AURORA BOREALIS. 197 



borealis upon the magnetic needle, does not constantly attend 

 that phenomenon, and that a very feeble aurora borealis has 

 sometimes a more marked influence upon the magnetic needle, 

 than a very brilliant aurora borealis. It likewise happens not 

 unfreqaently, that the latter produces no sensible effect upon 

 the magnetic needle. Upon the preceding facts, besides the 

 question first above stated, I would propose the following : 



Whether the seat of the aurora borealis, in our latitudes 

 before the year sixteen hundred and sixty-six, when the mag- 

 netic variation was easterly, was likewise in the eastern part of 

 the atmosphere. And whether the times, when the variation 

 is stationary, concur with the times of the disappearance of the 

 aurora borealis ; and those in which the variation of the needle 

 is most rapidly changed, concur with the times of the most fre- 

 quent appearance of the aurora borealis. 



The want of accurate observations, before sixteen hundred 

 and sixty-six, in both respects, renders the second question , - 

 insoluble. 



With regard to the first and the third question, the following Minutes of the 



table affords an outline of the observations which have been d . Istinct _P e " 



nous oi ap- 

 made upon the progress of the western declination of the pearance and 

 needle since the year sixteen hundred and sixty-six, and the disappearance 

 greater or less frequency of the appearance of the aurora borealis. 

 borealis, for periods of ten years each. 



' Times. 



From 1666 to 1680 increase 1 30 the aurora borealis 7 



1680 to 1689 5 20 13 



1689 to 1700 2 J 2 22 



1700 to 1710 2 38 59 



1710 to 1720 2 10 112 



1720 to 1724 stationary ■» 



- 1724 to 1730 increase 1 25 J — "— 531 



- 1730 to 1740 1 5 349 



• 1740 to 1750 1 45 84 



1750 to 1760 1 15 V nQ observations 



176O to 17/0 1 25 > 



- 1770 to 1780 « • 50 the aurora borealis 402 



. 1780 to 179O 1 1 (>9 



— — 179O to 1800 " O 26) disappearance 



— — 1800 to I8O9 diminish O 12J nearly total. 



After 



