In seeking for the origin of the lines he begins by discarding natural causation on the ground first of their straightness, and second of their uniform width, regularities not to be found to any such a degree in the processes of nature. His third ground is “that the lines form a system; that, instead of running anywhither, they join certain points to certain others, making thus, not a simple network, but one whose meshes connect centres directly with one another.... If lines be drawn haphazard over the surface of a globe, the chances are ever so many to one against more than two lines crossing each other at any point. Simple crossings of two lines will of course be common in something like factorial proportion to the number of lines; but that any other line should contrive to cross at the same point would be a coincidence whose improbability only a mathematician can properly appreciate, so very great is it.... In other words, we might search in vain for a single instance of such encounter. On the surface of Mars, however, instead of searching in vain, we find the thing occurring passim; this a priori most improbable rendezvousing proving the rule, not the exception. Of the crossings that are best seen, all are meeting places for more than two canals.”
He then takes up the question of cracks radiating from centres of explosion or fissure, and points out that such cracks would not be of uniform breadth. There are cracks on the moon which look like cracks, while the lines on Mars do not. Moreover, the lines fit into one another which would not be true of cracks radiating from different centres. The lines cannot be rivers for those would not be the same width throughout, or run on arcs of great circles. Nor can the lines be furrows ploughed by meteorites, since these would not run straight from one centre to another; in short the objection from the infinitesimal chance of several lines crossing at the same point applies. “In truth,” he concludes, “no natural theory has yet been advanced which will explain these lines.”
The development, or order in the visibility, of the canals throws light on their nature. Early in the Martian spring they were invisible, then those nearest to the melting snows of its south pole appeared, and in a general succession those farther and farther away; but when they did appear they were always in the same place where they had been seen before. Each canal, however, did not darken all at once, but gradually; and this he accounts for by saying that what we see is not water but vegetation which takes time to develop. “If, therefore, we suppose what we call a canal to be, not the canal proper, but the vegetation along its banks, the observed phenomena stand accounted for. This suggestion was first made some years ago by Professor W. H. Pickering.
“That what we see is not the canal proper, but the line of land it irrigates, disposes incidentally of the difficulty of conceiving a canal several miles wide. On the other hand, a narrow, fertilized strip of country is what we should expect to find; for, as we have seen, the general physical condition of the planet leads us to the conception, not of canals constructed for waterways,—like our Suez Canal,—but of canals dug for irrigation purposes. We cannot, of course, be sure that such is their character, appearances being often highly deceitful; we can only say that, so far, the supposition best explains what we see. Further details of their development point to this same conclusion.” Such as that with time they darken rather than broaden.
To the objection that canals could not be built in straight lines because of mountain ranges he replies that the surface of Mars is surprisingly flat, and this he proves by careful observations of the terminator, that is the edge of that part of the planet lighted by the Sun, where any considerable sudden changes of elevation on the surface of the planet would appear, and do not.
He then tells of the discovery by Mr. Douglass of the canals in the dark regions toward the south pole. They could not be seen while those regions remained dark, but when they faded out the canals became visible, and supplied the missing link explaining how the water from the melting polar cap was conveyed to the canals in the arid space north and south of the equator. Mr. Douglass found no less than forty-four of them, almost all of which he saw more than once, one on as many as thirty-seven occasions.
Then came the phenomenon that convinced Percival of an artificial system of irrigation: “Dotted all over the reddish-ochre ground of the desert stretches of the planet ... are an innumerable number of dark circular or oval spots. They appear, furthermore, always in intimate association with the canals. They constitute so many hubs to which the canals make spokes”; and there is not a single instance of such a spot, unconnected by a canal, and by more than one, with the rest of the system. These spots are in general circular, from 120 to 150 miles in diameter, and make their appearance after, but not long after, the canals that lead to them, those that appear first becoming after a time less conspicuous, those seen later more so. In short they behave as oases of vegetation would when a supply of water had reached them, and thus give “an end and object for the existence of canals, and the most natural one in the world, namely, that the canals are constructed for the express purpose of fertilizing the oases.... This, at least, is the only explanation that fully accounts for the facts. Of course all such evidence of design may be purely fortuitous, with about as much probability, as it has happily been put, as that a chance collection of numbers should take the form of the multiplication table.” He does not fail to point out that great circles for the canals, and circular shapes for the oases, are the forms most economical if artificially constructed; nor does his reasoning rest upon a small number of instances, for up to the close of observations at that time fifty-three oases had been discovered.
Finally he deals with the corroborating phenomena of double canals and the curious dark spots where the canals in the dark regions debouch into those that run through the deserts.
In his conclusion he sums up his ideas as follows:
“To review, now, the chain of reasoning by which we have been led to regard it probable that upon the surface of Mars we see the effects of local intelligence. We find, in the first place, that the broad physical conditions of the planet are not antagonistic to some form of life; secondly, that there is an apparent dearth of water upon the planet’s surface, and therefore, if beings of sufficient intelligence inhabited it, they would have to resort to irrigation to support life; thirdly, that there turns out to be a network of markings covering the disk precisely counterparting what a system of irrigation would look like; and, lastly, that there is a set of spots placed where we should expect to find the lands thus artificially fertilized, and behaving as such constructed oases should. All this, of course, may be a set of coincidences, signifying nothing; but the probability points the other way.”