1906. No.3. AMUNDSEN’S OCEANOGRAPHIC OBSERVATIONS IN 1901. 97 
If it be assumed that the true salinity be only 35'00 900 its density (61) 
at —o 9’ C. would be about 28'17, and if the salinity be 35'10 0/00 the 
density would be 28'25. At that time it therefore seemed necessary to 
assume that this particular bottom-water was from the intermediate 
warmer strata (between 250 and 700 metres) which had been cooled 
down to its lower temperatures somewhere in the still unknown parts 
of the North.Polar Basin itself. But this sea is covered by a layer, at 
least 200 metres thick, of lighter cold water of very low salinity, de- 
creasing upwards from about 34'7 %00, at 200 metres, towards 30 %oo or 
32 %0 at the surface. It seemed therefore hardly probable that water 
with a salinity of abouth 35 %0, could occur near the surface anywhere 
inside the area of this sea!. It was therefore thought that the only way in 
versing thermometers on cold winter-days may have a tendency to be too low; and the 
series of temperatures taken during the summer, on days when the temperature of the 
air was very nearly the same as that of the water, are therefore the more trustworthy. 
The final values of these series, given in the Memoir pp. 244—256, may be expected 
to have a fairly high degree of accuracy, at least as compared with those of most 
other expeditions. The writer thinks that especially the series of temperature taken 
.in June and August, 1894, and those taken in July, 1895, are very good. As the 
conditions are so extremely uniform, especially in the deep layers of this basin, as 
proved by all the observations, these series give very reliable information about the 
vertical distribution of temperature in the bottom-water of that sea, and as it is known 
that the density of this bottom-water must be very nearly uniform at the same levels 
throughout the whole basin, there cannot possibly be much difference in the temperature 
of the deep layers in any other part of it; unless it be assumed that the salinity also 
differs much which is highly improbable. 
It must therefore be assumed that hardly anywhere in the North Polar Basin can 
the minimum temperature of the bottom-water be much below —0'9? C. 
In the connection a temperature reading of —r'ı4°, on October 27, 1894, for 
2500 metres may be mentioned here. Corrected by the instrumental error and error 
by cooling in water (—0'04, see I. c. pp. 241— 242), this reading would give a tem- 
perature of —1'18? C., which is, however, an impossible value since a temperature of 
—0'84° C. at 3000 metres was observed on the same occasion, it would give the water 
a too high density, and would make it sink rapidly to the bottom, unless it be also 
assumed that just at this one depth the water happened to have been of a salinity 
much lower than that of the bottom-water as found at all other Stations in the Polar 
Basin; which is hardly possible. The air-temperature on that day was —33° C., with 
a wind of 3 metres per second; and it is clear that the termometer has been exposed 
too long to the air before being read off; the broken off mercury was cooled down 
some twenty degrees to that the correct temperature must have been between —0'8 
and —o’g’C. If the mercury had had time to assume the air temperature, the reading 
corrected for instrumental error should have been reduced by + 0‘45° C. which would 
have indicated a temperature of —o'73° C., and that is evidently too high. (In the 
footnote about this temperature-reading, op. cit., p. 252, there is a mistake, the reading 
being stated to be — ı'24° C., instead of —ı'14° C.). 
+ 
The only possibility would be that north of Spitsbergen and north of Novaya Zemlya, 
the vertical circulation during the winter, might give the surface water a comparatively 
high salinity (see later). 
Vid.-Selsk. Skrifter. I M.-N. Kl, 1906. No. 3. 
