OCEANOGRAPHICAL RESULTS, MICHAEL SAKS, 1900. 147 



But in the sea still farther west we found the temperature 

 of the upper water-strata higher in July and August, 1900, than 

 observed in previous years. At Stat. 29 (August 9, 1900) our 

 temperatures at depths between m. and 200 m. were higher 

 than those of the North Atlantic Expedition at Stat. 219, on July 

 28th, 1900 (PI. 4, Fig. 8). 



In the sea near Iceland, our temperatures of the upper water- 

 strata are conspicuously higher than those observed by the Nor- 

 wegian North Atlantic Expedition in August 1876 (see our Stat. 

 10 compared with Stat. 48 of Aug. 6, 1876, PI. 4, Fig. 3) 

 and by the Danish Ingolf Expedition in July and August, 1896 

 (see PI. 4, Figs. 4 and 5). But a fact of special interest is that 

 at the same time our temperatures at the lower depths are at 

 several stations strikingly lower than observed at nearly the 

 same places during the Ingolf Expedition, see especially our 

 Stat. 14, northwest of Iceland, compared with Stats. 99 and 129 

 of the Ingolf Expedition (PI. 4, Fig. 5). 



This proves evidently that while the warm Irminger Current 

 at the sea-surface has been more developed in August, 1900, 

 than in August, 1896, the cold under current, at depths below 

 200 m., has also been more developed in 1900. At Stat. 17 we 

 find very much the same conditions, if our temperature curve 

 (see PI. 4, Fig. 6) be compared with those of the three sur- 

 rounding Stations of the Ingolf Expedition. 



If the salinities found at various depths of the sea in the 

 several years be compared in a similar manner, we might pos- 

 sibly find similai- differences. But the difficulty is that the 

 salinities of the various expeditions have not been determined 

 with sufficient uniformity ^ to make the results of such a com- 



^ I believe for instance, that ttie great • differences found by Dr. Hjort 

 in tlie values of the salinity of the upper strata of the Gulf Stream 

 betw^een Norway and Iceland in May, 1896, and May, 1897, may to a 

 great extent be due to differences in the solutions used for the titrations. 



