THE VOYAGE OF H.M.S. CHALLENGER. 
220 
D'4 to 1 but in most cases it is about 15° C. The mean loose carbonic acid for these 
is 42*35, i.e., barely more than is found in the Atlantic at high temperatures. 
The Pacific surface water would appear to be less rich in carbonic acid than that of 
the Atlantic. 
From all the evidence afforded by the Challenger research we see : — 
1st. Free carbonic acid in sea-waters is the exception. As a rule the carbonic 
acid is less than the proportion corresponding to bicarbonate. 
2nd. In surface waters the proportion of carbonic acid increases when the 
temperature falls, and vice versa. 
3rd. Within equal ranges of temperature it seems to be lower in the surface 
water of the Pacific than it is in the surface water of the Atlantic Ocean. 
We explained on page 213 how the atmosphere contributes towards keeping up the 
balance of the carbonic acid in the ocean. As a consequence the ocean, but in a far 
higher degree, acts as a regulator of the carbonic acid in the atmosphere. 
To Schldsing belongs the credit of having been the first to point this out clearly. 
In speculating on the quantity of carbonic acid in the atmosphere, he comes to the 
conclusion that its balance is maintained chiefly by the terrestrial volcanic carbonic acid 
springs and by the ocean, the combined influence of plant and animal life sinking into 
insignificance in comparison with these two agencies. 
1 think he is right ; but if so, then the proportion of carbonic acid in the 
atmosphere should be more in the tropics than in the temperate zones and polar regions. 
In reference to this conclusion, the numerous determinations of atmospheric carbonic 
acid which were made by Th:rpe in the Irish Channel in August 1865, and subsequently 
during a voyage to Brazil in 1866, may be quoted. 
'Flic air over the Irish Channel (lat. 54° 2 1 7 ), in 26 analyses, gave a mean value of 
3*089 volumes of earbonic acid per 10,000 volumes of air analysed, the extremes being 
2*92 to 3*32. In the air collected during his voyage to Brazil he found (in 51 samples) 
from 2*70 to 3*26 volumes per 10,000 ; mean 2*953. The difference is in favour of the 
Iri-h Channel, but it is less than the limit fluctuations on either side. Possibly the differ- 
'*ii'*e would have been less or even have become negative if the Irish Channel analyses 
bad b«-<-n made in winter instead of in August. In a most suggestive memoir on the 
•ubjoot by Dumas, I find it quoted that Franz Schulze of Kostock, in a very numerous 
ly*" -■ made there during 1^69 71, found the proportion of carbonic acid to 
I* subject to very slight fluctuation, and to have amounted on an average — 
In 1 H 60 (the whole year) to, ....... 2*8668 
.. I*™ ....... 2*9052 
„ 1*71 (tint six month*) to, ....... 3*0126 
