38 F. L. Ekman, 



Even if hydrodynamical science were sufficiently developed to 

 treat this complicated problem, it would in the outset be stopped by 

 want of a competent body of observations. Although therefore there is 

 but slender prospect of science being soon in a condition, fully to com- 

 prehend all the causes of the vast current-motions in the open ocean, 

 we may nevertheless seek to approximate to that interesting subject of 

 human research, in order if possible at least to discover, what circum- 

 stances must in the first place be investigated in order that the ques- 

 tion may receive its proper solution. 



If we set out from the principle, that with exception of the 

 streams, caused b}^ the winds or by other streams, the ocean-currents 

 are originally produced b}^ differences of level, and if it be further ad- 

 mitted, that it is the unequal distribution of heat, evaporation and rain- 

 fall, which chiefly gives rise to these differences, it would seem desirable 

 in the first place to inquire, how much the level of the sea in a certain 

 region may be altered in a unit of time by each of these causes, sup- 

 posing no motion for the moment to take place. 



As regards the first, viz: tlie chcmi/es of level cauf^ed by heat, there 

 is only one way of determining them, namely by calculating the abso- 

 lute quantity of heat, which the sea within a given region acquires or 

 loses in a given time; from the increase or decrease of warmth, thus 

 obtained as a result, the change of level must be deduced. As regards 

 the })ossibility of such a calculation, we may in the present state of 

 science assume, that no sensible amount of warmth can now be commu- 

 nicated to the ocean through the crust of the earth. But the amount of 

 heat, derived from the sun, is not only dependent on the sun's position 

 and the loss of heat, caused by the atmosphere's absorption and by the 

 reflection of the heat-rays from the surface of the sea, but also on the 

 more or less regular nature of these. If the surface is smooth, the sun's 

 heat penetrates deeper; if agitated, a greater portion is retained in the 

 superficial stratum, and heat is more employed in forming vapour than 

 in the other case. The mean force of the wind has therefore here in- 

 fluence, so much the more, as it is also for other reasons favourable to 

 evaporation. The heat lost by evaporation and also by radiation, which 

 quantities are very considerable and the greater the warmer the water 

 previously was, must be deducted from the sunheat received. In such 

 a calculation it would be a mistake to neglect the atmospheric heat. 

 The heat, which the sunbeams leave in the atmosphere, and that, which 



