With regard to the origin of the lunar craters there has been much discussion. Some have considered them to be evidence of violent volcanic action in the dim past; others, again, as the result of the impact of meteorites upon the lunar surface, when the moon was still in a plastic condition; while a third theory holds that they were formed by the bursting of huge bubbles during the escape into space of gases from the interior. The question is, indeed, a very difficult one. Though volcanic action, such as would result in craters of the size of Ptolemæus, is hard for us to picture, and though the lone peaks which adorn the centres of many craters have nothing reminiscent of them in our terrestrial volcanoes, nevertheless the volcanic theory seems to receive more favour than the others. In addition to the craters there are two more features which demand notice, namely, what are known as _rays_ and _rills_. The rays are long, light-coloured streaks which radiate from several of the large craters, and extend to a distance of some hundreds of miles. That they are mere markings on the surface is proved by the fact that they cast no shadows of any kind. One theory is, that they were originally great cracks which have been filled with lighter coloured material, welling up from beneath. The rills, on the other hand, are actually fissures, about a mile or so in width and about a quarter of a mile in depth. The rays are seen to the best advantage in connection with the craters Tycho and Copernicus (see Plate XI., p. 204). In consequence of its fairly forward position on the lunar disc, and of the remarkable system of rays which issue from it like spokes from the axle of a wheel, Tycho commands especial attention. The late Rev. T.W. Webb, a famous observer, christened it, very happily, the "metropolitan crater of the moon." [Illustration: PLATE XI. THE MOON The systems of rays from the craters Tycho, Copernicus and Kepler are well shown here. From a photograph taken at the Paris Observatory by M.P. Puiseux. (Page 204)]