sands, of miles in
length.

The water supply being limited, the inhabitants can not undertake to
irrigate the entire surface of the thirsty land, and convenience of
circulation induces them to extend the irrigated areas in the form of
long lines. The surface of Mars, according to Lowell's observation, is
remarkably flat and level, so that no serious obstacle exists to the
extension of the canal system in straight bands as undeviating as arcs
of great circles.

Wherever two or more canals meet, or cross, a rounded dark spot from a
hundred miles, or less, to three hundred miles in diameter, is seen. An
astonishing number of these appear on Mr. Lowell's charts. Occasionally,
as occurs at the singular spot named Lacus Solis, several canals
converging from all points of the compass meet at a central point like
the spokes of a wheel; in other cases, as, for instance, that of the
long canal named Eumenides, with its continuation Orcus, a single
conspicuous line is seen threading a large number of round dark spots,
which present the appearance of a row of beads upon a string. These
circular spots, which some have regarded as lakes, Mr. Lowell believes
are rather oases in the great deserts, and granting the correctness of
his theory of the canals the aptness of this designation is apparent.[2]

[Footnote 2: The reader can find many of these "canals" and "oases," as
well as some of the other regions on Mars that have received names, in
the frontispiece.]

Wherever several canals, that is to say, several bands of vegetation or
bands of life, meet, it is reasonable to assume that an irrigated and
habitable area of considerable extent will be developed, and in such
places the imagination may picture the location of the chief centers of
population, perhaps in the form of large cities, or perhaps in groups of
smaller towns and villages. The so-called Lacus Solis is one of these
localities.

So, likewise, it seems but natural that along the course of a broad,
well-irrigated band a number of expansions should occur, driving back
the bounds of the desert, forming rounded areas of vegetation, and thus
affording a footing for population. Wherever two bands cross such areas
would be sure to exist, and in almost every instance of crossing the
telescope actually shows them.

As to the gemination or duplication of many of the lines which, at the
beginning of the season, appear single, it may be suggested that, in the
course of the deve