Auroras and Sunset. 147 
“Sept. 3d, Brown Co., fll. We hada splendid Aurora Borealis. 
The light first appeared in the northeast, of a yellow color, and 
spread round the horizon each way, nearly to the west on one 
| side and southeast on the other side. 'The aurora then began to 
shoot up in brilliant pillars of yellow light below and rose-colored 
above. 'These pillars converged to a point, 3° or 4° S. of the 
Zenith ; there was no waving or rolling motion. As the brillianey 
increased above, the northern horizon became dark, like a bank 
of fog unilluminated, and through it the stars were visible. Grad- 
wally it faded away, and bright places E. S. E. and W.N. W. 
| were all that would attract attention. In about half an hour, 
most brilliant columns shot upward from these points, of yellow 
and crimson light, and all over the northern horizon pillars gradu- 
ally developed themselves, and became extremely bright. Mean- 
while, deep crimson light appeared in the southeast and stretched 
ie 
_ Over to the N. W., forming a complete arch about 50° high, and 
under that another arch of a white light about 30° high, both dis- 
tinct, regular and well-defined. The crimson one was absolutely 
intense in its color, as palpable as blood. ‘This continued several 
minutes. All this time the dark bank was black in the northern 
horizon and probably 25° high. At length brilliant yellow pillars 
rose from the northern horizon through the dark bank at several 
_ points, and faded and rose again. Gradually the whole became 
= “Tess brilliant, and soon the splendor of the phenomenon was gone. 
_ Itexceeded in splendor all that I ever saw, except that in January, 
two years ago.” In this, as in many cases, the Aurora seems to 
be black, as well as colored. The two arches were not nearly so 
_ distinct as described in Illinois, though one was nearly complete 
Pee for some time. 
“Is not the color of the light depending upon the height at 
| “which the electric fluid or Aurora is passing, the red making its — 
| _ Way through the lower and denser parts of the atmosphere com- 
| pared with the other? That the phenomenon has any connec= 
| tion with spicula of ice in the upper regions, appears most remo=- 
| ved from any thing tangible. 
IL. Aurora of Sept. 3d, as observed at Olean, N. Y. 
About half an hour after sunset a mild lighting up of the 
heavens in the north, as is usual, or very frequent at this season, 
