I912. No, 12. THE SEA WEST OF SPITSBERGEN. 21 



metres) and — 1.3^ C, his salinities vary between 35.00 and 35.08 "00 

 (only in one case has he a salinity of 34.94" 00 in 2000 metres). In these 

 cases the error has evidentl}' varied between 0.08 %o and 0.16 "^/oo, at 

 least. It is therefore hardl}- possible to compare Hamberg's salinities with 

 those of Isachsen's stations. The salinities of Arrhenius' Stations 3 — 6 

 north of Isachsen's Section \'I are not sufficiently trustworth}- for a com- 

 parison. 



We think that our above discussion of the variations in the salinity 

 of the Atlantic water of the Norwegian Atlantic Current (in the Lofoten- 

 section, and the Sognefjord-section) proves the improbability of any great 

 annual variations in the salinities of the Spitsbergen Atlantic Current. 

 The views held by the Svenska Hydrografisk-Biologiska Kommissionen 

 [191 1] in this respect cannot therefore be correct, as we think is even 

 indicated by the observations of the Swedish expedition of 1908, given 

 in their paper [191 1, p. 18], The temperature of the Spitsbergen 

 Atlantic Current was probably comparativel}' high that year in about 

 77*^44' N, Lat,, as will be pointed out later. If the views of the Swedish 

 oceanographers were correct, we might therefore expect unusuall}' high 

 salinities in the current that summer, but the maximum salinity at their 

 only station in the current (see Fig. 21, G i) was 35.10 "00 in 100 metres, 

 while the maximum salinity at Isachsen's Stat, 28, some distance to the 

 north, was 35.03 "00 in July, 1910, when the current was probably somewhat 

 colder^. The Swedish oceanographers have evidently not been aware of 

 the great stability of the salinities of the Norwegian Atlantic Current in 

 the region to the south; a stability which is proved by all observations 

 during recent years. 



The Swedish oceanographers maintain [191 1, p. 7] that the high sali- 

 nities observed at Arrhenius' Stats. 3 — 6 are perfectl}' trustworthy. We 

 think that the vertical series of observations themselves at these stations 

 prove sufficiently that this cannot be the case. By computing the densities 

 from the observed temperatures and salinities one finds absurd values, not 

 to mention that at most stations water with comparatively high density is 

 frequently resting on much lighter water, A few examples may be men- 

 tioned. 



^ The Swedish Stat. II was near the coast (in 78" 5' X., 12" 50' E.) and within the 

 waters of the Atlantic Current, The saUnity of 35.14 *^, o» found in 200 metres at this 

 station is erroneous, as is clearly proved by the density of 28.03, 'he density of the 

 underlying water being 27.89, and that of the overlying 27.86. 



