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



a westerly wind. There were great changes of wind in 

 this locality, a strongly developed cyclone passing by. It 

 seems to be quite obvious that the air observed by us 

 had come from northerly and not from southerly regions 

 a short time (only some hours) before. In this way the 

 relatively great difference (sea temperature niinus air 

 temperature) on June 28th is easily explained. On June 

 29th and the following two days the air came, according 

 to the charts, from southerly regions, giving everywhere 

 in the neighbourhood a higher temperature in the air than 

 in the sea surface. We had then rather suddenly passed 

 from tiie warm water of the "Gulf Stream" into the cold 

 Arctic waters. In the chart for June 30lh an observation 

 is introduced from a place about 100 naut. miles to 

 the S of the "Michael Sars" position, showing an air 

 temperature of 21'' C, sea temperature 18-4° C, and a 

 southerly wind of 4 Beaufort. Our observations were respec- 

 tively 14.0" C, 12-05= C, S, 5 Beaufort (0—12). Between the 

 two places there was a difference in air temperature of 

 T C and in surface temperature of 6-35 ' C. The air had 

 probably traversed the distance in about 7 hours (velocity 

 7-5 m.p.s.). 



For the third series of observations our A-curves 

 (p. 80'-) show a great difference of temperature between 

 water and air on July 14th. The synoptic chart for that 

 day shows that the wind in those regions came directly 

 from the area of the cold Labrador Current to the West. 

 Observations from a ship within the latter area gave a 

 higher air temperature (10° C) than sea temperature 

 (7-5" C), while the simultaneous "Michael Sars" observations 

 further east gave 13-6" and 16-2" C. Here, too, the transition 

 from cold to warm water takes place within a short 

 distance; and in such cases relatively great temperature- 

 differences between water and air must be expected. The 

 heating or cooling of the air takes place very quickly 

 though not instantaneously. 



By comparing the A-curves for sea temperature and 

 air temperature we find confirmation of the well-known 

 rule that the sea surface is generally warmer than the air. 

 Out of all the averages only 32 "/o show an air temperature 

 higher than the surface temperature, even though the 

 investigations were made in summer when the general 

 rule is often reversed. When we study the details exhibited 

 by the wind curves and vectors and the synoptic charts 

 we find, practically without exception, that the air temperature 

 increased relatively to the sea temperature when the wind 

 came from warmer areas, and decreased when the wind 

 came from colder ones. The increase is particularly marked 

 in cases when the horizontal temperature gradient was 

 comparatively strong, and especially when in addition the 

 wind velocity was great. 



We need only take a few cases as examples to show 

 the details. In the third series (A-curves p. 80*) we find: 



July 9th. Air temperature relatively much higher than 

 sea temperature. Wind from warmer regions. 



July 10th. Smaller difference between air temperature 

 and sea temperature, air warmer than sea, wind from 

 warmer regions, but very light. 



July nth. Wind quickly veering from southerly 

 directions (warm) to south-westerly, westerly and north- 

 westerly (cold) with increasing velocities, air temperature 

 quickly decreasing relatively to sea temperature. 



July 12th. In the afternoon wind veering to SW (warm), 

 air temperature increasing relatively to sea temperature. 



July 13th. In the forenoon increasing wind from 

 warmer regions, air warmer than sea. In the afternoon 

 comparatively fresh wind from colder regions, temperature 

 quickly decreasing and becoming much lower than sea 

 temperature. The cold and fresh wind also prevailed 

 on the following day. 



.luly 15th. Wind from cold areas but with decreasing 

 velocities, air colder than sea but with decreasing differences. 



In this way we may follow the variations from day 

 to day and practically everywhere find a clear verification 

 of the following general laws referred to the 24-lwur 

 means of temperature: 



T/ie air temperature is in most cases lower than the 

 sea temperature and always when the wind has quite 

 recently come from colder regions. The difference — sea 

 temperature minus air temperature — is then greater 

 with a rapid air-current than tvith a slow one. 



The temperature of the air approaches the sea tempera- 

 ture zuhen the wind has come from warmer regions and 

 becomes higher than that of the sea when the air moves 

 with comparative rapidity. 



We shall now consider the more individual and the 

 short-period variations exhibited by the B-curves. As stated 

 above, the surface temperature, even in the details, coincided 

 with the salinity and the daily period was often indistinct. 

 The detailed variations of the air temperature, we find, 

 show a more marked daily period, with a maximum in 

 the middle of the day or in the afternoon, and a minimum 

 in the night. But these daily variations were often greatly 

 modified by local variations of the surface temperature 

 and might sometimes be entirely blurred. Consequently 

 there is also quite a good correspondence between the 

 B-curves for sea temperature and air temperature, and 

 there is even a considerable resemblance between the 

 B-curves for salinity and air temperature. In many cases 

 the extremes appear a little earlier in the curves for sea 

 temperature than in those for air temperature, sometimes 

 indicating that the surface conditions were primary. The 

 daily period in the air temperature, with its greater 



