114 
Til E VOYAGE OF H.M.S. CHALLENGER. 
>\ 1 ' ' _ in a current of air, ami the air carrying with it the carbonic acid cauglit in the 
lMiyta-Ha.sk which has meanwhile been re-exliausted. The rest explains itself. 
T method was applied to a fresh artificial sea-water, the analysis of which had 
rir pi I 
_ \ ii (in mgrms. per litre) for the alkalinity 53*37 and 53*72 ; mean, 53*54 x mgrms. 
The total carbonic acid had been found to be llO’O mgrms. The alkalinity (on the 
assumption of lid): 2C0 S ) demands 107*1 mgrms. Hence, the free carbonic acid 
amounted to 2*9 mgrms. 
Three distillations, carried out as just explained, gave — 
(7) (8) (9) 
19 0 45 1 48-5 
m .M ins. of carbonic acid per litre of water, the loose carbonic acid according to calculation 
should have been 1 10*0 — 53*5 = 5G*5. The very low result in experiment (7) may have 
I* rn owing to an unobserved gross error in manipulation, although apparently the experi- 
ment went -in quite rightly to the end. Of the other two even No. 9 gave only about 
7 of the loose carbonic acid ; and it docs not agree with No. 8. 
The number 110*0 for the total carbonic acid was the result of a single analysis; 
hence, after experiment (9), I again determined the carbonic acid in the remainder of 
the water and found these values for it, — 
103-G5 and 105 35; mean, 104 5 
iiijrnn. of carbonic acid per litre. This is a considerable falling off, even against the 
■ 'in, - r l<»71 calculated from t he alkalinity, which is to some extent accounted for by 
1 • I t that tin- v at r, ince the execution of the original carbonic acid determination, 
had 1 ‘tiding in an only partially filled, though corked, vessel. But even supposing 
; * t • 1 "*. r one of tin two last carbonic acid determinations, there remains 103*G5 
— 1 :, l 5011 fi,r th loo-e carbonic acid, sis against the 48*5 found by Buchanan’s 
process in No. 9. 
• < "ii id. ml ilc quantity of the water left, which I utilised for studying the 
• dh ' ? on it of repeated treatment with air. The liquid (containing 104*5 mgrnis. per 
i m ic a< id) v accordingly shaken thrice, each time with 5 times its 
'* ' •md tic remaining carbonic acid determined. The quantities found were 
1 ; He m, 99*3 mgrms. of carbonic acid. The residue was subjected again 
in the ,-;inic manner : the carbonic acid now amounted to 94*8 and 
• i‘- 94*4. Alter another repetition of the operation, the quantity of 
tvirlwnic acid was 90*1 and 91*3; mean, 90 7. 
