94 



BJ0RN HELLAND-HANSEN 



[REP.,eF THE "MICHAEL SARS" NORTH 



negative values of E are found at Stals. 4, 16, 25 B, 26, 

 30, 35, 50, 53, 69 and 92, wfiere the instability is due 

 to an increase of the salinity at the surface. At Stats. 

 6, 29, 31 and 83 the negative values of E between the 

 surface and 10 metres are due to a comparatively low 

 temperature at the surface. At Stats. 34, 44 and 46 the 

 temperature is lower and the salinity higher at the sur- 

 face than at 10 metres. Some of the stations where the 

 temperature at the surface is lower than at 10 metres were 

 worked during the night, while in some cases when the 

 stations were worked in the day-time the temperature of 

 the air was relatively low, which probably indicated cooling 

 of the surface water. Negative values of E are only rarely 

 met with at greater depths than 25 metres below the sur- 

 face. Nowhere at these depths is the negative value of 

 E greater than may be accounted for by errors of observ- 

 ations, especially as regards the salinity. In each of 

 these cases an error of 0-01 °/oo in the determination of 

 salinity may explain why the value of E has been re- 

 corded as negative. 



We can obtain a clear picture of the seasonal vari- 

 ations of stability in the upper water-layers by comparing 

 the stations in the eastern part of the North Atlantic off 

 the Bay of Biscay worked in April (for instance Stats. 

 4—9), with those worked in July (Stats. 90~93)- In the 

 open ocean the stability is practically always low in the 

 upper 10 metres on account of the mi-xing caused by 

 wind. At greater depths, and especially between 25 and 

 50 metres, the stability is very high in summer, indicating 

 a quasi-discontinuity. The upper boundary of the discon- 

 tinuity-layer may be regarded as the lower boundary for 

 the effective mixing by wave-motion. 



There is an intimate correspondence between the 

 vertical variations of stability and the value of the virtual 

 coefficient of temperature conductivity and diffusion. Low 

 values of E mean comparatively high values of v and 

 vice versa. 



As has been mentioned before, the vertical variations 

 of salinity have generally less effect upon density than 

 the vertical variations of temperature. The vertical and 

 horizontal variations of stability, therefore, show a marked 

 coincidence with the variations of temperature, and the 

 stability exhibits seasonal variations parallel to those of 

 temperature. In our examination of the distribution of 

 stability, therefore, we must consider the season when 

 the observations are made, when dealing with, say, the 

 upper 200 metres. 



The horizontal variations of stability are intimately 

 connected with the local variations of the currents, as they 

 are with the distribution of temperature. On account of 

 the local variations in velocity and vertical extension of 

 the currents the horizontal distribution of stability is rather 

 irregular, as will be seen from the charts on p. 98*. 



The chart for 75 — 100 metres shows great variations. 

 These are largely due to seasonal variations, a fact espe- 

 cially apparent in the area south of Ireland. Some high 

 values of the stability in the vicinity of the Straits of 

 Gibraltar are due to the transition from the relatively light 

 Atlantic water above to the heavy water from the Mediter- 

 ranean below. The very pronounced stability at some 

 places in the western part of the North Atlantic is con- 

 nected with the appearance of the Great Atlantic Current 

 or the great vertical variations below the arctic water 

 near New Foundland. 



Between 150 and 200 and between 300 and 400 metres 

 the stability is as a rule considerably less than between 

 75 and 100 metres. Between 500 and 600 metres and 

 between 700 and 800 metres the stability becomes, to 

 some extent, greater again, because the lower limits of 

 the great currents are, in many places, met with at 

 these depths. At 900 and WOO metres and deeper dozvn 

 the values of the stability are as a rule comparatively 

 low, decreasing with the depth until the stability ap- 

 proaches nought in the deep water. 



IX. DYNAMICS OF THE SEA. 



42. Theoretical Considerations. 



It is a well-known fact that the tangential stress of 

 the wind along the sea surface causes a displacement of 

 the surface water, and that a prevailing wind of appre- 

 ciable strength may give rise to ocean currents of even 

 great extension. It is also an established fact that diffe- 

 rences in the horizontal distribution of temperature and 

 salinity in the sea cause a disturbance of the hydrostatic 



equilibrium and provoke convection currents. The actual 

 ocean currents are the combined effect of wind and dif- 

 ferences of density. We shall not here discuss the spe- 

 cial kinds of currents caused by the tides. 



In a series of important papers Professor Walfrid 

 Ekman has discussed the problem of the wind currents. 

 The chief results of his investigations are that the wind 

 in open sea will create a current at the surface which is 

 deflected about 45 ° cum sole from the direction of the 



