NO. 4 TEMPERATURE VARIATIONS IN THE NORTH ATLANTIC 8/ 



in chapter VII. Here we shall only say that a high temperature 

 at the mouth of the Channel points to a current in the water toward 

 the north or northeast, which may be set up by such a state of the 

 air pressure distribution as may cause low surface temperatures in 

 the middle of the Atlantic Ocean. There occurs therefore a dif- 

 ference between these middle and the most easterly regions (see 

 curves fig. 43). This difference is correlated with the masses of 

 water in the North Sea and the Baltic Sea. Accordingly one may 

 predict with the help of temperature observations in the Atlantic 

 Ocean in February whether in that year the water level in the North 

 and Baltic Seas will be on the whole higher or lower than common. 

 This leads to further consequences for the Baltic Sea where evi- 

 dently the rise or fall of the water within the basin plays a con- 

 siderable role in the entire circulation as well as in all that relates 

 to it. 



VARIATIONS IN THE AIR TEMPERATURE OVER THE ATLANTIC OCEAN 



On account of the difficulties attending accurate measurements of 

 the temperature of the air, we must expect that the air temperatures 

 amongst our observational material will contain many inaccuracies 

 and accidental errors. On this account, it must be supposed that 

 our temperature values for the air will not correspond very ac- 

 curately with the real conditions. Nevertheless, it appears that our 

 mean values even for the 2° fields go very well. We have not drawn 

 any curves for the air temperature for the single fields, but only 

 the curves for the difiference between the surface temperatures and 

 the air temperatures in each 2° field, as shown in figures 44 to 46. 

 If the air temperatures were altogether untrustwo'rthy these curves 

 would not show good agreement one with another. We find, how- 

 ever, a very good similarity between them, and we see that they, 

 like the curves for the surface temperature, show a gradual transi- 

 tion the further we go from the most easterly fields toward the west. 

 But westward of 44° to 46° west longitude they, like the surface 

 temperatures, begin to show greater irregularities and less correspon- 

 dence and this was indeed to be expected. Since these curves 

 show such a great agreement each to each, at least in the eastern 

 part of the ocean, it is clear that we may infer that our values for 

 the air temperature in the dififerent fields correspond very well 

 with the truth, and this conclusion will be even more justified for 

 the mean values of our 10° fields (see pis. 42 to 45, curves L). 



