On account of the small amount of water vapour in the atmosphere of Mars true clouds are rare. Figure 22 shows such a cloud at the edge of the planet. The aforementioned mists are often called clouds, for instance by Pickering. [Illustration: Fig. 22. Cloud at the upper right edge of Mars observed by Molesworth, March 7, 1901 ] [Illustration: Fig. 23. Mars as observed by Lowell, July 11, 1907. Even the dark portions appear coursed through by “canals.” ] [Illustration: Fig. 24. Mars, as observed by E. M. Antoniadi, October 6, 1909. ] That elevations really may be found on Mars is evident from the fact that snow or hoar-frost often remains in patches near the pole and occasionally quite far therefrom, for instance on the large island Hellas (40° S. Lat.), while it disappears from the surroundings and sometimes from the pole itself (the south pole). Such a highland covered by ice exists near the south pole, and is shown near the upper edge of Fig. 24. In places where snow always remains, a feeble glacier formation may occur. Most investigators assume that mountains and plateaus exist on Mars, although of modest altitude (Campbell believes that he has observed peaks 3000 m. (9800 ft.) high). Lowell who diligently has looked for mountains at the edge of the illuminated part of Mars, has reached the conclusion that they, if present, cannot rise more than 600 to 900 m. (2000 to 3000 ft.) above the surrounding plains. It were indeed improbable that all inequalities of the Martian surface should have been removed in the process of disintegration, which although at work for enormous extensions of time, has long been extremely feeble and is unassisted by torrents of rain which might rapidly wash the products into the valleys. At present, it is mainly the sand carried by the desert wind that slowly reduces the roughnesses and in this process extensive highlands are hardly touched. But, without the assumption, in itself very unlikely, of a nearly level surface on Mars it becomes difficult to comprehend how the canals, if filled with pure water, can proceed in straight lines without reference to existing differences in altitude. Like the rivers on the Earth they ought to bend according to the topography, even if constructed by engineers.