Atmospheric Origin of the Aurora, Sf. 157 
auroras Would -have been the precursor. As tending to eluci- 
date this new and interesting field of inquiry, I will state the re- 
sults of observations on thirty two auroras observed at Schenec- 
tady, N. Y., between Oct: 5, 1830, and Nov. 3, 1833—the tables 
being prepared for these alone, although the results of subsequent 
observations are, I am persuaded, not less striking. My observa- 
tions are made at 9 A.M. and 9 P.M. The proportions of sky 
clear at the times of observation, are set down in tenths, About 
one day before an aurora, the sky usually begins to increase in 
clearness.. In the following results, reference was had only to 
clearness as compared with the corresponding hour of the prece- 
ding day, and only to mean results. During the 24 hours prece- 
ding the morning of the day on the evening of which the aurora 
occurred, the sum of the increments of clearness was to that of 
the decrements as two to one.* During the 24 hours immedi- 
ately preceding the aurora, the increments are to the decrements 
as six toone. Similar results would be obtained by taking the 
number of instances in which the clearness increased or dimin- 
ished in case of different auroras, instead of the amount of tenths, 
as above. Within the) two days preceding an aurora, and on 
some part of the night-of it, we observe all the circumstances 
preparatory to and connected . with crystallizations in the high 
regions, developing themselves; such as increasing atmospheric 
pressure, increase of cold, and the disappearance of clouds. On 
the other hand, during the day or two succeeding it, are devel- 
oped all. those circumstances which attend a more advanced stage 
and lower descent of the products, whether crystalline or melted ; 
such as a diminution of atmospheric pressure and clearhess, and 
an elevation of the temperature and dew point. The latter 
changes, oceupying less time, are more rapid than the former, and 
appear more striking. For example, during the 24 hours 
‘succeeding an aurora, the decrements of clearness are to the in- 
crements as 37.to 1. But this high ratio requires in reality to be 
further increased, in conformity with the principles above estab- 
lished. fot, the mee increase of Caney which occurred, 
- * The sum of the tenths ‘which. saan express ss the amount by chad 
sky became clearer on the respective days immediately preceding the atin. au- 
Toras, is : called, in, expressing the mean results, the sum of the increments during 
a hours iminediately aie the aurora. A similar Soeaeareye is used 1 
Other epochs and for the decrements. - 
