1906. No. 3. AMUNDSEN’S OCEANOGRAPHIC OBSERVATIONS IN 1901. 63 
at Station II, the releasing propeller having obviously not functioned at 
the desired depths. During the first cruise of the “Michael Sars” in the 
same summer, 1900, there were on board several water-bottles and among 
them two instruments of exactly the same form as that used by Öster- 
gren, and made simultaneously by the firm L. M. Ericsson in Stock- 
holm. The bottles were closed by a releasing propeller; but it was 
found that the propeller very often failed to release and close the bottle 
at the desired moment, especially at great depths; a fault which was 
afterwards remidied, and does not occur with bottles of the newer 
pattern. Östergren has evidently had exactly the same experience 
with his water-bottle of this same make (at Station II) without 
noticing it. His temperatures for 1000 and 2000 metres (see Section IX) 
are much too high!; the temperature and salinity as given for 1000 
metres indicate that the bottle has on that occasion been closed some- 
where between 500 and 700 metres, where there was also a salinity 
similar to that of the bottom-water. The salinity optained from 2000 
metres is impossible, and indicates that the bottle has probably been 
closed at a depth of about go metres instead of 2000 metres. According 
to the list of observations, which Prof. Pettersson kindly sent me, another 
observation was also taken at 1500 metres, which, however, gave a 
temperature of 028" C., indicating that the bottle had been closed 
somewhere in the intermediate warm water-stratum. The authors left 
out this observation in their published Table. It is consequently seen 
that at this Station the water-bottle has probably worked properly down 
to 700 metres, but below this depth no observations taken are trust- 
worthy. 
The salinities published by Pettersson and Östergren are evidently 
computed according to Pettersson’s former method, and are consequently 
higher than those obtained by Knudsen’s Tables. The difference is some- 
what higher than might be expected. To judge from the salinities given 
of the cold bottom-water at both Stations, which are very uniform, the 
salinities are evidently about 0'10 %/,, too high and have therefore here 
! According to the observations published, Östergren already found the typical bottom- 
water of the Norwegian Sea at 700 metres, with a temperature about —o'5° C. and a 
salinity about 34°93 °/,0, and at this depth the water-bottle has probably closed 
properly. But if so, it is impossible that the temperature could have been higher 
at Iooo and 2000 metres but the salinity evidently very nearly the same; such 
conditions are against physical laws, and are found nowhere in the bottom-water of the 
Norwegian Sea. It is of course possible that in the deepest hollows of the basin there 
might be a slight rise of temperature towards the bottom, owing to the uuderground 
heat of the lithosphere. 
