Inasmuch as in any research the assurance that a phenomenon is real is the first point about it to be established, it is a scientist’s duty, not only to scan the phenomena with jealous care to that end, but to scrutinize every theory which would seek otherwise to account for them—the testing such being only second in importance to observing the things themselves. Accordingly I have examined each of the optical theories that have been advanced and critically compared what they assert and require with the results of observation. The outcome of this research has proved as negativing to any other origin for the double canals than reality as direct observations at the telescope are positive on the point. To show this I shall review each in its consequences, confronting it with what the telescope has to say on the subject; for it is of the pith of the matter that the reality should be as demonstrable on demand as on sight. Furthermore, I shall do this before embarking on the general account of these strange things, because it is vital to any interest that one should be assured from the start of the truth of what he is to read. The preface may seem to him abstruse and prosy, but it will introduce him to some curious optical properties and will eventually enhance his concern by proving to him that what reads like fiction is all the more wonderful for being fact. _I. The Diplopic Theory_ Diplopia is the property of seeing double with one eye. Surprising as it sounds it is an effect not unknown to students of optics, though it usually requires training to produce. It is possible only when the eye is not focused on the object, and is not always possible then. From my experiments its feasibility seems to depend upon whether the focus be beyond or before what it should be. If the eye be focused for a point beyond the object, the object is doubled, if for a point this side of it, the latter is simply blurred. When the double is formed, the amount of the separation of the two images is a function of the distance the focus is out. The greater the discrepancy, the wider apart is the ensuing double. Nor does the image, of a line, for example, stop at doubling. After a certain breadth of separation is reached a third line appears, bisecting the interval between the other two. With yet greater widening the third line itself splits into a pair and so the resolution goes on. In my own experiments I have gone so far as to suspect a fifth line. Far from being unconscious, the process of producing the phenomenon is, with some people, of difficult accomplishment. Mr. Lampland, for instance, of the Flagstaff Observatory, to whom we owe the first photographing of the canals, and who sees the doubles of Mars without difficulty, has hitherto found diplopic vision an impossible feat. Even with the most practiced diplopia is never unconscious except when the object viewed, as a micrometer wire, has nothing to locate it in space.