Aurora Borealis of September 3, 1839. 263 
for a considerable time, as if caused by the gentle flow of a fluid 
on its surface. The first column from the N. was evidently either 
in or very near the plane of the magnetic meridian. The banks 
continued to increase along the horizon until 9 P. M., when they 
extended from the E. to within 10° of the W. The extremities 
were not bright, but had the usual appearance of light clouds. 
At this most westeily point, a shower had arisen about sunset 
which had been driven toward the S. W., and during the Aurora 
flashes of lightning at a great distance were occasionally seen. 
These became less and less vivid, the storm being driven away 
by aN. E. wind, blowing at the rate of three miles, and inereasing 
to about six miles an hour during the phenomenon. ‘The needle 
(about 10 inches long, and very sensitive,) vibrated through an 
arc of 1° 10’, while the columns were apparent, and on the fol- 
lowing morning it had rested in an intermediate position only 20’ 
W. of its greatest eastern limit during the evening. The plate 
of an electrical machine, in an adjoining room, showed more than 
usual activity, giving after two turns, pungent and loud sparks to 
the knuckles three inches distant. 'The season has been exceed- 
ingly dry since the middle of July. ° 
“Taking the direction of my transit telescope, I found the 
magnetic variation this day, (Sept. 7, 1839,) to be 5° 56’ E.” 
3. At New Orleans, La., (N. lat. 29° 58’) the Auroral display 
Was quite conspicuous, and appeared so much like a large conflia- 
gration, that the fire engines were called out to extinguish the 
flames. ‘The altitude of the streamers is not meutioned. No 
corona was probably formed at this place. Being desirous to as- 
certain how far south-a corona was visible, I made special inqui- 
ries of a friend at Claiborne; Ala., (N. lat. 314°) where the Au- 
rora was very splendid, and learn that it could scarcely be said 
that a corona was formed there, although several times the Auro- 
ral columns were nearly united overhead. It is probable that a 
corona might have been seen within a hundred miles north of this. 
A. Carlyle, Ill. (N. lat. 384; W. lon. 894.) Prof. John Locke 
has published in the Daily Missouri Republican, St. Louis, Sept. 8, 
1839, a description of this Aurora, from his own observations. 
The display was of the most splendid character, and the point to 
which the streamers converged was determined by him to coin- 
cide exactly with that to which the dipping needle is there di- 
rected. Dr. L. had a few days previous found the magnetic 
