f its original high temperature.

_Astronomical causes of fluctuations in climate._--Sir John Herschel has
lately inquired, whether there are any astronomical causes which may
offer a possible explanation of the difference between the actual
climate of the earth's surface, and those which formerly appear to have
prevailed. He has entered upon this subject, he says, "impressed with
the magnificence of that view of geological revolutions, which regards
them rather as regular and necessary effects of great and general
causes, than as resulting from a series of convulsions and catastrophes,
regulated by no laws, and reducible to no fixed principles." Geometers,
he adds, have demonstrated the absolute invariability of the mean
distance of the earth from the sun; whence it would at first seem to
follow, that the mean annual supply of light and heat derived from that
luminary would be alike invariable: but a closer consideration of the
subject will show, that this would not be a legitimate conclusion; but
that on the contrary, the _mean_ amount of solar radiation is dependent
on the eccentricity of the earth's orbit, and therefore liable to
variation.[205]

Now the eccentricity of the orbit, he continues, is actually
diminishing, and has been so for ages beyond the records of history. In
consequence, the ellipse is in a state of approach to a circle, and the
annual average of solar heat radiated to the earth is actually on the
_decrease_. So far this is in accordance with geological evidence, which
indicates a general refrigeration of climate; but the question remains,
whether the amount of diminution which the eccentricity may have ever
undergone can be supposed sufficient to account for any sensible
refrigeration. The calculations necessary to determine this point,
though practicable, have never yet been made, and would be extremely
laborious; for they must embrace all the perturbations which the most
influential planets, Venus, Mars, Jupiter, and Saturn, would cause in
the earth's orbit, and in each other's movements round the sun.

The problem is also very complicated, inasmuch as it depends not merely
on the ellipticity of the earth's orbit, but on the assumed temperature
of the celestial spaces beyond the earth's atmosphere; a matter still
open to discussion, and on which M. Fourier and Sir J. Herschel have
arrived at very different opinions. But if, says Herschel, we suppose an
extreme case, as if the earth's orbit sh