86 FRIDTJOF NANSEN. M.-N. KI. 
too great a quantity of heat and having too rapid horizontal circulation 
to make it possible for the whole bulk of water to be cooled suf- 
ficiently by radiation to attain a density as high as that of the underlying 
bottom-water. No bottom-water can therefore, as a rule, be formed in 
these regions, and the same is also the case in the regions towards the 
south. å 
It is of interest to examine what may probably occur during the 
winter at a place like Amundsen's Station 14, which is the place where 
the bottom-water was found nearest the surface. The surface-water with 
a salinity of 34°39 oo would be soon cooled down below —1:5° C. (its 
freezing point is at —ı'875° C.) and become heavier than the underlying 
water. It would sink and be replaced by somewhat warmer water with 
higher salinity. But this new surface-water will be cooled down in its 
turn till it becomes heavier than the previous surface water; it will sink 
still deeper and be replaced by warmer water with a still higher salinity 
from below. In this manner the salinity of the uppermost strata will 
be continually raised and approach that of the bottom-water (about 
34'9 °/oo); the depth of the vertical circulation will increase until it be- 
comes operative through the upper lighter strata and reaches down into 
the typical bottom-water. All strata, from the surface downwards, will 
then have attained a nearly uniform temperature, salinity, and density. 
After this the cooling at the surface will produce such heavy water 
that it may sink far down into the bottom-water and even to the bot- 
tom itself. 
As the water which is to form the deepest and coldest layers of 
the bottom-water has to sink down through all intermediate layers, it 
must be expected that the temperature of this bottom-water, when first 
formed at the surface and before beginning to sink, is lower than the 
lowest temperatures observed near the bottom. It must be remem- 
bered that the temperature of the water cannot but rise slightly during 
the sinking, for it has to pass through strata with slightly higher 
temperatures, with which it will be more or less intermixed, and the 
! If ice be formed on the sea-surface, the salinity will be more rapidly increased; but if 
this ice remain, and form a continuous cover gradually growing in thickness, the cooling 
of the underlying water will be restrained considerably (see above p. 81). 
The writer has observed that a layer with an almost uniform temperature and salinity 
is formed in this manner down to 20 and 25 metres in the North Polar Basin; during 
the winter; but there, the vertical circulation cannot reach deeper, during one winter, 
owing to the rapid rise of salinity downwards. (Cf. Nansen, Oceanography of N. P. 
Basin). 
