ATLANT. DEEP-SEA EXPED. 1910. VOL. i] PHYSICAL OCEANOGRAPHY AND METEOROLOGY 



13 



to be formed in the warm air ascending the inchned 

 polar front surface, either during an onset of warm air or 

 of retreating cold air, or during a cold onset forcing the 

 warm air to escape upwards. 



Clouds formed in this way cannot be expected to 

 stand in close correlation with the momentary meteorological 

 conditions deep below them at the sea surface. They are 

 the product of a more complex interaction of air masses 

 of different densities brought together from widely distant 

 regions. The clouds lower down have a more direct 

 connection with the conditions at the sea-surface. 



The water-vapour ab.sorbed from the sea is transported 

 upwards into the atmosphere when the latter has become 

 unstable by heating from below. The longer and more 

 intense this heating has been, the greater is the amount 

 of moisture in the upper air, and the lower and denser 

 the clouds formed. In the region now discussed the air 

 to a great extent comes from the American (Continent or 

 from between Labrador and Greenland being at first cold 

 or cooled, then heated by the "Gulf Stream" water and 

 afterwards making its way northwards again as an air-current 

 which is relatively warm in its lowest part. A detailed 

 survey of the wind-vectors and curves, and of the 

 synoptic charts, shows that the maxima of cloudiness 

 especially appeared in southerly winds which transported 

 air that had first passed the cold parts of the sea and then 

 the warm areas before returning northwards (see e. g. 

 July 13th, 21st and 24th). 



With northerly winds the air would mostly come 

 directly from the cold regions without having been heated 

 for any length of time. We have previously mentioned 

 that such air must make the observed temperature and 

 humidity sink. At the same time the cloudiness will 

 become relatively small if the cold air dominates up to 

 high altitudes, because the upper air is not sufficiently 

 moist for the formation of thick clouds. As time goes 

 on the cloudiness will probably increase, owing to the 

 instability. This cloudiness, however, which is due to 

 local ascending "currents'' through unstable air, gives the 

 typical broken sky with Cu and CuNb and seldom com- 

 pletely overcast heavens for any length of time. 



On July 18th the wind was north-westerly in the 

 afternoon, nevertheless the air temperature rose and the 

 humidity increased to such an extent that even a light 

 fog was formed and low clouds covered the sky. The 

 charts for July 18th and 19th show, however, that the 

 air had originally come from W or SW and only made 

 a short visit to the north, keeping a relatively high 

 temperature and probably being very humid for some 

 distance upwards. The fog had probably been formed 

 further to the north. 



In our first series of observations, to the south of the 



Azores (p. 76*), the wind came mostly from between north 

 and east, and the air was essentially in a state of instability. 

 We were within the precincts of the North Ea.st Trade 

 winds with their relatively dry air, generally with no great 

 extension of clouds (the scale for cloudiness is larger in 

 this series than in the others). The A-curve for cloudiness 

 shows variations which seem to be rather irregular without 

 any definite connection with the variations either of 

 humidity or of wind. 



During the fourth series of observations (p. 82*) the 

 sky was overcast nearly all the time except in the neigh- 

 bourhood of Scotland and Norway. The upper air was 

 probably very moist, having come more or less directly 

 from the Atlantic as "old sea-air". 



The divergencies from the daily means demonstrated 

 liy the B-curves correspond in most cases to the variations 

 in wind velocity. It should be remembered that the wind 

 direction, and not only the velocity, may vary at the level 

 of the clouds as compared with the stratum at the sea 

 surface. Full agreement, therefore, cannot be expected. 

 An increase of wind, however, especially along our route 

 from America to Europe, meant in the majority of cases 

 the proximity of depressions, and therefore great cloudiness. 

 An increase in wind velocity may often cause an alteration 

 in the stale of equilibrium. We accordingly find, that 

 the variations in wind-velocity and cloudiness exhibited 

 by the B-curves generally went nearly parallel when 

 there were "northerly" winds (e. g. June 6th to 12th), 

 while with "southerly" winds (e. g. June 22nd to 25th) 

 the variations were liable to be reversed. 



The variations were sometimes apparent earlier in the 

 cloudiness than in the wind. This is partly due to the 

 fact that the wind-velocity may be much greater at the 

 cloud level than at the surface, and partly to the fact that 

 the estimation of the cloudiness embraces wider areas. 



11. The interaction of the Ocean and the 

 Atmosphere. 



The observations dealt with here were made in June 

 —August 1910 only, and were locally restricted to some 

 few lines. The results set forth above are therefore, 

 strictly speaking, very limited in regard to time and 

 locality. They seem, however, to give a fairly systematic 

 and clear view of processes taking place between the sea 

 and the atmosphere, so they probably possess a more 

 general application, all the more so because they coincide 

 in many respects with results already known. We proceed 

 to give a general review of these processes, along the 

 lines of the detailed discussion in the previous sections. 



