401 



2. DiTTMAR mentions {Challenger, Physics and Chemistry, 

 vol. 1, p. 213) the possible existence in the sea of veritable 

 springs of carbonic acid. That such springs may exist in the 

 sea as well as on land is, indeed, extremely probable but, so 

 far as I know, none have as yet been discovered. 



3. The river-water entering the ocean no doubt contains 

 a considerable part of its alkali in the state of bicarbonates, 

 and the amount of loose carbonic acid contained in these is a 

 source of gain to the sea. Mellard Reade [Journ. of Geol. vol. 7, 

 1899, p. 569) estimates the yearly quantity as 1.35 x lO'' tons. 



4. Interchange witli the atmosphere. Any difference in 

 tension existing between the surface of the sea and the atmo- 

 sphere will give rise to diffusion, the rate of which may be 

 determined from Bohrs formulas and constants of invasion and 

 evasion (quoted above p. 383) when the tension-difference is 

 known. In this respect the ocean may be compared with the 

 freshwater-lake mentioned above, but there is this great difl'e- 

 rence that, while the total amount of carbonic acid in the lake 

 is extremely insignificant compared with that of the atmosphere, 

 the amount of loose and free carbonic acid in the ocean is many 

 times greater than in the air. If ditferences in tension are found 

 the interchange of the gas between the sea and the atmosphere 

 will therefore have by far the greater influence upon the latter, 

 and I must refer for a more exhaustive treatment of the matter 

 to the following paper on the composition of the atmosphere. 

 Here I shall confine myself to give only the tensions actually 

 found in a series of water-samples from the Davis-Strait and 

 the North-Atlantic collected in Septbr. 1903 during the home- 

 voyage of a steamer. 



The first of these tables contains the tensions as they 

 Avere actually observed in the samples of water at 12°. 5. Two 

 samples of water were always collected simultaneously, and 

 1 gr. of sublimate was added to one of them. 



The tensions of the pure-water-samples, given in the first 

 XXVI. 2G 



