12 
THE VOYAGE OF H.M.S. CHALLENGER. 
Th. crude pyrophosphate, as it gave only 99*49 (say 99-5) per cent of “pure” sub- 
• -h- >iiKl l • diminished by 0*5 per cent, and then be increased by the magnesia con- 
• d t!..' lira. . Now the quantities of lime averaged about 34 mgrms.; those of pyro- 
34 x ’022 
►out 3 1 0 mgrms. : hence the correction should be = —-1*55-4 ^gg— = +0*53 
ni jnu. I f it v. rv diffident about the validity of this slight correction, and at last decided 
fitting it. In the tabular statement of analytical results included in this publica- 
amounts of lime and magnesia are all corrected (or recorrected) according to this 
• it. no nt : the magnesia only for the filter-ash, the crude lime by multiplication 
by 0*91. 
Determination of the Potash. 
Tli i ~ o iv, me a great deal of trouble. The customary method, as applied to sea-water, 
: Id 1 .. . aft.-r (or perhaps without) elimination of the sulphuric acid, magnesia, and lime, 
• . , . j i \ rt alt the bases into chloroplatinates by evaporation with a large excess of chloride 
]•! it ilium, and from the residue (which must contain the Na 2 PtCl 6 in the hydrated form) 
t" 1; \ivi:ite out what is not chloroplatinate of potassium, either by means of aqueous 
.]■ h 1 (Fresenius), or by means of small instalments of chloride of platinum solution 
( dning 5 per cent, of metal), followed by alcohol (Tatlock). I prefer Tatlock’s form 
• f tb.- process, a.s it is in a high degree independent of the presence of salts of lime and 
_• and of sulphates, and consequently directly applicable to sea-water. That this 
pp. - gives good results with anything that fairly falls within the heading of even 
ip “ potash salts,” is well enough proved, but I felt very diffident as to its giving a 
-ufti. ;• ntly close approximation in the case of sea-water, where the potash forms such 
. sin 11 proportion of the whole. My suspicions were confirmed by the following two 
t- ' vdy •• -, in which known weights of pure chloroplatinate of potassium, after being 
lv 1 in a solution containing about the same amount of chloride of sodium as the 
im of the former is associated with in sea-water, were recovered by Tatlock’s 
method : — 
(1) 
(2) 
Chloroplatinate of potassium taken, 
. 123 
115 
mgrms. 
Chloride of sodiain taken, . 
. 1350 
1270 
99 
Metallic platinum in the precipitant used, . 
. 2500 
2300 
»> 
( hloroplatinate of potassium recovered, 
. 108-2 
102 
O 
* 99 
Loan, ..... 
12 
11 
per cent. 
\ * • p It v. i-. obtained on the application of the method to an artificial sea-water 
• ; ip : * :.■• . dly from pure materials. Ten cubic centimetres of the “ sea-water ” 
‘ yi* , 9*40 mgrms. of potash (K 2 0), gave 41*8 mgrms. of chloroplatinate 
* to S 0G m.'rms. of potidi (K,/l). Loss =14 percent. These three test- 
