dea has coloured or governed all his writings on the
subject. The innumerable difficulties which it raises have been either
ignored, or brushed aside on the flimsiest evidence. As examples, he
never even discusses the totally inadequate water-supply for such
worldwide irrigation, or the extreme irrationality of constructing so
vast a canal-system the waste from which, by evaporation, when exposed
to such desert conditions as he himself describes, would use up ten
times the probable supply.

Again, he urges the 'purpose' displayed in these 'canals.' Their being
_all_ so straight, _all_ describing great circles of the 'sphere,' all
being so evidently arranged (as he thinks) either to carry water to some
'oasis' 2000 miles away, or to reach some arid region far over the
equator in the opposite hemisphere! But he never considers the
difficulties this implies. Everywhere these canals run for thousands of
miles across waterless deserts, forming a system and indicating a
purpose, the wonderful perfection of which he is never tired of dwelling
upon (but which I myself can nowhere perceive).

Yet he never even attempts to explain how the Martians could have lived
_before_ this great system was planned and executed, or why they did not
_first_ utilise and render fertile the belt of land adjacent to the
limits of the polar snows--why the method of irrigation did not, as with
all human arts, begin gradually, at home, with terraces and channels to
irrigate the land close to the source of the water. How, with such a
desert as he describes three-fourths of Mars to be, did the inhabitants
ever get to _know_ anything of the equatorial regions and its needs, so
as to start right away to supply those needs? All this, to my mind, is
quite opposed to the idea of their being works of art, and altogether in
favour of their being natural features of a globe as peculiar in origin
and internal structure as it is in its surface-features. The explanation
I have given, though of course hypothetical, is founded on known
cosmical and terrestrial facts, and is, I suggest, far more scientific
as well as more satisfactory than Mr. Lowell's wholly unsupported
speculation. This view I have explained in some detail in the preceding
chapter.

Mr. Lowell never even refers to the important question of loss by
evaporation in these enormous open canals, or considers the undoubted
fact that the only intelligent and practical way to convey a limited
quantity o