242 Dr Richardson on the Aurora Borealis. 



4ith, That the nearest end of the needle is drawn towards the 

 point from whence the motion of the aurora proceeds, and that 

 its deflections are greatest when the motion is most rapid, — the 

 effect being the same whether the motion flows along a low arch 

 or one that crosses the zenith. 



5th, That a low state of temperature seems favourable for the 

 production of brilliant and active corruscations, it being seldom 

 that, we witnessed any that were much agitated, or that the pris- 

 matic tints were very apparent when the temperature was above 

 zero. 



6thf That the corruscations were less frequently visible be- 

 tween the first quarter day, and the full moon, than in any other 

 period of the lumination, and that they were most numerous be- 

 tween the third quarter and the new moon *. 



*7th. That the appearance of the aurora was registered at 

 Bear Lake in 1825-26, 343 times, without any sound having 

 been heard to attend its motions. 



Sth, The height of the aurora was not determined by actual 

 observation, but its having been seen on several occasions to il- 

 luminate the under surface of some dense clouds, is conclusive 

 that its elevation could not have been very great. When Dr 

 Richardson and Mr Kendall made their excursion on Bear 

 Lake, in the spring of 1826, the former saw the aurora very 

 brilliant and active, displaying prismatic colours in a cloudless 

 sky (on 23d April) ; while Mr Kendall, who was watching at 

 the time, by agreement, for its appearance, did not see any 

 corruscation, though he was only twenty miles distant from Dr 

 Richardson. 



9^/*, The gold-leaf electrometer, which was kept in the obser- 

 vatory, was never affected by the appearance of the aurora. 



10th, On four occasions, the corruscations of the aurora were 

 seen very distinctly before the day-light had disappeared, and 

 we often perceived the clouds in the day-time disposed in streams 

 and arches, such as the aurora assumes. 



• The proportion of corruscations seen at th^se periods, from the month 

 of October 1825 to April 1826, was 38 to 125. The moonlight being strong 

 between the first quarter and the full moon at those hours when we more 

 particularly watched for the Aurora, may, perhaps, account for our not hav- 

 ing seen its corruscations so often during this part of the lunation. 



