the balance, by 0‘4 mgrm. Maximum observed 
= 0 ' 6 ; minimum = 0 ' 2 mgrm. 
This method was applied to the three mixtures of Challenger waters referred to, and 
also to a surface-water which had been collected for me at Port Louis, King’s Cross, 
Arran, for the bromine investigation. The analyses were executed in four sets, each of 
which included the four different waters, so that each set, in reference to the question in 
hand, could be discussed by itself and independently of the other three. The object of 
this was to eliminate, if necessary, the influence of any involuntary change in the modus 
operandi that might have taken place in passing from one set of analyses to another. 
The filtrates from the crude oxalates, and from the pure oxalates, of lime ; also the 
“ sesquioxides ” and the “pure lime precipitates ” obtained in the several analyses were 
collected, and, at the end, examined as follows : — 
The sesquioxides, from 16 analyses, w r ere ground up, re-ignited and weighed. Weight 
found = 16 ’6 mgrms., which yielded on analysis : — 
Mgrms. 
Residue left on dissolving in hot hydrochloric acid, . 
6-9 
Purified sesquioxides, ..... 
7-2 
Lime, ....... 
1-7 
Magnesia, ....... 
0-4 
* The following is the worst specimen of inconstancy observed: — Crucible + CaO = 19'84 + [3 0, 2'4, 2 - 7, 2‘C, 2 - 3, 
27, 2’8, 2-4, 2 - 4, 2‘1, 2 - 6, 27 mgrms.] 
(PHYS. CHEM. CHALL. EXP. PART I. 1884.) A 5 
