All agree upon the desert climate of the mainland on Mars. Like most deserts on the Earth, it is, therefore, probably a table-land, where one plateau mounts above the other, each one nearly level. By the action of the wind, the upper layers have been transformed into fine sand. On the dead planet no further sediments are deposited by the sea. The only accretions to the planet are meteorites and cosmic dust which slowly rains down. It contains among other substances iron, partly metallic and partly in the form of protoxides (which have a light green colour).[5] The oxygen in the atmosphere of Mars transforms these compounds to ferro-oxide which has different colours according to its coarseness, but generally is ochre. The surface of Mars is also described as possessing this colour. Dross has, therefore, assumed that the Martian soil is mingled with ferro-oxide. The finest dust, however, is yellow while larger crystals tend toward violet. We often observe on Mars that the details are covered by a yellow veil. This is of course finely powdered ferro-oxide probably mixed with less coloured sand which the desert wind whirls up over large portions of Mars. Vast sections of the planet bore such wrappings in the autumn of 1909, as observed and described by Antoniadi in Paris (see Fig. 21). Similar observations have previously been made by W. H. Pickering and others. [5] In the deepest portions of the oceans on the Earth where no sediments from the coasts are deposited, large accumulations containing iron compounds have been found and the presence of certain minerals therein points to their meteoric origin. As a rule, only the central and the polar regions of the surface of Mars can be seen. Territories near the equator more than 40 to 50 degrees removed from the point in line with the Sun and the centre of the planet are generally hidden behind a thin, white veil of mist. As soon as the Sun leaves the zenith and reaches half-way to the horizon the moisture of the air is precipitated near the ground. This shows that the planet does not possess any quantities of heat-conserving vapours in its gas shell. The mist does not extend to the poles, whose white caps always appear distinctly, because the Sun cannot greatly affect the evaporation in regions where the Sun’s altitude is neither very high nor very variable. The same holds true for other snow-covered tracts, even if they are not located in the immediate vicinity of the pole. When the supply of water vapour is scant, only the most salient canals come into view. As a rule they do not then appear double, as one of the mates is always less prominent. Lowell showed, he believed, that it always is the same canal out of a pair which first comes to sight and that its position always remains unchanged in contrast to Schiaparelli who has reached the opposite conclusion. This, of course, is quite natural.