n, in the same manner as the solar rays, and refracted and
reflected by the drops; but the colours are faint in consequence of the
feeble light of the moon compared with that of the sun. A lunar rainbow
has been thus described by an observer:--"The moon was truly 'walking in
brightness,' brilliant as she could be, not a cloud was to be seen near
her; and over against her, toward the north-west, or perhaps rather more
to the north, was a rainbow, a vast arch, perfect in all its parts, not
interrupted or broken as rainbows frequently are, but unremittedly
visible from one horizon to the other. In order to give some idea of its
extent, it is necessary to say, that, as I stood toward the western
extremity of the parish of Stoke Newington, it seemed to take its rise
from the west of Hampstead, and to end perhaps in the river Lea, the
eastern boundary of Tottenham. Its colour was white, cloudy, or greyish,
but a part of its western limb seemed to exhibit tints of a faint sickly
green. After some time the moon became darkened by clouds, and the
rainbow of course vanished."

[Picture: Lunar Rainbow]

The brilliant colours of the solar rainbow are frequently produced by the
clouds without any prismatic arrangement. The light of the sun is
decomposed by a process called absorption: for example, white light is
composed of red, yellow, and blue rays, in certain proportions; now, if
in passing through, or falling upon any substance whatever, the red rays
are stifled or absorbed, while the yellow and blue are allowed to pass or
to be reflected, it is obvious that such a substance cannot appear white,
because one of the elements of white light, namely, the red, is wanting;
it must therefore appear of such a colour as results from the combination
of yellow and blue; the substance will therefore appear green. So, also,
when white light falls upon what we call a _red_ surface, the yellow and
blue rays are stifled or absorbed, leaving the red only to be reflected.
Now, when we consider the various ways in which this absorption may take
place; one or two, or all of the coloured rays being absorbed in every
possible proportion, it is easy to form some idea of the manner by which
the innumerable tints of the sky are produced.

It has been calculated, that, of the horizontal sunbeams which pass
through two hundred miles of air, scarcely a two thousandth part reaches
the earth. A densely formed cloud must there