AURORA BOREALIS. 367 



the earth's surface, there is comparatively little 

 lightning where the temperature is uniform, as 

 over the tropical ocean, and where the wind blows 

 uniformly in the same direction. 



It is highly probable, that the electric fluid 

 is nearly all given out, before the upper current 

 reaches the centre of maximum cold ; which ex- 

 plains why the aurora is less brilliant at Mel- 

 ville Island, than at Bear Lake, the Shetlands, 

 Orkneys, and many other places in lower lati- 

 tudes. It also explains why in the coldest 

 regions, as at Port Bowen, Winter Island, &c. 

 the magnetic needle is less often disturbed by it 

 than in lower latitudes ; and why in the former 

 places, it was almost always seen by the British 

 navigators southward of them.* 



Another important fact, that shows the in- 

 timate connexion between the aurorae and the 

 change of temperature which the equatorial air 

 undergoes, is, that the northern lights are more 

 numerous and brilliant during the coldness of 

 winter, when the upper currents from the tro- 

 pics, in their passage to the Polar regions are 

 greatly condensed ; and that they are then 



* It is stated by Captain Scoresby, that the aurora was dis- 

 covered more frequently south than north of Spitzbergen and 

 by Sir Edward Parry, that it was generally seen south of Mel- 

 ville Island from which we may conclude, that the upper cur- 

 rent of the atmosphere has acquired nearly its maximum density, 

 and deposited nearly all its vapour, before arriving at the limits 

 of greatest cold. 



