Upon this characteristic, indeed, depends the climate at the surface of Mars: whether it is at all similar to our own, and whether fluid water is a possibility on Mars or not. While the cosmic relations of the planet in its orbit are quite the same as ours, nevertheless the greater distance of Mars diminishes his supply of direct solar heat to about half what we receive. On the other hand, his distance from the sun during his year of motion around it varies much more widely than ours, so that he receives when nearest the sun about one-half more of solar heat than he does when farthest away. Southern summers on Mars, therefore, must be much hotter, and southern winters colder than the corresponding seasons of his northern hemisphere. Indeed, the length of the southern summer, nearly twice that of the terrestrial season, sometimes amply suffices to melt all the polar ice and snow, as in October, 1894, when the southern polar cap of Mars dwindled rapidly and finally vanished completely. Very interesting in this connection are the researches of Stoney on the general conditions affecting planetary atmospheres and their composition. According to the kinetic theory, if the molecules of gases which are continually in motion travel outward from the center of a planet, as they frequently must, and with velocities surpassing the limit that a planet's gravity is capable of controlling, these molecules will effect a permanent escape from the planet, and travel through space in orbits of their own. So the moon is wholly without atmosphere because the moon's gravity is not powerful enough to retain the molecules of its component gases. So also the earth's atmosphere contains no helium or free hydrogen. So, too, Mars is possessed of insufficient force of gravity to retain water vapor, and the Martian atmosphere may therefore consist mainly of nitrogen, argon, and carbon dioxide. As everyone knows, the axis of the earth if extended to the northern heavens would pass very near the north polar star, which on that account is known as Polaris. In a similar manner the axis of Mars pierces the northern heavens about midway between the two bright stars Alpha Cephei and Alpha Cygni (Deneb). The direction of this axis is pretty accurately known, because the measurement of the polar caps of the planet as they turn round from night to night, year in and year out, has enabled astronomers to assign the inclination of the axis with great precision. These caps are a brilliant white, and they are generally supposed to be snow and ice. They wax and wane alternately with the seasons on Mars, being largest at the end of the Martian winter and smallest near the end of summer. The existence of the polar caps together with their seasonal fluctuations afford a most convincing argument for the reality of a Martian atmosphere, sufficiently dense to be capable of diffusing and transporting vapor.