iderable distances, which sometimes reached thousands of miles.

The canals seemed to form a kind of network, which connected the various
seas with each other. A few of the more conspicuous of these so-called
canals appeared indeed on some of the drawings made by Dawes and others
before Schiaparelli's time. It was, however, the illustrious Italian
astronomer who detected that these narrow lines are present in such
great numbers as to form a notable feature of the planet. Some of these
remarkable features are shown in Figs. 51 and 52, which are copied from
drawings made by Professor William H. Pickering at the Lowell
Observatory in 1894.

Great as had been the surprise of astronomers when Schiaparelli first
proclaimed the discovery of these numerous canals, it was, perhaps,
surpassed by the astonishment with which his announcement was received
in 1882 that most of the canals had become double. Between December,
1881, and February, 1882, thirty of these duplications appear to have
taken place. Nineteen of these were cases of a well-traced parallel line
being formed near a previously existing canal. The remaining canals were
less certainly established, or were cases where the two lines did not
seem to be quite parallel. A copy of the map of Mars which Schiaparelli
formed from his observations of 1881-82 is given in Plate XVIII. It
brings out clearly these strange double canals, so unlike any features
that we know on any other globe.

[Illustration: PLATE XVIII.

SCHIAPARELLI'S MAP OF MARS IN 1881-82.]

Subsequent observations by Schiaparelli and several other observers
seem to indicate that this phenomenon of the duplication of the canals
is of a periodic character. It is produced about the times when Mars
passes through its equinoxes. One of the two parallel lines is often
superposed as exactly as possible upon the track of the old canal. It
does, however, sometimes happen that both the lines occupy opposite
sides of the former canal and are situated on entirely new ground. The
distance between the two lines varies from about 360 miles as a maximum
down to the smallest limit distinguishable in our large telescopes,
which is something less than thirty miles. The breadth of each of these
remarkable channels may range from the limits of visibility, say, up to
more than sixty miles.

The duplication of the canals is perhaps the most difficult problem
which Mars offers to us for solution. Even if we admit that the canals