ATMOSPHERIC ELECTRICITY IN HIGH LATITUDES. 



91 



tions of the radio-activity there for four weeks (October 17 to November 12), in 

 exactly the same way as I had done in Karasjok. 



The result was in entire agreement with ELSTER and GEITEL'S observations. The 

 mean for the month was only 58, which must be compared with the Karasjok winter 

 value of 126, the numbers for the different times of observing being 



TABLE XXV. Radio-activity in Hammerfest. 



But what is much more interesting and important is the great variation of the 

 radio-activity with the wind direction. When it is remembered that Hammerfest is 

 free to the open ocean on the north and west, while to the south lies the whole 

 stretch of Norway and Sweden, the following table tells its own story : 



TABLE XXVI. Radio-activity and Wind Direction in Hammerfest. 



It must be admitted that these results lend great support to ELSTER and GEITEL'S 

 hypothesis. 



OBSERVATIONS OF THE AURORA. 



It was not my intention on going north to make a particular study of the aurora, 

 but I naturally followed it with as much attention as possible. The necessity of 

 making my regular observations during the daytime, beginning at 7.30 A.M., made it 

 impossible to stay up to watch the aurora late into the night. Each evening I noted 

 down the chief variations in the aurora's form and brilliancy, but did not go into 

 minor details. I intend here to shortly record a few of the things which struck me, 

 and which are rather of a general than particular interest. 



During the year of my stay there were not many exceptionally fine auroras, and 

 coloured auroras were very rare. From the one or two I did see the colours appeared 

 to be of two distinct kinds (by colours in this connection I mean colours other than 

 the greenish-white light of the ordinary aurora). There is first the mass of coloured 



N 2 



