Aurora Borealis of September 3, 1839. 261 



instantly vanishing. This was often repeated. The southern part 

 of the heavens was occupied with streamers to very unusual ex- 

 tent. The arch bounding the streamers on the South gradually 

 descended, so that at lOh. its vertex was not more than 10^ above 

 the horizon. We were consequently led to infer that this occur- 

 rence extended very far to the south of us, which has been found 

 to be the fact. 



At 7h. 37m. T stationed within the house, a surveyor's compass, 

 so that the needle coincided with the N. and S. points. At 8h. 

 7m. it stood at 30'' W. of N. At 9h. 7m. a splendid red blaze in 

 the E., needle N. 30' E. : 9h. 27m. needle 0. At noon on the 

 4th, the needle stood at N. 1° 30' W., so that, (if we assume, as 

 is most probable, that the needle had then regained its usual place,) 

 its north end was not observed during the Aurora, to be carried to 

 the west of its mean position. Circumstances rendered it incon- 

 venient to retain the instrument any longer in the place it occu- 

 pied during the Aurora, so that it can not be confidently asserted 

 that the influence of the Aurora was entirely ended at noon on 

 the 4th. The needle is much less sensitive than that of the va- 

 riation compass formerly employed. Of course I can not com- 

 pare the magnetic effects of this display with those of other great 

 occasions of this kind. During the evening, the temperature 

 was from 70° to. 60°. At 9h. the dew point was 58°, the air be- 

 ing 65°. 



We discontinued our observations a little after llh. at which 

 time the display had greatly declined. A person who was abroad 

 after midnight, informed me that about 1 A. M., (4th,) the spec- 

 tacle was, if possible, more splendid than before. At 4 A. M., I 

 found numerous streamers in active undulation, about the North- 

 ern horizon ; but not reaching to a greater elevation than 20°. 



Dnring the night of Wednesday, the 4th, the sky was densely 

 overcast. A moderate Auroral display was seen at Albany, N. Y., 

 at Middiebury, Vt., and probably at many other places where the 

 state of the weather permitted observation. 



2. Nashville, Tenn. N. lat. 36° 9' 33''; W. Ion. 86° 49'. 

 The following observations are contained in a letter to Prof. Sil- 

 liman from Prof. James Hamilton, of the University of Nashville. 

 " Although a resident in this city for more than six years, be- 

 tween 1827 and 1835, I have never seen so beautiful an exhibi- 

 tion of the Aurora Borealis, as that which occurred in the evening 



