44 F. L. Ekman, 



soon increase in proportion as the beginning evaporation canses a simi- 

 lar motion. The difference of level, which evaporation and subsequent 

 rainfall produces in a unit of time, will be after a while, in consequence 

 of the increased temperature of the water, tirst equal to and then greater 

 than that arising from heat; and it would seem as if the warm surface- 

 current ought in consequence to diminish and ultimately cease. Without 

 entering into the question, whether at any stage of the current's deve- 

 lopement any such event take place, it is clear that when the tempera- 

 ture of the sea-water has reached the maximum, which is determined by 

 the limited accession of heat and the constantly increased loss of heat, 

 the evaporation and rainfall will generally spoken be constant, and thus 

 also the change of level, which they produce at a certain place in a 

 unit of time, will be a constant quantity. But in consequence of this, 

 the difference of level, caused bj' the above named changes of level, must 

 after a time obtain its maximum, on which the afflux to the evaporating 

 region must become constant. The sinki)ig of the surface through eva- 

 poration is then always compensated by afflux, and the level in the eva- 

 porating locality is thus maintained constant. If we now even suppose, 

 that the surface had assumed a slightly hollowed form, the hollow 

 would soon be tilled up by the effect of the heat. The arriving water 

 would remain there, its saltness and weight would be more and more 

 increased b}^ evaporation, and soon a vertical circulation would arise; 

 this circulation would indeed be limited because of the greater density, 

 which the lower water-strata have from their low temperature, but that 

 limit would .be continually lowered by the sinking molecules, which 

 would also carry with them into the deep the warmth they had acquired 

 at the surface. Thus the thickness of the warmed stratum must always 

 increase and consequently its level be raised, until it came to coincide 

 with that that of the surrounding water. The heated water would then 

 begin to flow out over this last, and the heat to produce its surface- 

 stream, just as if no hollow had been formed by evaporation. 



We must not overlook the circumstance, that in this case the le- 

 vels of the water-strata have not been, according to the common rule, 

 determined by the proportion of their specific gravities. The warmer 

 water beneath the especially evaporating region assumes a much lower 

 level, than it according to its specific weight ought to occupy, and thus 

 exercises a far less hydrostatic pressure than the surrounding colder 

 water. But just on this account this latter can flow in as if into a 

 hollow formed by evaporation, though no such hollow be there. An 



