538 



HIKKEI.AND. THE NORWEGIAN' AURORA POLARIS EXPEDITION, 19021903. 



For Dyrafjord it is only in the horizontal force that the two types are well separated; for the 

 other two magnetic elements the storminess is much smaller, and each group of storminess is divided be- 

 tween the positive and negative storminess, so that the TV and P groups get mixed up. In consequence the 

 time of maximum has been determined from the horizontal force only. 



For AxelDen S'^ has a large value during a long interval, and has in fact two maxima. The 

 number given is the mean of the time for the two maxima. 



The calm period is long and well defined for Matotchkin-Schar and Kaafjord, for Dyrafjord shorter 

 and not so quiet, while for Axeloen it has more the character of a minimum than that of a quiet 

 interval. 



We see from the table that in spite of the rather large differences in longitude of the stations, the 

 TV storminess for the various stations has its maximum very nearly at the same local hour, about half past 

 twelve at night. Also the P storminess has its maximum at the same time of day about six o'clock in the 

 evening, and the calm period is always found in the forenoon. Except at Axeloen, the middle of the 

 calm interval is at half past ten in the morning. 



Tin's result shows that the storminess near the auroral zone follows the diurnal motion of the .;//;/. 



PROPERTIES OF THE "AVERAGE POLAR STORM" 



103. To obtain comparable numbers we must consider the storminess for a period common to all 

 four stations. The storminess for the three months December, January and February, is given in table 

 XCV and graphically represented in fig. 196. As we have already mentioned S J can he considered as 

 a vector representing the perturbing force of the average polar storm, and in the usual way it can be 

 represented by a vector diagram of some sort. 



TABLE XCV. 

 Mean Storminess for the Period December 2 March i. 



