102 



BJ0RN HELLAND-HANSEN 



[REP. OF THE "MICHAEL SARS" NORTH 



rection, as applicable to the corresponding standard level 

 surfaces (columns 6 and 7) or isobaric surfaces (columns 

 8 and 9). For the standard values of dynamic metres the 

 density in situ (compression included) and the pressure 

 are computed, and for those of pressure the specific volume 

 in situ and the dynamic depth are found. The numeri- 

 cal calculations have been made by means of the tables 

 published by V. Bjerknes and J. W. Sandstrom [V. Bjerk- 

 NES, 1910]. Table IV contains anomalies of specific volume 

 and depth of isobaric surfaces, found by subtracting from 

 the actual values such values as would have been found 

 if the temperature had been 0° C. and the salinity 35-0 "/oo 

 everywhere in the sea. 



The charts on p. 99* show the dynamic depth of dif- 

 ferent isobaric surfaces according to the observations on 

 the "Michael Sars" Expedition and other expeditions in the 

 North Atlantic. Dynamic isobaths are drawn for every 5 

 dynamic cms. They show the direction of the gradient 

 currents at the different isobaric surfaces relatively to the 

 currents at the surface. The difference in velocity between 

 the sea surface and the isobaric surfaces is inversily pro- 

 portional to the distance between the isobaths. If there 

 were no currents at a surface of, for instance, 1400 deci- 

 bars, the dynamic chart for this surface would approxi- 

 mately illustrate the currents in the surface water when the 

 pure wind currents are not taken into account. Figs. 39 

 and 40 reproduce two dynamic charts for the eastern 

 North Atlantic, dynamic isobaths being drawn for every 

 2-5 dynamic centimetres. They show the topography of 

 the 200- and 400-decibar surfaces relatively to the surface 

 of 2000 decibars and not to sea-level [Helland-Hansen and 

 Nansen, 1926]. 



The dynamic charts illustrate quite well the agreement 

 between different depths in accordance with the theorem 

 of parallel solenoids. Ekman's law of the connection be- 

 tween gradient currents and variations in bottom-depth 

 seems to be verified in many places. 



The conditions in the Faeroe-Shetland Channel are 

 quite remarkable. It is a well-known fact that the distri- 

 bution of temperature and salinity is very irregular in the 

 Channel, and it has often been difficult to interpret it. 

 As previously mentioned a comparatively large number of 

 stations were worked in August 1910 from the "Michael 

 Sars" and the "Goldseeker" (cf. section 22). The horizontal 

 distribution of temperature and salinity at different depths 

 is illustrated on p. 95*, and the corresponding topography 

 of different isobaric surfaces is seen from the charts on p. 

 lOO*. Isobaths are drawn for every dynamic centimetre in 

 the charts for 200, 400 and 600 decibars and for every 

 dynamic decimetre in the charts for 800 and 1000 deci- 

 bars. These charts indicate the existence of a series of eddies 

 in the Faeroe-Shetland Channel. The Atlantic current 



comes from the west in the southern part of the Chan- 

 nel while another current comes from the north-west along 

 the northern Faeroe banks. Now it is quite interesting to 

 see the agreement between our dynamic charts of the 

 Channel and the picture inserted on p. lOO* illustrating 

 an experiment once made by Professor KrOmmel [1911, 

 Fig. 128]. Krummel's experiment was made by means of 

 a water-tank, where the water was set in motion by air- 

 blasts in the directions shown by the thick arrows in the 

 figure. The intention was to demonstrate experimentally 

 the currents in the central part of the North Atlantic be- 

 tween Africa and South America. The original illustration 

 by Krummel is reversed, so that the right hand side of 

 it comes to the left in our reproduction. The experiment 

 showed the formation of a double eddy between the pri- 

 mary currents as well as an eddy on each side of them. 

 The resemblance to the conditions in the Faeroe-Shetland 

 Channel is striking. Our dynamic charts show a double 

 eddy between the two main currents and another eddy 

 farther to the north. The possible existence of an eddy 

 in the south-eastern part of the Channel cannot be proved 

 for lack of observations. 



Having calculated the pressure at level surfaces or 

 the dynamic depth of isobaric surfaces for different sta- 

 tions we can find, by means of the equations (e) and (f) 

 in the preceding section, the vertical differences of the 

 velocity-components of gradient currents. To facilitate the 

 calculations the value of x = ^k w. sinrp is tabulated below 

 for different latitudes. 



10-^ 



^ 10- 



^'2 



ID. Sl/l (p 



The anomalies of depth of isobaric surfaces recorded 

 in Table IVB are expressed in 10 ■* dynamic metres as 

 a unit. In taking the differences of the values recorded 

 for two stations we obtain numbers which, when multiplied 

 by lO^-*, correspond to Dj - D., in equation (f). In order 

 to explain the further calculations we may take the following 

 example: In Table IVB we find a difference of 6175 



