NO. 4 TEMPERATURE VARIATIONS IN THE NORTH ATLANTIC 25 



Europe over that part of the Atlantic Ocean of which we are treat- 

 ing-. The low continental winter temperature is particularly notice- 

 able in the west, but the air is considerably warmed above the 

 Gulf Stream and hence it also takes a higher temperature over the 

 middle part of the ocean. This higher temperature sinks a little 

 toward the European coast, but not so much as the temperature 

 of the water-masses, which are cooled by outward radiation and by 

 the air. Hence the difiference between the temperature of the 

 water and the air diminishes nearly uniformly from west to east. 

 These results are distinctly seen in curves of figure lo, which relate 

 to the single 2° fields. There also is seen another peculiarity. The 

 relations of the cold " wedge " can easily be explained from this 

 general view. Here the water comes from the north and is rela- 

 tively very cold, while the air, on the other hand, comes in the great- 

 est part from the west. It is already considerably warmed by the 

 Gulf Stream water west of the cold " wedge ". The air tempera- 

 ture therefore does not show so marked a minimum as the water 

 temperature, and the consequence of this is that the difference 

 between the water and air temperatures at this place is relatively 

 small. 



In the easterly and middle parts of the ocean, the surface tempera- 

 ture shows a minimum in the middle of March, but in the western 

 part of the ocean the minimum comes toward the end of February. 

 The curves for the temperature of the water (see fig. 12 W) have 

 a comparatively regular course. A difference in some of the tem- 

 peratures of a tenth of a degree or perhaps even less would be 

 sufficient to make the curves completely regular. 



The air temperature shows in the eastern part a long extended 

 minimum from about February to the middle of March as shown 

 in figure 12 L. In the western part the air temperature rises rapidly 

 and gradually during the whole time and the minimum comes ap- 

 parently in January. In the middle part, there are certain irregu- 

 larities. Here there appear to be two equally low minima, one in 

 the middle of February and one in the middle of March, with 

 a well marked secondary maximum at about the first of March. In 

 agreement with this the difference between the temperatures of 

 the water and the air also shows irregularities (see fig. 12 W-L). 

 If our mean values really correspo'nd to the truth for this eleven- 

 year period, the cause of this irregularity is probably that the above 

 mentioned secondary depression in the air temperature is not 

 smoothed out because there is not a sufficientlv large number of 



