XO. 4 TEMPERATURE VARIATIONS IN THE NORTH ATLANTIC 33 



the same years when the drift ice and icebergs of the Antarctic 

 Ocean have wide distribution, and icebergs even drift as far north 

 as the latitude of the Cape of Good Hope. 



In later continuations of his work (1909, 1910, 1914) Hilde- 

 brandsson strongly maintains concerning the variations of air pres- 

 sure, temperature and precipitation, particularly in winter, that 

 there is a well-marked opposing relatio'n between those action centers 

 where there is an air pressure minimum and those where there 

 is an air pressure maximum. Examples of such opposing centers 

 are Iceland and the Azores, Alaska and Siberia, Tierra del Fuego 

 and Tahiti. On the other hand a well-marked correspondence exists 

 between the action centers of the same kind, as for example between 

 the variations of the two air pressure maxima of the Azores and 

 Siberia. 



Hildebrandsson thinks that the principal cause of these varia- 

 tions which occur in opposing senses in the action centers of opposite 

 kind, as for example air pressure minima and air pressure maxima, 

 is not to be sought in the very regular tropical climates, and not 

 even in the temperate zones. No such far-reaching phenomena of 

 great variations from year to year are to be found in these, suffi- 

 cient to be the cause of such considerable difTerences as exist be- 

 tween the different types. The cause must therefore, he thinks, be 

 found in the polar oceans, in the condition of the polar ice. Dur- 

 ing a warm summer in the northern regions, according to his view, 

 the ice is broken up and partially melted, and consequently in the 

 next winter, in February and March, great masses of ice are com- 

 mon near Iceland. This reduces the temperature of the ocean 

 between Iceland, Scotland and Norway, which again in its turn 

 causes an increase of the air pressure in the same ocean region. 

 This, again, influences not only the temperature in the parts of the 

 earth which are directly afifected by these action centers either in 

 the same or opposite direction, but also action centers on the earth 

 may be influenced at great distances. 



How Hildebrandsson would explain that a warmer summer 

 widens the distribution of the ice and on this account brings a 

 greater quantity of ice to Iceland in the next following winter, he 

 does not fully state. He does not appear to have observed that 

 the variations in the distribution and the drift of the polar ice 

 are influenced to a great extent by the variations in the prevailing 

 winds, that is to say, in the distribution of air pressure, whereas 

 the temperature has, directly, very little to do with it. The con- 



