ATMOSPHERIC ELECTRICITY IN HIGH LATITUDES. 93 



sky a long narrow aurora beam stretched, showing bright and dark patches owing to 

 the clouds. It looked exactly as if the aurora beam ran along under the clouds, 

 brightly illuminating the patches of cloud which it met. In reality the bright 

 patches were the openings and not the clouds. It took me a long time to make 

 quite certain of this, and it was only by at last seeing a star in the middle of a 

 bright patch that I could be quite certain. 



LEMSTROM strongly supported the idea that the aurora often penetrates down to 

 the earth's surface, and described how on one occasion the aurora line appeared in a 

 spectroscope pointed at a black cloth only one or two metres away. I was able to 

 repeat this observation on several occasions, and found that the line which then 

 appeared in the spectroscope was not due to an aurora discharge in the air between 

 the spectroscope and the black cloth, but was due to reflected light, which it was 

 impossible to prevent entering the spectroscope, as the whole landscape was lit up 

 with the monochromatic light of the aurora. 



All the time I observed the aurora I could not detect the slightest noise accom- 

 panying the discharge. 



I cannot close this account of my work in Lapland without expressing my deepest 

 thanks to each and every one of the small Norwegian colony in Karasjok in 

 particular to my host and hostess, Handlesmand arid Fru NIELSEN ; and to Lensmand 

 and Fru HEGGE all of whom did their very best to make my stay amongst them a 

 source of the greatest pleasure and real enjoyment. 



APPENDIX. 



Potential Gradient. The potential gradient was measured, as stated in the paper, 

 by means of a Benndorf self-registering electrometer. The electrometer is of the 

 quadrant type, the quadrants being kept at a constant voltage by means of small 

 cells, and the needle itself connected to the collector. To the bifilar suspension of 

 the needle a long aluminium arm is attached, which swings freely above a strip of 

 paper drawn along by clockwork. Every two minutes an electrical contact is made 

 which causes a bar to descend and to press the end of the aluminium arm down upon 

 the paper, where a dot is left showing the position of the arm and so the potential 

 gradient. The zero of the instrument was so arranged that on the normal side a 

 potential gradient of 800 volts/metre could be registered. On the negative side 

 only 100 volts/metre could be registered; but as all the days on which negative 

 potential gradient occurred were disturbed days, and the results on such days not 

 used, the range was quite great enough. 



The collector was arranged in the following way : My bed-sitting room in which 



