118 Disturbance of the JVeedle during an Aurora. 



2. Aurora of July 10, 1833. 



(with a copperplate.) 



In the number of the Journal of the Franklin Institute for July last, 

 I gave an account of a disturbance in the direction of the horizontal 

 needle, during the occurrence of the Aurora Borealis, on the 17th 

 of May. On the occasion of the brilliant aurora of the 10th of July, 

 I had a remarkably favorable opportunity of observing a similar dis- 

 turbance, and as the subject is one which is considered to require 

 further observations for its elucidation, I send you an account of those 

 made on the evening referred to. 



The needles with which the observations were made, are the same 

 which were referred to in my note of observations made in May. 

 Their places had remained unchanged from that time ; the horizontal 

 needle out of doors, under a small wooden enclosure ; the horizontal 

 needle within doors on a table placed against a partition wall in my 

 study ; the dipping needle out of doors in a small observatory con- 

 structed entirely of wood, copper, and brass. The needles out of 

 doors have only the local attraction unavoidably incident to a city lo- 

 cation ; that within doors has, of course, an irregular and more con- 

 siderable attraction to affect it. The latter needle is exposed to very 

 slight variations of temperature. These particulars are probably suf- 

 ficient for the present purpose. 



1 first saw the aurora on the evening of the tenth of July, at three- 

 quarters past nine o'clock ; it appeared then as a low nebulous light, 

 resting upon dark clouds, and interspersed by them, but was not 

 sufficiently brilliant to make me entirely certain that it was an au- 

 rora : the test afforded by the magnetic needle shows that the phe- 

 nomenon began before this time, and the attention of my observer had 

 been so far drawn towards the sudden diminution of variation at nine 

 o'clock, that an inaccuracy in the observation was suspected, and the 

 needle recurred to, at five minutes past nine, to verify the result be- 

 fore obtained. At ten o'clock, the light in the north was more dis- 

 tinct, extending upwards, nearly 30° from the horizon. At this time 

 I began a systematic set of observations, which were continued until 

 one o'clock on the morning of the 11th. By going a short distance 

 from my dwelling to the east, I have a tolerably good view of the north 

 eastern portion of the horizon ; by going a greater distance to the 



