586 



APPENDIX. 



[No. III. 



through several holes. Another small compass belonging to the transit instru- 

 ment, to the needle of which no card was attached, was placed in a recess at 

 the opposite side of the house, in the open air. The variation of the latter 

 was inspected by Mr. Franklin, and of the former, by me, at several hours of 

 the day and night. Both were very frequently affected by the Aurora, but 

 in different degrees ; and they recovered their former positions at different 

 times. Upon some occasions they were affected in the day, and upon others } 

 one compass alone was affected. Mr. Franklin remarked, that a certain 

 position of the Aurora was always attended by a variation of his compass in 

 the same direction, and that the motions of the needle were so immediate, as 

 to be often visible during his observation. Mine, on the contrary, however 

 various the shape and situation of the Aurora, in the course of two or three 

 hours, declined gradually to the eastward or westward, without ever retro- 

 grading ; and I could not perceive that the increase or decrease of the variation 

 depended upon any particular position of the Aurora, for it never was stationary 

 for one-fourth of the above-mentioned period. However, all the remarkable 

 aberrations of both instruments took place during the same nights, and were 

 found to be at their maximums on the following mornings, and they generally 

 recovered before 8h. p. m. As these maximums were the last result of the 

 impulses given by the Aurora, and their duration the longest, they were com- 

 pared together, and about half of them found to be in contrary directions from 

 the magnetic meridian, but not equal in quantity. Considering the Aurora as 

 magnetic effluvia, the aberrations of both needles should have been always in 

 the same direction with respect to the magnetic meridian ; because all bodies 

 which are not ferruginous, are equally conductors of magnetism, and will 

 transmit it in the same straight line — whereas, electricity conveys itself through 

 the channels which are the best conductors, and may act from different quarters 

 upon two needles differently situated : and its effect continues, till the charge 

 received by the needle is carried off by the pivot supporting it, which, by 

 experiment with an excited electric, appears to require not less than 10 or 

 12 hours. 



The common cork-ball electrometer not having on any occasion given signs 

 of a charge, I tried the following experiment, in order to attain further evidence 

 on the subject. A brass needle was attached to a compass card, and balanced 

 on a copper pivot in a wooden box. It was about four inches in radius, and a 



