46 



BJ0RN HELLAND-HANSEN 



[rep. OT the "MICHAEL SARS" NORTH 



and other differences between tlie stations we may only 

 call attention to the fact that in both cases the isotherms 

 are crowded below 25 metres, indicating a quasi discon- 

 tinuity. We have, here, in quick succession all three cases 

 mentioned above. Such sharp vertical transitions of tem- 

 perature (and density) are in summer often found over 

 wide areas just below the warm surface-layers. In the 

 open ocean with comparatively small vertical variations 



Sr 



Sz 



has a ma.ximum. At Stat. 92, Fig. 7, it corresponds to 

 [he depth marked "A"; at the other station it is situated 

 a little deeper. At levels above this depth the distribu- 

 tion of temperature corresponds to our case I. We have 



0, which tends to make >■ decrease downwards as 



in salinity, the stability has then a maximum where 



dz 



5^ r 



far as "A". Near this level approaches 0, and it is 



nought at the level itself (case 11). The surface waters 

 when stirred by wind are the seat of a vivid turbulent 

 motion, but it is a matter of course that the energy of 

 this turbulence decreases downwards. In fresh wind we 

 must e.xpect that )• decreases fairly rapidly with depth, 

 just above the level of "A", ^vith the result that a consider- 

 able rise of temperature takes place. Below this level 

 the conditions correspond to those in our example 111. 

 The stability decreases rapidly. At these depths the tur- 

 bulence due to wind directly is but small (if it occurs at 

 all). On the other hand, the vertical variations in the 

 velocity of the currents may be considerable, to judge 



Sv 

 from dynamical calculations. It is most likely that - is 



Sz 



positive, and therefore, — ~ negative. Then the tempera- 

 ?2r Sz 



ture below "A" sinks or rises according as 



c'z- 



SzSz 

 may probably in many 



But even if it rises, the value of — 



St 



cases be smaller than the corresponding value for the 

 water just above "A". In such cases the discontinuity 

 layer must be intensified, and at the same time its ver- 

 tical thickness often increases. 



We have hitherto dealt with the most common case 

 that the temperature decreases downwards. In some iii- 



IS r 



stances the temperature may increase with depth I 



\Sz 



The salinity must, then, increase with depth too 



z 

 S z 











as a state of instability cannot exist in any noteworthy 

 degree. Sometimes we find a minimum or a maximum 



of temperature at an interinediate depth, the vertical 

 gradient of temperature having different signs above and 

 below this depth. Such a distribution of temperature is 

 met with, for instance, at the Newfoundland Banks and, 

 at larger depths, in those parts of the eastern North At- 

 lantic where water from the Mediterranean is specially 

 proiTiinent. 



fig. 8. Scliematic illustr.ntion oi different Uinds of variation in the 

 vertical dis'ribution of temperature. 



Fig. 8 deiTionstrates the different cases to be encoun- 

 tered, with regard to the vertical variations of temperature. 

 We obtain the following scheme, referring to our initial 

 equation: 



Intermediate minima or maxima of temperature will 

 gradually vanisii if they are not renewed by horizontal 



