A number of attempts have been made to explain the _doubling_ of the canals merely as effects of refraction or reflection; and it has even been suggested that it may arise from the telescope not being accurately focussed. The actual doubling of the canals once having been doubted, it was an easy step to the casting of doubt on the reality of the canals themselves. The idea, indeed, was put forward that the human eye, in dealing with detail so very close to the limit of visibility, may unconsciously treat as an actual line several point-like markings which merely happen to lie in a line. In order to test this theory, experiments were carried out in 1902 by Mr. E.W. Maunder of Greenwich Observatory, and Mr. J.E. Evans of the Royal Hospital School at Greenwich, in which certain schoolboys were set to make drawings of a white disc with some faint markings upon it. The boys were placed at various distances from the disc in question; and it was found that the drawings made by those who were just too far off to see distinctly, bore out the above theory in a remarkable manner. Recently, however, the plausibility of the _illusion_ view has been shaken by photographs of Mars taken during the opposition of 1905 by Mr. Lampland at the Lowell Observatory, in which a number of the more prominent canals come out as straight dark lines. Further still, in some photographs made there quite lately, several canals are said to appear visibly double. Following up the idea alluded to in Chapter XVI., that the moon may be covered with a layer of ice, Mr. W.T. Lynn has recently suggested that this may be the case on Mars; and that, at certain seasons, the water may break through along definite lines, and even along lines parallel to these. This, he maintains, would account for the canals becoming gradually visible across the disc, without the necessity of Professor Lowell's "pumping" theory.