Chloride solutions, if concentrated, freeze at the following temperatures; that of calcium at -55° C. (-67° F.) that of magnesium at -44° C. (-48.2° F.) and that of sodium at -22° C. (-7.6° F.). If now, as we previously have seen, the mean temperature of Mars as a whole is about -40° C. (-40° F.), of the equatorial belt about -10° C. (+14° F.), and of the pole in the height of summer about 0° C. (32° F.), it is evident that a liquefaction of the ocean surface and of the canals, particularly where salts are deposited, very readily may take place. We should in this connection remember that the ice on Mars is stationary while on Earth it is in motion. The consequence is that sand and dust in the course of thousands of years have accumulated on the bottom of the shallow basins in the polar ice. These seas appear therefore dark in spite of their exceedingly small depth and the white salt and ice-crystals remaining undissolved are unable to display their light colour. Even in the “ocean,” Lowell was persuaded that he had observed canals (see Fig. 23), and it may be possible that cracks are in evidence there, particularly in the most shallow sections, as is the case in the Tyrrhenian sea north of Sicily. It is significant that Flammarion has reached the conclusion, which at first appears highly hazardous, that the freezing point of water is lower on Mars than on Earth. This is entirely correct, if we let water stand for salt solutions. It is customary to point to the strictly uniform breadth and the rectilinear appearance of the canals as clear evidence of their being artificial, _i. e._, the work of engineers. The Italian astronomer Cerulli strongly objected to this conception. “In the exceedingly rare cases when both sides of the canal plainly may be seen,” states Schiaparelli, “I have observed curves and notches in the borders.” This occurred with the canals Euphrates and Triton in 1879, and with the Ganges in 1888. And it would seem obvious that watercourses produced in old furrows would not, as a rule, be of uniform breadth. Antoniadi, by his observations in the autumn of 1909 (see Fig. 17a and 24), has confirmed this opinion, as has le Coultre, who found twice as many irregular canals as rectilinear ones. Antoniadi remarks that some canals appear to be collections of lakes strung out in a certain direction while others are narrow lines which bend and twist. “The complicated network of straight lines is probably illusory.” The spots on Mars, he continues, are very irregular, and “present by no means any geometrical form” (on which the belief largely is founded that they are the product of intelligent beings). “The appearance of the planet reminds one of that of the Moon (except that the latter is dead, _i. e._, unchangeable) or of a terrestrial landscape viewed from a balloon.”--“In a word, the ‘geometry’ of Mars is revealed as a pure illusion.” Exceedingly instructive is a comparison between the two maps