From this arrangement it is at once evident that the canals are not fortuitously placed. That lines should thus meet exactly and in numbers at particular points, and only there, shows that their locating is not the outcome of chance. If very thin rods be thrown haphazard over a surface, the probability that more than two will cross at the same point is vanishingly small. Increasingly assured is it that this would not happen generally. The result we see is therefore not a matter of chance, but of some law working to that end. To the detection of what that law is precedes the easier ascertainment of what it is not. The lines, for example, cannot be rivers, which was the first explanation offered of them by Proctor many years ago, because of their peculiar straightness. Nor can they be channels, the name given to them by Schiaparelli, except in the non-committal sense in which he used the term. For here again their geometric regularity is bar to any estuary-like hypothesis. For quite another reason they cannot be cracks, because of their uniform size throughout. Their unbroken character is another fatal objection to the same suggestion. For cracks in ground never pursue for any great distance a continuous course, any more than they keep uniform or straight. The state of an old ceiling is a case in point. When it breaks, it does so in fissures that proceed a certain way, then give out to be continued by others roughly parallel to the first, but parted from them. The same character is shown by the rills on the moon. The ‘Straight Wall,’ so called, is composed of three such sections, and the little rill to the right of it, west of Birt, of four. Thus were they seen at Flagstaff, and as, to the writer’s knowledge, they have not been so depicted elsewhere, the fact may serve to give some idea of the definition there. That the underlying cause is not explosion or contraction is also evidenced by the canals collectively as well as individually, their arrangement into a system, for cracks, however produced, could only originate from certain centres and could not fit into those starting from others, as the canals invariably do. For each canal goes as undeviatingly to one terminal as it left forthrightly from another. If one wishes to see what explosion or contraction can do, he has only to look at the moon through an opera-glass, when he will be shown a very different sight from what the drawings of Mars detail. Thus just as, considered individually, the lines cannot be watercourses because of their straightness, so they cannot be cracks because of dovetailing into one another.