8 



If it be remembered that the limit for the accuracy of the 

 method of investigation may be set at 1 in the fourth decimal, it 

 will be seen that there has been an almost homogeneous layer, in 

 respect to density, from the surface to the bottom. Under such 

 circumstances the power of the wind, as well as that of cold above, 

 must play a great part. Cooling from above would naturally upset 

 the balance, as the upper particles would become heavier and con- 

 sequently sink down through the lower and light layers. According 

 to MOHN, the cooling of sea water from 6 to 5 would thus in- 

 crease the specific gravity by 1.2, from 6 to 4 by 2.3, and from 

 6 to 3 by 3.3 in the fourth decimal. A cooling of the surface 

 layers by 3 would thus, from the above example, cause them to 

 sink to a depth of 200 in. 



Even the distribution of temperature shews, moreover, that a 

 vertical circulation must take place. Warmth spreads very slowly 

 through water when its diffusion is not encouraged by circulating 

 currents. The vertical circulation may be checked when the salinity 

 at the surface is much less than at the bottom, this causing a 

 great difference in its specific gravity. Thus, in the upper part of 

 the Christiania Fjord, in September 1897, the temperature, at a 

 depth of 60 metres, was but 6, whilst the surface temperature was 

 13 to 14. On the west coast where the continental warmth is 

 less, the temperature at a similar depth was, in September, about 

 12 (see PI. 5, Fig. 3). There the vertical circulation must, there- 

 fore, have been very considerable, much greater indeed than in the 

 Christiania Fjord. 



MURRAY [88] reports some interesting examples of the power 

 of the wind. A Scotch freshwater loch was examined before and 

 after a gale. Whilst the isotherms prior to the storm were all 

 parallel with the surface (a proof of the quiet state of the lake) it 

 was observed, shortly after the gale, that a great portion of the 

 warm water had been blown to one end of the lake. 



As will be seen from the following outline, all the isotherms 

 after the gale run towards the bottom in the direction the wind. 



Fig. 2. Section of a Scotch lake after a gale, in the direction indicated by 



the arrow. From a copy of MURRAY'* Treatise in the Scotch Geographical 



Magazine 1888 by VON ROHX. 



MURRAY has also shewn, from other observations, that gales 

 can even force the warm water beneath the cold, and in such a 



case a mixture takes place. The Norwegian coastal waters are, 

 during winter, just as homogenious in regard to specific gravity 

 as those of a lake where, in summer, various degrees of tempera- 

 ture are met with at various depths; and when it is reim i inliriv<l 

 that the sea will break off the west coast of Norway in 20 fathoms 

 of water, it will be perceived that also there a great mixture must 

 take place. 



If, however, it proves to be the case that the formation of 

 the various layers of water takes place in the manner previously 

 indicated from mere local causes, great fluctuations can, nevertheless, 

 according to one opinion, take place in the bodies of water. The 

 Scandinavian coastal waters, from the Scaw to Finmarken, must be 

 regarded as one hydrographical system which certainly is, mainly, 

 in motion northwards along the shores of Norway, but which by 

 the force of wind may be turned in various directions. It is espe- 

 cially interesting to know from the Swedish investigations, that great 

 changes can take place in the Swedish fjords, as at one moment a 

 fjord may be filled with fresh Baltic Water, while at another the 

 salter water, the so called "Bank Water", of a salinity of 32 to 

 30 %o, can force its way in. It thus proves, what the Swedish 

 Scientists have indicated, that where great differences exist in the 

 specific gravity of the layers, they remain like two very different 

 seas who do not associate the one with the other. The summer 

 conditions are very instructive. In the Christiania Fjord, for in- 

 stance, the cold water of the previous winter will be found along 

 the bottom, while the warm fresher layers appear on the surface. 



Even if the amalgamations thus play a great part, it is, how- 

 ever, seen that they do not occur rapidly, and that layers, once 

 formed, can often exist for long periods of the year, without being 

 much affected by the influence of the super-jacent waters. In the 

 course of a long time, months for instance, the great differences 

 may, successively, be overcome, and two layers become amalgamated 

 to one. 



We have thus come to the conviction that the changes in the 

 Norwegian coastal waters are dependent, mainly, on the Gulf Stream 

 on the one hand, and, on the other, on local causes. The Polar 

 current has no direct influence; but our investigations prove that 

 it is of a thickness, which, indirectly, must have the greatest in- 

 fluence on the entire current-system of the North Atlantic, and 

 the climate of Northern Europe. And its importance is still further 

 increased through the periodical variations in its thickness and 

 distribution. 



Whilst the Polar current withdraws during the Summer to 

 the north eastern coast of Iceland, and the Gulf Stream fills 

 the whole of the central portion of the Northern Ocean, the Polar 

 current increases in thickness during the winter, the Gulf Stream 

 retires, and its waters become partly blended with water from 

 the coast. 



NANSEN likewise reports [97] that, during the winter, the flow 

 of the surface waters of the Arctic Ocean was in the direction of 

 the Northern Ocean and the Atlantic, whilst the currents during the 

 Summer are varying, in part, the reverse of those of the winter. 



Still more interesting than the annual periodical fluctuations of 

 the Polar current, is the fact that, from our observations, it varies 

 in its development in different years. 



