ATMOSPHERIC ELECTRICITY IN HIGH LATITUDES. 97 



The instrument worked splendidly, and I had very little trouble with it. Its only 

 drawback was that it could not be left to itself for more than 20 minutes, for then 

 the weight required winding up again, and sometimes the voltage would vary if not 

 attended to. 



For my insulators I used amber enclosed in a metal case, so designed that air had 

 a long way to travel from the oiitside before it could reach the amber. The 

 insulators acted very well even in rain and fog. It was seldom that I had any 

 difficulty with them, and I was never compelled to give up an observation on account 

 of insulation troubles. 



Meteorological Measurements. Karasjok is a second-order station of the Norwegian 

 Meteorological Service, and 1 was granted full use of the observations made there. 

 I depended solely on these observations for the height of the barometer. For 

 temperature and relative humidity I used an instrument (the polymeter) made by 

 LAMBKKUHT, Gottingen. The hygrometer consists of a bundle of human hairs 

 mounted in such a way that a pointer is made to move over a very open scale 

 showing directly the relative humidity. About once a month the zero was tested by 

 painting the hairs with water in the way usual with such instruments. Treated 

 in this way the instrument proved quite reliable. A mercury thermometer was 

 attached to the metal frame of the hygrometer. The instrument was hung outside 

 the window of a porch, the door of which always stood open. In this way the 

 thermometer was not influenced by the radiation from a warm room, and. as the 

 window looked north, the sun did not shine on it during the day time. The metal 

 back of the instrument prevented the thermometer reading too low on a clear evening. 

 I had no instrument for measuring the wind strength and had to estimate it as well 

 as possible from the "feel." As the wind strength is only used qualitatively, the 

 absolute values are of little importance. 



VOL. ccv. A. <> 



