AURORA BOREALIS. 449 



their ends ; but even here the rays could not have been quite 

 parallel unless their extremities were infinitely high. 



In Kotzebue Sound the Aurora was seldom visible before 

 ten o'clock at night, or after two o'clock in the morning. We 

 never heard any noise, nor detected any disturbance of the 

 magnetic needle : but here I mvist observe that Kater's com- 

 pass was the only instrument employed for this purpose, and 

 then on board the ship only, the exposed situation in which 

 we were anchored not admitting of any establishment on shore, 

 either for this purpose or for astronomical observations. 



Mr. Collie, the surgeon of the Blossom, whose attention to 

 meteorological phenomena was unwearied, has given an in- 

 genious hypothesis on the subject of the Aurora. After ex- 

 pressing his opinion that this meteor occurs in the region of 

 the thin and higher clouds of the earth's atmosphere, he ob- 

 serves, that " it is highly probably that the two strata of at- 

 mospheric fluid proceeding in opposite directions — the one 

 from the equinoctial toward the polar regions, and the other 

 in the reverse direction — are charged with opposite electri- 

 cities, and that they are in different degrees of temperature 

 and of humidity : the upper stratum, flowing from the equator 

 toward the poles, being of a higher temperature and more 

 charged with vapour than the lower, proceeding from the pole 

 to the equator. They might thus be charged with opposite 

 electricities, which would communicate and neutralize each 

 other. 



" The opposite temperatures would be reduced to their 

 mean, and under certain circumstances these changes might be 

 attended with the evolution of electrical light, and with the 

 condensation of transparent vapour into thin clouds (stratus- 

 cirrus, or cirro-stratus). As the watery particles of these 

 clouds form, a certain degree of electric conductibility would 

 be established, by which this subtle fluid might be propagated 

 to short distances ; but the greater dryness of the air, both 

 above and below this region of thin mist, would oppose an un- 

 conducting barrier to its escape. As soon as one thin cloud, 

 a thin stripe of cirrus, or fleecy portion of cirro-stratus or cirro- 

 cumulus, became charged with electricity, it would occasion,. 



