420 
THE VOYAGE OF H.M.S. CHALLENGER. 
rapid ovei'sight of the work done or remaining to be done. When completed, the diagram appeared 
to give so full and clear a representation of the whole course of the analysis that it was decided to 
print it without alteration rather than to attempt a written account, which could hardly have failed 
to bo difficult to follow. The diagi-am is therefore reproduced on the accompanying sheet. A brief 
account of the cpiantitative determinations not included in this diagram must, however, be given. 
Analysis of the Insoluble Residue left on treating 200 grammes of the Nodules ivith Hydro- 
chloric Acid . — The composition of this residue, which was dried in vacuo over sulphuric acid to 
constant weight, proved, after exhaustive qualitative airalysis, to be comparatively simple. No 
special difficulties were therefore met with in the course of the quantitative analysis. 
The loss on ignition was found to correspond exactly with direct estimations of the water con- 
tained in the residue after drying in vacuo over sulphuric acid to constant weight. 
Table III. gives the jDercentage composition of the water-free residue. 
A prelimiuaiy determination of silica gave 13’22 per cent. A second most careful determination 
with a larger quantity gave 13 '3 8 per cent. This latter number was adopted as being certainly the 
more accurate. 
On treating the crude silica with ammonium fluoride and sulphuric acid, a residue, consisting 
chiefly of titanic acid, was left. This residue was of course allowed for in calculating the percentage 
of silica. 
The titanium was estimated with great care, being precipitated repeatedly by boiling the dilute 
sulphuric acid solution. 
Analysis of the Aqueous Extract . — 46 732 grammes of the powdered nodules were exhausted 
repeatedly with boiling water. 
The complete extraction with water proved to be exceedingly tedious, and the solution was 
only obtained clear after repeated filtration. Aliquot portions were used for the determination 
of — (rt) Total bases as sulphate ; (b) potassium and sodium ; (c) lithium ; {d) chlorine ; (e) sulphuric 
acid. Traces of calcium and magnesium were also found and estimated. 
For the quantitative composition of the extract see Table I., column III. 
The residue, insoluble iii water, still contained sulphates, but no chlorides. The sulphuric acid 
was determined in a separate portion of the nodules. 
Estimation of ^Yater. — 7 5337 grammes of the powdered nodules lost 1*7330 grammes on 
drpng at llO® C. to constant weight. This is equivalent to 23'00 per cent. 
A direct determination of the water evolved on heating to redness in a platinum boat gave, on 
the other hand, 29'83 per cent, of water. As the water collected in the bulb of the absorption tube 
was slightly acid, two further direct estimations were made, using freshly-ignited oxide of lead to 
keep back acid vapours. These determinations gave 29 64 and 29'67 per cent, of water respectively. 
The mean of these determinations, 29 65, was adopted. 
EstinuUion of Peroxide-Oxygen . — These determinations were made by Bunsen’s well-known 
method. 
'I’he standard thio-sulphate solution used was titrated against pure iodine, prepared according 
to StfUi. 
Three determinations gave 4 67, 4’71, and 475 per cent, of peroxide-oxygen respectively. The 
mean of these determination.s, 4 71 per cent., was adopted. 
A.Hsuming the whole of this peroxide-oxygen to be present as MnOg, the percentage of this 
comjKmnd would be 25'61, which corresponds to 20 90 per cent. MnO, as against 21'46 actually 
