14 
THE VOYAGE OF H.M.S. CHALLENGER. 
the precipitate (a mixture of chloroplatixiate of potassium and sulphates) with ether-alcohol 
< l \ In:. ■ f .ili uln»l+ ! volume of ether), — decanting through a filter, — then dry the pre- 
. } i;it . .tii.l r lm • it in the porcelain basin in which it has been produced, by placing a 
funnel over it, passing hydrogen in by the stem, and heating to about 300°C. From 
. it nu-t what is not metallic platinum (1) by water, and (2) by hydrochloric 
ollect the metal on a filter, ignite and weigh. The weight multiplied by 0*4747 
m ght of the potash (K 9 0) contained in the water. The method, when applied 
tii. synthetically prepared sea-water above referred t". gave the following results : — 
Water taken. 
Grins. 
(1) 21-28 
(2) 53-27 
Potnanded by synthesis, 
Platinum obtained. 
Mgrms. 
38-6 
94-7 
Potash found 
per 1000 parts. 
0-861 
0-844 
0-879 
I -• r> suits, although not what I could have wished them to be, contrasted favourably 
’ h tl ■ - previously obtained by the ordinary method, and encouraged me to enter 
up '» a systematic series of test analyses, chiefly with the view of obtaining a higher degree 
1 pr« < i-ion by a greater familiarity with the modus operandi. Known weights of the 
pur< sulph it< - "f potash and soda were weighed out, dissolved in water and wrought accord- 
ii- t<> ti modified form of Finkener’s method. The results are given in the following 
table, in which the unit of weight is the milligramme : — 
© © 
Taken of Obtained. 
PoUaciam »ulplute. 
Sodium gulp hate. 
plathnun. corresponding to 
Potassium 
sulphate. 
<») 
A’iV. 
500 
24 
2-1 
( 2 ) 
562-0 
Nil. 
63G-5 
562-2 
( 3 ) 
406-8 
1000 
449-2 
397-0 
(4) 
521-0 
500 
589-6 
520-7 
< 5 ) 
418-5 
1000 
471-4 
416-4 
(6) 
399-2 
1000 
450-5 
397-9 
Att. r tie . \p. rim* nts, 1 felt sure that the method when applied to sea-waters, with 
die n m o to the method once laid down, would at anyrate give fairly constant 
r *1 n, i < ept ibl<- of .-ubsequent correction, and sure to be closer approximations to the 
J-1* ’ h m < * add have been obtained by any other known process. The method, accord- 
: 1 1 i) ple d t<. tie- first series of waters as above explained. The results, however, 
' • / • ; . . 1 1 1 thou -quoted niMiveie first trials (with synthetical sea- water), were not 
: ’ <•' ‘ tie v would necc-sarily have la-en if they had been affected by only accidental 
:r r I •' r. f .r< wle-n the second ■ riesof waters came to hand, J caused Mr. W. G. Jolm- 
• f " n - ' ’*e niy i - -it . mts, to try experimentally wlu-t her the substitution, for the adopted 
i r i r ‘ } -.it . iii< ehloride (I*t('l 4 ), of another proportion, or the use of chloroplatinate 
• 4 : 'bum or lPiiium, would ensure greater constancy in the results. The results of these 
"■ ih ar* o,\. n in th« following table, for the interpretation of which it is only necessary to 
