they choose. The Eumenides--Orcus, for example, pursues the even tenor of its unswerving course for nearly 3500 miles. Now, it might be possible so to select one's country that one canal should be able to do this; but that every canal should be straight, and many of them fairly comparable with the Eumenides- Orcus in length, seems to be beyond the possibility of contrivance.

In this dilemma between mountains on the one hand and canals on the other, a certain class of observations most opportunely comes to our aid; for, from observations which have nothing to do with the lines, it turn; out that the surface of the planet is, in truth, most surprisingly flat. How this is known will most easily be understood from a word or two upon the manner in which astronomers have learnt the height of the mountains in the Moon.

The heights of the lunar mountains are found from measuring the lengths of the shadows they cast. As the Moon makes her circuit of the Earth, a varying amount of her illuminated surface is presented to our view. From a slender sickle she grows to full moon, and then diminishes again to a crescent. The illuminated portion is bounded by a semicircle on the one side, and by a semi-ellipse on the other. The semicircle is called her limb, the semi-ellipse her terminator. The former is the edge we see because we can see no farther; the latter, the line upon her surface where the sun is just rising or setting. Now, as we know, the shadows cast at sunrise or sunset are very long, much longer than the objects that cast them are high. This is due to the obliquity at which the light strikes them; the same effect being produced by any sufficiently oblique light, such as an electric light at a distance. Imperceptible in themselves, the heights become perceptible by their shadows. A road illuminated by a distant arc light gives us a startling instance of this; the smooth surface taking on from its shadows the look of a ploughed field.

It is this indirect kind of magnification that enables astronomers to measure the lunar mountains, and even renders such vicariously visible to the naked eye. Every one has noticed how ragged and irregular the inner edge of the Moon looks, while her outer edge seems perfectly smooth. In one place it will appear to project beyond the perfect ellipse, in another to recede from it. The first effect is due to mountain tops catching the sun's rays before the plains about them; the other, to mountain tops further advanced into the lunar day, whose shadows still shroud the valleys at their feet. Yet the elevations and depressions thus rendered so noticeable vanish in profile on the limb.

Much as we see the Moon with the naked eye do we see Mars with the telescope. Mars being outside of us with regard to the Sun, we never see him less than half illumined, but we do see him with a disk that lacks of being round,-- about what the Moon shows us when two days off from full. It is when he is in quadrature--that is, a quarter way round the celestial circle from the Sun-- that he shows thus, and we see him then with the telescope at closer range than we ever see the Moon without it. So observed we notice at once that his terminator, or inner edge, presents a very different appearance from the lunar one. Instead of looking like a saw, it looks comparatively smooth, like a knife. From this we know that, relatively to his size, he has no elevations or depressions upon his surface comparable to the lunar peaks and craters.

His terminator, however, is not absolutely perfect. Irregularities are to be detected in it, although much less pronounced than those of the Moon. His irregularities are of two kinds. The first, and by all odds the commonest phenomenon, consists in showing himself on occasions surprisingly flat; not in this case an inferable flatness, but a perfectly apparent one. In other words, his terminator does not show as a semi-ellipse, but as an irregular polygon. It looks as if in places the rind had been pared off. The peel thus taken from him, so to speak, is from twenty to forty degrees wide, according to the particular part of his surface that shows upon the terminator at the time. The other kind is short and sharp. Now it will be remembered that we considered both kinds under the question of atmosphere, and we found both to be explicable as the effect of clouds, but not the effect of mountains. We may therefore feel tolerably certain that Mars is a flat world; devoid, as we may note incidentally, of summer resorts, since it possesses, apparently, neither seas nor hills. To canals we will now return.