The two white caps about the poles increase greatly during severe months of winter and vanish to small spots during the warmth of the Martian summer. These are the reservoirs of moisture for the planet. A series of photographs illustrating the phenomenon of the increase and decrease of these caps is very interesting, for at the height of the Martian winter, for instance, the north polar cap extends to the temperate zone while during the summer it dwindles until it is but a tiny patch a few hundred miles across. As the polar caps melt, they are bordered by a blue belt and this blue belt is only visible as the brilliant caps disappear. Judging from the same phenomena which are manifest on our earth, these polar caps were supposed to be snow. Through the aid of the spectroscope, they are now known to be snow. The ruddy areas on Mars are interpreted as desert regions--great unfertile tracts of land but little affected by changes in climatic conditions, while the large irregular dark or blue-green regions are regarded as marshes or areas covered with vegetation, since they invariably grow darker during the Martian summer and reveal themselves in massed colors changing from the initial blue-green to ocher and finally to chocolate-brown. These changes follow the melting of the polar caps and each hemisphere undergoes the change in turn. Mars, however, is best known and famous for the theory of its so-called "canals." The "canals" were first observed by Schiaparelli, the distinguished Italian observer, in 1877. Schiaparelli thought that the dark irregular areas were oceans and that they were connected by these narrow streaks which he called "canale" or "channels." Unfortunately the word was not translated as "channels" but as "canals," which implies artificial construction, and this latter does therefore not convey the idea which the astronomer sought to convey. These black lines, or "canale," are generally straight and although spoken of as narrow, must be at least 12 miles in width and from a few hundred to three or four thousand miles in length. Starting from points near the polar caps, they follow the arcs of great circles, proceeding to what seem to be centers in the middle of the continent where, most surprisingly, they meet other lines which have come to the same spots. Schiaparelli says that "every channel opens at its end either into a sea or into a lake or into another channel, or else into the intersection of several other channels." "None of them as yet have been seen cut off in the middle of a continent, remaining without beginning or without end ... by preference they converge toward the small spots which we have given the names of lakes." These lakes are sometimes several hundred miles in diameter, although many are very much smaller.