48, F. L. Î^KMAN, 



as one can see, acquire any increase from any- separate circumstances. 

 The stream, already from its commencement diffused over a tolerably 

 wide surface, presses itself in consequence of westward deviation towards 

 the American coast line, by friction against which it must lose a portion 

 of its power. As regards the ascending stream, the differences of level, 

 which drive it, are also originally but small, consisting of a change of the 

 water's volume through heat. Of itself that change of volume would 

 probably only have produced a very shallow and feeble stream. But we 

 have seen how that stream by the agency of the winds has had its 

 starting point removed to the Eastern coast of America, there acquired 

 an increased driving force in the difference of level, caused by the trade- 

 winds, and then, after having been further reinforced by the effects of 

 other circumstances, let out with collected forces to find its way through 

 the free ocean. I suppose that, when all these circumstances are taken 

 into consideration, the grounds here assigned as the origin of the streams 

 will not be found contrary to our experience of the coiirse of these sur- 

 face-currents, even though we cannot give a numerical proof of their 

 correctness. 



To resume now the thread of our explanation, which has been 

 for a moment broken off, I must remind the reader of what I stated 

 some time back, that every ocean-current must have its course in a cer- 

 tain measure dependent on the possibilities, that are presented of its 

 combining with other ocean-currents. Both the ascending warm surface- 

 stream and the descending cold must by mechanical reaction give rise 

 to an afflux of water to the region, where they pass. The descending 

 stream has its origin partially from the Icy ocean on the North of Asia ; 

 this stream unites with one descending between Spitzbergen and Green- 

 land, and lastly we may add the stream coming from Baffin's Bay and 

 the neighbouring tracts. Throughout the whole of this long line the 

 cold surface-current must evoke reaction-streams, with which the warm 

 ascending stream can combine, and in this I see a separate cause, why 

 the Gulf-stream after having long previously lost its proper intensity 

 and its marked character as a stream, nevertheless continues as a drift 

 stream to extend itself past the coasts of Scandinavia up even to Spitz- 

 bergen and into the Asiatic Icy Ocean. The Gulf-stream must in turn 

 bring about a reaction-stream in the opposite direction, with which the 

 cold descending surface-stream can partially unite. Whe have thus, in 

 addition to the above adduced principal cause for a descending cold 



