74 FRIDTJOF NANSEN. M.-N. Kl. 
It is also clear that water like the bottom-water cannot arise by an 
intermixture of cold Polar with warmer Atlantic water, because the former, 
having temperatures below —1°C., has too low salinities of between 
34'3 and 347 /oo. The water with a temperature of about —1° C. near 
the under-side of the East Greenland Polar Current has a salinity of 
about 347 %0. Below this water-stratum both temperature and salinity 
rise downwards, whilst above it they both decrease. Water with a 
temperature of about —ı'4°C. and —ı'5°C. has salinities below 346 
and 34'5 %0. If Atlantic water be intermixed with this water to obtain a 
bottom-water with a salinity about 34'9 %00, so much of the former would 
have to be added, that the temperature could not possibly become very 
low, even assuming that the Atlantic water had been much cooled be- 
forehand. 
In the regions where the bottom-water arises, there must be expected 
very nearly the temperature and salinity of this water from the bottom 
upwards and towards the depth at which it is formed; it cannot pos- 
sibly be separated from the temperature minimum of the top-layer by 
any intermediate warmer layer. The cooling can only come from the 
sea surface, and the water must, as a rule, be gradually heated after 
it has sunk to greater depths. 
Before discussing the necessary conditions for the formation of the 
cold bottom-water, the conditions that may be required for the existence 
of an intermediate warmer water-stratum, will be considered. 
The Intermediate Warmer Water-Stratum underlying the East Greenland 
Polar Current. 
It is clear that, where the underlying warmer water is not protected 
by an overlying water-stratum with a much lower salinity, it cannot 
exist for any length of time, for the cooling during the winter would 
make the top layer heavier than the underlying warmer water. An 
active vertical circulation would be thus produced, with the result that 
the whole bulk of water to considerable depths, would be cooled down. 
Where, however, the underlying warm water is protected by an overlying 
lighter stratum of cold Polar water, it cannot be cooled down during 
the winter, because the vertical circulation, caused by radiation of heat 
is the ice chiefly met with by the sealers in spring, melts in regions where it may 
come into more direct contact with Atlantic water; but as the ice is originally formed 
inside the same sea-basin, the final direct effect upon the temperature of the sea, by 
its formation and melting will be #11. Indirectly it will, however, reduce the cooling of 
the sea, by protecting the upper layers against radiation during winter. 
