Contrasted with the behavior of oxygen, no certain effect of the atmosphere on the of the water as indicated by the pH was observed. At station A the oxygen curves were flattened out on July 13-15, due to high winds, but no flattening is observed on the pH or CO2 curves. On going north into colder waters no fall in pH was noted, except on a very cloudy day at the mouth of the Chesapeake Bay and in the pol luted water off New York Harbor, neither of which can be considered normal conditions. Palitzsch, however, noted that pH fell from 8.08 to 8.13, with more extreme drop in temperature in the North Atlantic and North Sea. This is not due to a difference in standards, since he found the pH of the Atlantic at about 41° north latitude and of the Bay of Biscay to be 8.22 and 8.25 respectively.
No local effect of the tension of the sea on the CO2 of the air was observed. The wind reaching Tortugas blew over the Florida Keys and alternate areas of water of CO2 tension= 3 to 3.5 in the morning and 2 to 3 in the afternoon. The of the air varied from 2.8 to 3.5 without any relation to time of day. Part of this variation may have been merely the expression of technical errors, but a large number of duplicate experiments, and of experiments designed to show the limit of error, failed to support the idea that the variations were entirely due to technical errors. The of the air was determined colorimetrically with 24 mm. sealed tubes containing the standards. The objection might be raised that the expanse of water of the same tension was not great enough to affect the air and that considerable time was neces sary to effect a change, and while one body of air was passing over the sea, both morning and afternoon variations in the water affected it alternately with neutralization of effects. The variations from day to day might have been due to slight shiftings of the wind and also to vertical currents in the air. In order to avoid such objections, Dr. A. G. Mayer made determinations over the Pacific Ocean. The standard tubes that he used were only 10 mm. in diameter, but the bicarbonate solution was kept in a gold-lined Jena flask and had not changed when tested on his return. The absolute standards may have been different from those used at Tortugas, but the method was adequate to show differences, provided equilibrium was always reached. The determina tions were taken at noon each day, except where otherwise stated. The results of my calculations from his data are given in table 11.
From table 11 there probably may not be recognized any correlation between CO2 tension of the sea and of the air. The cause for this may be that the apparatus was too crude or that exchange between air and sea is entirely masked by currents in the air. If we regard the air as sufficiently circulated to be of practically uniform composition, the changes in barometric pressure would not cause significant changes in CO, tension and the latter might be considered constant. If there
were sufficient exchange between sea and air the CO, tension of the sea would be practically constant. In table 11, the sea-water ex amined by Dr. Mayer on the trip from San Francisco to Hawaii (if we except the first station, where the water was polluted from land drainage and came in an insensitive region of the thymolsulfon phthalein tubes) had an almost uniform tension, although the tem perature rose 6.2°, thus necessitating in total a fall of 1.3 c.c. per liter and rise in pH of 0.08. This may not, however, have been entirely due to exchange with the air in situ, but partly to the remote history of the water. Palitzsch has shown that the pH of the open sea decreases as the depth increases. It is well known that temperature decreases as the depth increases, and the two factors would have antagonistic effects on the tension and tend to keep it constant. The reason for this stratification of the water may be that each stratum of water reached a certain temperature and tension while at the surface, and the temperature effect on the density caused the superposition of the strata. After being buried out of reach of surface currents, the temperature and total CO, change very slowly and hence the CO2 tension changes very slowly and may be practically the same when it reaches the surface the second time as when it left the surface the first time. Palitzsch's data are shown in table 8.
It is not intended to create the impression that the air regulates the tension of the sea. There is 20 to 30 times as much total in the sea as in the air. If there were no in the air and this condition could be maintained by some external agent, all of the CO2 in the sea would pass into the air, thus reducing the carbonates and bicarbonates to hydrates. In other words, the total of the sea is available for regulating the of the air. From the available data, however, it appears that the partial pressure of in the air at sea-level is less than the average tension of the sea. The reason for this anomaly is as follows: Part of the carbon fixed by photosynthesis at depths of less than 300 meters gravitates to greater depths and is oxidized, thus increasing the content of deep water. Some of this deep water is constantly being carried to the surface by vertical currents, but on reaching the light the CO2 is reduced again by photosynthesis and the original state of affairs is regained. Thus the sunlight interposes a barrier between the air and the main body of sea-water, so that the air contains less than it would if the sea were more rapidly stirred.