388 Alf Trolle. 



How then can we explain the relatively high temperature of the 

 bottom-water, which is considerably higher than the average tem- 

 perature of the sea-water outside? 



The heat can scarcely have been carried down to the lower layer 

 from above, as there is always a downflow from a glacier here, and 

 it is hardly probable, that the heat of summer, during the short time 

 the lake is not covered with ice, can penetrate down through the 

 whole layer. Even if we imagine the surface-layer to be considerably 

 warmed up in summer, however, this heat will nevertheless not be 

 able to penetrate down to the lower layer, as it cannot lead to convec- 

 tion currents, the uppermost salt layer on being heated only becoming 

 lighter. In such a case the heat must be propagated further down- 

 wards by direct conduction, but this is such a long process, that it 

 must be left quite out of consideration. 



There is still the possibility, that the water became heated when 

 it flowed into the lake. 



The most probable explanation of the origin of the warm and salt 

 bottom-layer in the lake is, in my opinion, that it flowed in during 

 exceptionally high water in the summer some 50 years ago at most. 

 The inflowing sea-water may be considered to have passed compara- 

 tively slowly over the plain between the sea and the lake and become 

 heated by the atmosphere on the way. The lake has in this way be- 

 come fdled with warm water of nearly homogenous salinity. Later 

 the downflow from the glacier, which may be considered to reach 

 down to a depth of ca. 45 ^/^ m. below the level of the lake, has affected 

 the water down to this depth, whilst the bottom-layer, which does 

 not take part in the circulation, has kept its temperature and sa- 

 linity almost unchanged. 



As the salinity in the surface layer is ca. 0.35 p. m. and this sa- 

 linity comes for the most part from diffusion with the lower layer, 

 the influence of which is кполуп, we may thus form some estimate of 

 the rapidity with which the circulation in the upper layer takes place. 

 If we assume, namely, that the conditions have been stationary, when 

 the investigations were undertaken, we may make the following cal- 

 culations. 



The average salinity in the uppermost 50 m. of Sælsøen was 0.35 

 p. m. Of this amount 0.07 p. m. seems to have other causes than 

 the diffusion, as a salinity of 0.07 was found in freshwater lakes (see 

 series LV and LVII), where there was no salt bottom-water. 



If we calculate that a salinity of 0.28 p. m. was caused by dif- 

 fusion, a cylinder of 50 m. height and 1 cm^ in diameter would con- 

 sequently contain ca. 1.5 gm. salt But we know (as already shown), 

 that 0.4 gm. salt is separated from the lower layer in 1 year by dif- 

 fusion; consequently, the upper layer is renewed in ca. 3% years. 



