This aurora was due to the _inner vortex ascending_, whose period was at
this time 28 days.
There are several circumstances to be observed in this case. The bow
brightened and faded simultaneously with the aurora, and respected the
vertex of the auroral bank, being apparently concentric with it. The
bow, therefore, depends on the same cause, but differs from the aurora
in being limited to the _surface_ of the atmosphere in which the vortex
has produced a wave to the southward of its central path, as may be
understood by inspecting Fig. 2, Sec. 1,--the figure representing the
polar current of the central vortex. On the 29th of May, 1840,[32] the
author saw a similar phenomenon, at the same time of night, and passing
over the same stars southward until it reached within 5d of Jupiter and
Saturn, to which it was parallel. This atmospheric wave offers a greater
resistance to the passage of the ether: hence the light. On this account
it is, also, that when the passage of a vortex is attended by an auroral
display there will be no thunder-storm. There may be an increase of
wind; but the atmosphere at such times is too dry to make a violent
storm, and there is a silent restoration of the equilibrium, by the
ether passing through the dry atmosphere, without meeting any
condensable vapor, and becoming luminous on account of the greater
resistance of the air when unmixed with vapor. We thus see also the
connection between the aurora and the linear cirri, and we have a
triumphant explanation of the fact, that when the observer is north of
the northern limit of the vortices, he sees the aurora to the south and
not to the north; for, to see it to the northward, he would have to see
it in the same latitude as it appears in the south, and, consequently,
have to see across twice the complement of the latitude. We thus see,
also, why the temperature falls after an aurora; for, the passage of
electricity in any shape, must have this effect on account of the great
specific caloric of this fluid. We see, also, why the aurora should be
more frequent where the magnetic intensity is greatest and be
consequently invisible at the equator, and why the magnetic needle is so
sensibly affected at the time of its occurrence. We may, perhaps, here
be allowed to allude to another phenomenon connected with terrestrial
magnetism and electricity.