18 F. L. Ekman, 



same proportion as the mass increases, we shall have v^ "'•Tr* ^^^i 



as on the other hand the streem's vis viva must be constant, if we 

 neglect the qnantit}^ thereof, that by friction is converted into heat, we 



have also m . IT" = M . v" or v = W . -.—>= which is greater than IT. —, 



whence we see that the velocity of the stream decreases more slowly than 

 its mass increases. Hence the molecules in the following part of the 

 stream have a tendency to separate themselves from those in the pre- 

 ceeding, or in other words, the quantity of water at the locality we are 

 considering suffers a decrease. If the stream be a superficial one, the 

 level of the moving water should be thlis lowered, and, if an under- 

 current, the level of the water above should also be lowered. The hy- 

 drostatic pressure of the moving water-strata is then reduced, and in 

 consequence thereof an afflux of water from the surrounding strata should 

 take place in direction to the former. The afflux of Avater, required for 

 the replacement of molecules carried of by the stream, is then helped by 

 the fact, that the velocity of the thus increased current does not de- 

 crease in the proportion of the increase of the mass. 



In order in future to have a short name for the motions, which 

 a stream thus produces in the surrounding water, I have denominated 

 these motions reaction-streavis. It is easily seen that streams of this 

 kind, which in particular circumstances may be sharply defined and quite 

 powerful, must very frequently arise in the ocean. They must in the 

 first place accompany such streams, as originate in a disturbance of the 

 equilibrium, extending only to a certain depth, whether the disturbing- 

 cause be heat or rain, or both ignited. But also the superficial streams, 

 caused by the wind, and the under-currents, produced by the heavier 

 water's sinking and diffusing itself at the ocean's bottom, must give rise 

 to reaction-currents. The velocity of these streams ought chiefl}" to depend 

 on that of the original current and the area of the section, in which the 

 stream, arising from reaction, has to move. That area is often very 

 great and the reaction-stream consequently scarcely sensible as a sepa- 

 rate stream. The streams caused by a disturbance of equilibrium, ex- 

 tending down to the bottom, ought to be the least calculated to give 

 rise to reaction-streams; in fact, in these cases, the original motion 

 itself takes place in all strata, though with difterent velocity at ditferent 

 depths, and its velocity is moreover usually small. 



