NO. 4 TEMPERATURE VARIATIONS IN THE NORTH ATLANTIC I4I 



sun spot curve, particularly curve II, but in this period also are 

 included three or four shorter periods. 



The direct agreement between the temperature curves for the 

 eastern part of the North Atlantic and the sun spot curves is shown 

 in figure 59. Curves C were obtained by taking consecutive thirty- 

 three month means of the temperature values. The short periods 

 are thus to a great extent eliminated. These curves for Liepe's 

 stations I and IV (L-I and L-IV) show a distinct agreement with 

 the sun spot curve 83 which is the lowest of the figure, only it is 

 to be observed that this curve is inverted. The two temperature 

 curves have minima at sun spot minima and high temperatures 

 at sun spot maximum in the years 1893 and 1894. In addition the 

 corresponding curves B for L-I and L-IV show a strongly marked 

 division of the sun spot period in three or four shorter periods 

 similar to those of the prominences. 



It seems clear that in the surface temperature of the Atlantic 

 Ocean several periods occur for which one of about three years is 

 particularly notable and also a longer period which corresponds 

 with the sun spot period. The temperatures vary in these periods 

 in 'the middle part of the ocean oppositely to the sun spot num- 

 bers, while in the eastern parts they increase more or less directly. 

 As it has repeatedly been remarked, our observational series is 

 all too short in order to give certain conclusions with regard to these 

 matters. Considerably longer are the observational series of Peter- 

 sen and Liepe, but they are not sufficient, and still longer series of 

 ocean temperatures are unfortunately not available. In the lack 

 of sufficiently extensive observational material in the ocean and 

 because we have found great agreement in general between the con- 

 dition of the ocean and of the air, we have undertaken to investigate 

 the various meteorological elements which have the advantage of 

 having been published for a long period of years. 



We will first compare the variations in the surface temperature of 

 the Atlantic Ocean found by us with the variations of the air tem- 

 perature in different regions of the earth for the period of years 

 1898 to 1910. Such a comparison is given in figures 60 and 61. 

 Curves I to IV in these two figures show the variations in the air 

 temperature in diflferent regions according to Mielke's tables (19 13) 

 given in Koppen's investigation of 1914. The other curves show 

 the variations in the surface temperature in the different parts of 

 the Atlantic Ocean partly for the whole year, partly only for the 

 month of February. 



