No. III.] LIEUTENANT HOOD ON THE AURORA. 587 



copper arch of 60° to that radius, was fixed at one end of the box, which was 



■ 



closed by a wooden slide, and paper pasted over every crevice to exclude the air 



• 



To give it the same advantages for conducting electricity as the compass boxes 

 (which are made of brass,) I introduced an iron wire, eight inches in length, 

 perpendicularly through the lid, in such a manner, that its lower extremity was 

 in a horizontal plane with the needle ; and a pane of glass at that end of the 

 lid, enabled me to see into the interior of the box. Having previously ascer- 



■ 



tained that it contained no magnetism, the instrument was placed, on the 2d 

 of May, on a covered shelf, at the outside of the house, in a position nearly 

 east and west ; the brass needle being 25' from the conductor, and a small 

 glass bubble adjusted on the box, in order to prevent its otherwise unper- 

 ceived movement. At 12h. p. m., I examined the needle, and found its position 

 unaltered. No Aurora was then visible, but one was afterwards seen by Mr- 

 Franklin ; and at 8h. a. m., May 3d, the needle and conductor were in contact. 

 I moved the needle 40' from the conductor, and it was similarly affected at 

 some period on the nights of May 3d, 5th, 6th, 9th, 10th, and 11th. The 

 thermometer, during this period, ranged in the day between + 26° and + 56 

 and in the night, between + 10° and + 33°. I did not see the Aurora, except 

 on the nights specified above ; and did not perceive any alteration in the 

 needle till the succeeding mornings. 



The night of the 1 2th furnished a more satisfactory proof of the agency of 

 the Aurora. At 1 Oh. p. m. , the needle was not affected, and no Aurora was 

 visible. At Oh. 30' a. m., May the 13th, several arches appeared across the 

 sky from N.W. to S.E., and the needle was attracted to the conductor from 

 the distance of 1°. The temperature of the air was + 12 . I now determined 

 to convert the instrument into a kind of electrometer, by insulating the needle 

 and conductor. The pivot which supported the former was fixed upon sealing 

 wax, and the part of the latter which passed through the lid was covered with 

 the same substance. 



Paper was pasted on the box as before, and it was re-placed at 2h. p. m. on 

 the 1 4th, the temperature of the air being 54°. A heavy gale of wind from 

 N.N. W. with snow, immediately followed, and the temperature of the air, at 

 midnight, was reduced to 19°. At 9h. a. m. May 15th, the needle was removed 

 30° from the conductor, and both were still charged, so that I could not bring 

 them together till the conductor was accidentally touched. I believe thi 



4 F 2 



