, they may depend on
such conditions; but from the nature of the cause, we should look for
them as much in the zodial light, as in the aurora borealis, regarding
the different intensities.

There is also reason to suspect that the northern side is always the
brightest, both in spring and autumn. On the morning of October 4th,
1853, the light was very vivid and well defined, its northern margin
grazing Regulus and terminating at Mars, which was also to the north of
it. Now, although the _northern side_ was the brightest, the great mass
of light was to the south of the ecliptic, as far down as the cone shape
was preserved; but at 10d from the horizon, a still brighter mass
protruded from the cone towards the north, which was all _north_ of the
ecliptic, and of an irregular form, extending along the horizon. The
time was 4 A.M., and consequently was not due to any crepuscular light.
An explanation of the general fact of the brightest light being _always_
on the north side, is given in the present section, in connection with
another phenomenon. If, as some suppose, the light does not reach to the
sun, the annulus must at least fill all the space between Venus and the
earth, but it is far more in accordance with facts as well as with our
theory, to suppose it increases in density to the body of the sun.

Observations made at the observatory of the British Association,
detected, in 1850, sudden brightenings of the light, altogether
different from pulsations. The theory would refer these to that fitful
irregularity in the momentary intensity of the radial stream, which
gives the flickering and tremulous motion to comets' tails. But, the
steady variations in the intensity of this light must be due to other
causes. The longitude of the sun will here come in as a modifying cause;
for the obstruction caused by the body of the sun, when displaced from
the axis of the vortex, must necessarily exercise an influence on the
force and direction of the radial stream. A sudden influx of cometary
matter down the poles of the vortex, in more than usual quantities, will
also tend to brighten and enlarge the zodial light; and, in this last
cause, we have an explanation not only of ancient obscurations of the
solar light, but, also, of those phosphorescent mists, such as occurred
in 1743 and 1831, rendering moonless nights so light that the smallest
print could be read at midnight.

In total eclipses of the sun, the denser portion of the zod