regions.  But their appearance and their degree of visibility vary greatly, for all of them, from one opposition to another, and even from one week to another, and these variations do not take place simultaneously and according to the same laws for all, but in most cases happen apparently capriciously, or at least according to laws not sufficiently simple for us to be able to unravel.  Often one or more become indistinct, or even wholly invisible, whilst others in their vicinity increase to the point of becoming conspicuous even in telescopes of moderate power.  The first of our maps shows all those that have been seen in a long series of observations.  This does not at all correspond to the appearance of Mars at any given period, because generally only a few are visible at once.[C]

Every canal (for now we shall so call them) opens at its ends either into a sea, or into a lake, or into another canal, or else into the intersection of several other canals.  None of them have yet been seen cut off in the middle of the continent, remaining without beginning or without end.  This fact is of the highest importance.  The canals may intersect among themselves at all possible angles, but by preference they converge toward the small spots to which we have given the name of lakes.  For example, seven are seen to converge in Lacus Phoenicis, eight in Trivium Charontis, six in Lunae Lacus, and six in Ismenius Lacus.

The normal appearance of a canal is that of a nearly uniform stripe, black, or at least of a dark color, similar to that of the seas, in which the regularity of its general course does not exclude small variations in its breadth and small sinuosities in its two sides.  Often it happens that such a dark line opening out upon the sea is enlarged into the form of a trumpet, forming a huge bay, similar to the estuaries of certain terrestrial streams.  The Margaritifer Sinus, the Aonius Sinus, the Aurorae Sinus, and the two horns of the Sabæus Sinus are thus formed, at the mouths of one or more canals, opening into the Mare Erythraeum or into the Mare Australe.  The largest example of such a gulf is the Syrtis Major, formed by the vast mouth of the Nilosyrtis, so called.  This gulf is not less than 1,800 kilometers (1,100 miles) in breadth, and attains nearly the same depth in a longitudinal direction.  Its surface is little less than that of the Bay of Bengal.  In this case we see clearly the dark surface of the sea continued without apparent interruption into that canal.  Inasmuch as the surfaces called seas are truly a liquid expanse, we cannot doubt that the canals are a simple prolongation of them, crossing the yellow areas or continents.