8 
THE VOYAGE OF H.M.S. CHALLENGER. 
.11 I.f the values of N which is = 50'265 grms. The mean value for S — N is 1T29. 
For the deviations from the mean we have — 
(1) (2) (3) (4) 
+ 0002 -0002 +0005 -0005 
i - . at mast .., = 0 0001 of the most probable value. 
T - dews that my apprehensions, as far as my sea-water analyses were concerned, 
n ' foundation. At a subsequent date I resumed the question from a more general 
• nd-p..int, and found that there are a number of theoretically unobjectionable salts 
. it. of m.igm sia, alkaline phosphate (even ortho-phosphate), nitrate of lime, alkaline 
rat- t, which, when present in large quantities in a chloride solution obscure the final 
• . - in Vuillard's titration-process to such an extent as to render it useless. But even 
. :i.. i-. '. if the proportion of the impurities does not exceed certain limits, their 
• « m 1..- compensated by the addition of a very large proportion of iron alum. The 
!-, which would be out of place here, will be published in a separate memoir. 
Determination of the Sulphuric Acid. 
T . following method was rigorously adhered to : — 
•. f tie sea-water were weiyhed, mixed with 5 c.c. of a chloride of barium solu- 
ut lining . Bout ] Ba] mgrms. per c.c., and 2 c.c. of 20 per cent, hydrochloric acid ; 
n. \tur- w heated on a water bath, and then allowed to stand over night. The pre- 
• ■ •!:■ t. 1 on a Swedish filter, washed first with very dilute hot hydrochloric 
. ■ i • : i vi-’i le»t water, ignited in a platinum crucible, and weighed. Each series 
•••!■!!. In iti*«ns was controlled by a “ blank '’ with 20 c.c. of pure water instead of 
• r. m l tie- - nin- quantities of reagents as were used for the latter, and the filter- 
a fin nut- ti :■ •• of Ba.SO, subtracted from each of the precipitates as a correction. I 
•• !y r- fi lin' d from any purification of the sulphate of baryta, being afraid that 
. i lb!* lo.--.i-- involved in any such process would be greater than the impurities 
in the crude precipitate. 
\ i !’!■ number of determinations had been thus made, when a doubt arose 
t > tie eonvcl nens of my judgment in regard to the point just touched upon, 
I : m.i'h tin- following synthetical cxjierimcnts : — A kind of artificial sea- 
• ' pi - 1 i d l»v dissolving 1 '02 grms. of pure lime (CaO) and 3'3 grms. of 
'» (M in 20 c.c. of normal hydrochloric acid (1 c:c. = [HCl] = 36'5 mgrms.), 
d: j i , .' ' j.« "lution of 42 grms. of pure chloride of sodium, and diluting to 
•- 1 4 • '-V' q i< • for the 75 c.c. of standard sulphuric acid to be added in the 
trail*. 
' ' dphuric acid used had been prepared shortly before (for other 
* n /irding th«> [.rei«ralibti of f-ure magneiiia, see lelow, p. 1G. 
