746 THE AURORA BOREALIS. 



hundred metres from our abode, and in one corner of it the 

 electrometer was placed on a pile of stones, and in the opposite 

 corner the tube and scale. The collecting apparatus could be 

 raised through an opening in the roof. At certain periods a pole 

 25 feet long was employed, which was supported by stays, and the 

 collector could be raised the whole height of it. When the con- 

 duction of the air made it impossible to employ this method, the 

 pole was taken down, and the apparatus was only raised by hand, 

 the ball reaching about three feet above the roof. There were no 

 salient points which could have any disturbing influence. The 

 place of observation was a rather low island in Mussel Bay not 

 far from the 80th degree of latitude. There were no mountains 

 around which were at an elevation of more than three degrees 

 above the horizon. 



The collecting apparatus consisted sometimes simply of a hollow 

 knob or ball, of about three inches in diameter, mounted on a rod 

 of ebonite five feet long, at others of a lamp also, which could be 

 screwed on to the top of the ebonite rod by means of a cross piece 

 of ebonite one foot long. The lamp consisted of a cup of metal, 

 with its edges pierced, in the middle of which was a reservoir 

 with a rim of metal to hold the spirit. From the middle of the 

 cup of metal a metal rod a foot and a half long projected vertically 

 downwards, which could be screwed to the ball of the collector. 



The lamp was lighted and raised, and, when the spirit was burnt, 

 the lamp was lowered and broxight to the electrometer. When the 

 ball alone was employed, it was placed in communication with the 

 Earth by a metallic thread. However, in measuring the electricity 

 of the Earth the lamp with no spirit was often used, because of 

 the inconvenience of taking it off the ball. When there was only 

 a weak charge of electricity, constant contact was made between 

 the collector and the electrometer in order to take the measure- 

 ments rapidly. In this case a cotton-covered copper wire covered 

 with paraffin connected the electrometer to the collector. The 

 thread was stretched along inside the roof attached to some sup- 

 ports by a thread, with as good insulation as possible, and so as 

 not to disturb the electrometer when the collector was rapidly 

 raised or lowered. Several experiments were made to test this 

 arrangement. 



The observations were made by M. Wijkander and Lieutenant 

 Polander. 



In the autumn of 1872 experiments were made to discover 

 the electrical state of the air and of the Earth's surface, and there 

 were always feeble traces of positive electricity in the air and 

 negative electricity in the ground. The free air appeared to be a 

 very bad insulator under the circumstances of the experiment. 

 The deviation produced by the battery ceased at once when it was 

 removed. This was not from any defect in the electrometer, for 

 the insulation was always satisfactory in a warm room. The 

 experiments below show that the insulation was also good in free 

 air when it was very cold. 



