38 



lies he has been unable to refer to any obvious cause, except when 

 they were accompanied by the appearance of the aurora borealis, 

 which evidently affected the needle on many occasions. He also 

 thinks that the vibrations of the needle became less rapid with a 

 moist atmosphere, and more so when it was very dry. Changes of 

 the wind and snow storms appeared also to be attended with fluc- 

 tuations in the intensity of the magnetism. He endeavoured to as- 

 certain whether there existed any decided and constant difference in 

 the directive force of each pole ; conceiving that, on the hypothesis 

 of a central magnetic force, the north pole of the magnet would, in 

 these northern latitudes, be acted upon with much greater energy 

 than the south pole. From his observing that the relative intensity 

 of the two poles is not always the same, he infers the probability of 

 the earth's magnetism being derived from the agency of electric 

 currents existing under its surface as well as above it, and that the 

 rapid fluctuations in its intensity are owing to meteorological changes. 



The author is led to conclude that the aurora borealis is an elec- 

 trical phenomenon, and that it usually moves during the night nearly 

 from north to south, and in an opposite direction during the day ; 

 that it is of the nature of positive electricity ; and that its elevation 

 above the earth is much greater than a thousand, and perhaps thou- 

 sands of miles. 



Printed by Richard Taylor, Red Lion Court, Fleet Street. 



