ar atmosphere, it may be said, in a word, that even
those who advocate the existence of vegetation and of clouds of dust or
ice crystals on the moon do not predicate any greater amount, or greater
density, of atmosphere than do those who consider the moon to be wholly
dead and inert. Professor Pickering himself showed, from his
observations, that the horizontal refraction of the lunar atmosphere,
instead of being less than 2 sec., as formerly stated, was less than
0.4 sec. Yet he found visual evidence that on the sunlit side of the
moon this rare atmosphere was filled to a height of four miles with some
absorbing medium which was absent on the dark side, and which was
apparently an emanation from the lunar crust, occurring after sunrise.
And Messrs. Loewy and Puiseux, of the Paris Observatory, say, after
showing reasons for thinking that the great volcanic eruptions belong to
a recent period in the history of the moon, that "the diffusion of
cinders to great distances infers a gaseous envelope of a certain
density.... The resistance of the atmosphere must have been sufficient
to retard the fall of this dust [the reference is to the white trails,
like those from Tycho], during its transport over a distance of more
than 1,000 kilometers [620 miles]."[20]

[Footnote 20: Comptes Rendus, June 23, July 3, 1899.]

We come now to a brief consideration of certain peculiarities in the
motions of the moon, and in the phenomena of day and night on its
surface. The moon keeps the same side forever turned toward the earth,
behaving, in this respect, as Mercury does with regard to the sun. The
consequence is that the lunar globe makes but one rotation on its axis
in the course of a month, or in the course of one revolution about the
earth. Some of the results of this practical identity of the periods of
rotation and revolution are illustrated in the diagram on page 250. The
moon really undergoes considerable libration, recalling the libration of
Mercury, which was explained in the chapter on that planet, and in
consequence we are able to see a little way round into the opposite
lunar hemisphere, now on this side and now on the other, but in the
diagram this libration has been neglected. If it had been represented we
should have found that, instead of only one half, about three fifths of
the total superficies of the moon are visible from the earth at one time
or another.

[Illustration: PHASES AND ROTATION OF THE MOON.]

Perhaps it