106 FRIDTJOF NANSEN. M.-N. Kl. 



The values of Co//', i. e. the number of gram-equivalents of hydroxyl 

 ions (OH') per liter of sea-water, corresponding to the values found for 

 pn f are therefore given in the 5th column of Table III; but for the sake 

 of convenience the real values have been multiplied by io 7 . 



As the writer had no experience in this colorimetric method before 

 the cruise, and as some accidents happened to the apparatus during the 

 cruise, the results ought not to be considered as very trustworthy, 

 especially as the method in itself is not very accurate. 



Mr. PALITZSCH found [19123, p. 252] that the condition at some of 

 the stations of the Danish expeditions 1908 1910 suggested a relation 

 between the quantity of oxygen in the sea-water and its hydroxyl ion con- 

 centration (i. e. alkalinity), in this manner that a greater quantity of oxygen 

 is accompanied by a higher hydroxyl ion concentration (or a lower hydro- 

 gen ion concentration). Similar conditions have been found by Mr. GAARDER 

 by his numerous investigations in Norwegian fjords near Bergen, and by 

 Prof. B. HELLAND-HANSEN in the northern Atlantic in 1913 [1914]. 



Our observations may seem to agree with these results, in so far that 

 the hydroxyl ion concentration found shows, on the whole, a tendency to 

 decrease with increasing depth from the surface downwards; but this does 

 not seem always to coincide with the variations in the relative amount of 

 oxygen. The latter was found to be specially great, with supersaturation, 

 at 20 metres, and less near the surface. Our observations gave, however, 

 in most cases a distinctly higher value of the hydroxyl ion concentration 

 at the surface than at 20 metres 



It seems also strange that at Stat. 19 the hydroxyl ion concentration 

 decreases gradually from the surface and down to 310 metres, but below 

 that level it again increases with increasing depth. 



The much lower values of the hydroxyl ion concentration given by 

 our observations in the inner end of Cross Bay, at the Lilliehook Glacier 

 (in Porte Signe and at Stat. 13), as compared with those obtained at 

 Stat. 14 half way out the fjord, are somewhat puzzling, and seem diffi- 

 cult to account for, if the differences may not be due to some obser- 

 vational error; but they are rather big for such a possibility. 



