470 
THE VOYAGE OF H.M.S. CHALLENGER. 
(6) In Hydrochloric Acid Extract from Acetic Acid Residue. 
Silica, 
0.\ide of lead. 
. 0-01 
Oxide of copper, . 
0-272 
Oxide of cobalt. 
. 0-25 
Oxide of nickel, . 
Manganous oxide. 
Loose oxygen. 
Lime, 
Magnesia, 
Alkalies (R,0), 
Alumina, . 
Ferric oxide. 
0-40 
7-47 
0-93 
19-39 : 35-5 = 0'546 
3-95 : 8 = 0-494 
1-33 
1-42 
0-34 
3-03 
16-20 
(c) In Sulphuric Acid Extract from Hydrochloric Acid Residue. 
Alumina and ferric oxide, . . . . .1-62 
Silica, . . . . . . . .0-83 
{d) Ultimate Residue. 
Silicates and Silica, . . . . .14-91 
98-18 
Special Experiments cm. the State of Oxidation of the Manganese. 
The loose oxygen reported above had been determined in two ways, viz., firstly by Bunsen’s method : dis- 
tilling with hydrochloric acid, and titrating the iodine equivalent of the chlorine liberated by means of thiosul- 
phate — chemically pure iodine serving as a standard ; and secondly, by Fresenius and Will’s method : 
digestion of the substance with dilute sulphuric and oxalic acids, collecting the carbonic acid liberated in a 
tared potash bulb and soda-lime tube, and determining the increase of weight shown by the absorption 
apparatus. In the latter case the carbonic acid of the carbonates was determined in a separate portion of 
substance, setting it free by means of a mixed solution of ferrous chloride and hydrochloric acid and weighing 
it as above. In order to see whether the second method is affected by the presence in the substance of ferrous 
oxide (as Bunsen’s undoubtedly is), a quantity of a pure “ peroxide ” of manganese was made by heating pure 
nitrate first to about 200“ C., then to redness, and the percentage of loose oxygen in this preparation determined 
according to Fresenius and Will ; first in the usual manner and then after addition to the substance of a 
known weight of artificial ferroso-ferric oxide (FegO^) prepared in the wet way from ferrous sulphate. 
The results were as follows : — 
Percentage of Loose Oxygen fouiul. 
By the oxalic acid method, ....... 7 99 8-13 
By the same in presence of FcjO^,* . . . . , .7-98 
Hence the presence of ferrous oxide does not sensibly affect the oxalic acid method, which at the same time 
showed me that the substance of the manganese nodules analysed could not have contained much ferrous oxide. 
In fact the 3-95 per cent, of loose oxygen reported in the summary were deduced from the following deter- 
minations : — 
Oxygen found by oxalic acid. 
Oxygen found hy iodine method, 
DilTercnce, 
Manganous oxide found. 
4-02-0-502X “0" 
3-88-0-485X “0 ’’ 
0-017 X “0” 
19-39-0-546 xMnO 
The difference (0-017 x “ O ”), if not simply due to observational errrors, would correspond to 0-017 x FegOj 
— 0017 x72=l‘22per cent, of ferrous oxide = l-36 per cent, of ferric oxide, leaving 16-2 - 1-36 = 14"84 of 
' MnO . 0 — 0-6454 grm. ; FcjO^— 018 grm., CO, obtained — 0 2832 grm. —7-98 per cent, of oxygen. 
