h seem to me to bear the unmistakable impress of being due
to natural forces, while they are wholly unintelligible as being useful
works of art. I refer of course to the great system of what are termed
'canals,' whether single or double. Of these I shall give my own
interpretation later on.

CHAPTER III.

THE CLIMATE AND PHYSIOGRAPHY OF MARS.

Mr. Lowell admits, and indeed urges strongly, that there are no
permanent bodies of water on Mars; that the dark spaces and spots,
thought by the early observers to be seas, are certainly not so now,
though they may have been at an earlier period; that true clouds are
rare, even if they exist, the appearances that have been taken for them
being either dust-storms or a surface haze; that there is consequently
no rain, and that large portions (about two-thirds) of the planet's
surface have all the characteristics of desert regions.

_Snow-caps the only Source of Water._

This state of things is supposed to be ameliorated by the fact of the
polar snows, which in the winter cover the arctic and about half the
temperate regions of each hemisphere alternately. The maximum of the
northern snow-caps is reached at a period of the Martian winter
corresponding to the end of February with us. About the end of March the
cap begins to shrink in size (in the Northern Hemisphere), and this goes
on so rapidly that early in the June of Mars it is reduced to its
minimum. About the same time changes of colour take place in the
adjacent darker portions of the surface, which become at first bluish,
and later a decided blue-green; but by far the larger portion, including
almost all the equatorial regions of the planet, remain always of a
reddish-ochre tint.[4]

[Footnote 4: In 1890 at Mount Wilson, California, Mr. W.H. Pickering's
photographs of Mars on April 9th showed the southern polar cap of
moderate dimensions, but with a large dim adjacent area. Twenty-four
hours later a corresponding plate showed this same area brilliantly
white; the result apparently of a great Martian snowfall. In 1882 the
same observer witnessed the steady disappearance of 1,600,000 square
miles of the southern snow-cap, an area nearly one-third of that
hemisphere of the planet.]

The rapid and comparatively early disappearance of the white covering
is, very reasonably, supposed to prove that it is of small thickness,
corresponding perhaps to about a foot or two of snow in north-temperate
America and Europe, and