Pickering's views, it seems impossible to see any resemblance between this diminutive agglomeration of lines within a lunar crater, and the great geodetic lines sweeping for hundreds of miles across the face of Mars. [Illustration: PLATE III CHINESE BOWL, SHOWING CRACKLE] In the lunar crater, known as Flammarion's Circle, a most typical branching crack is seen. An examination of these lunar cracks, of which I made drawings through the great telescope at the Lowell Observatory, showed them to be cracks of the most unmistakable character, paralleled on the Earth's surface, by sunbaked fissures. If volcanic forces have caused these cracks in the Moon the same kind of energy should have produced the same general results in Mars, and circular craters should equally be in evidence, for many of the lunar craters are sufficiently large to be detected were they on Mars. They would certainly be indicated on the terminator, and yet not a trace of such markings has been found. It is rather extraordinary, too, that such earthquake fissures on any great scale should not have been filled with trap, silicate, or other injected material. Indeed it is strange that such a triangulating arrangement of cracks has not been found on the Earth's surface. [Illustration: PLATE IV MUD CRACKS ON SHORE OF ROGER'S LAKE, ARIZONA] In order to pronounce the lines on Mars as simply cracks one should study the various kinds of cracks in similar surfaces on the Earth. In such a study he would be amazed at the similarity of cracks. When there is a grain in the substance, as in wood, the cracks follow the grain, though even in this material they are discontinuous. In amorphous material they have essentially the same character; whether in the almost microscopic crackle of old Satsuma pottery, or huge cracks in sun-dried mud, the areas enclosed are generally polygonal. If the material be of impalpable fineness the edges of the cracks are smooth and clean-cut, as in Plate III, from a Chinese bowl; whereas if the material is coarse and pebbly the edges of the cracks are rough and irregular, as in Plate IV, from the muddy shores of a lake. Cracks arising from contraction never converge to a common centre, and when not connected with another crack they taper to a point. They begin at indefinite places and end in an equally indefinite manner. That there should be a common resemblance in cracks due to contraction is evident as they arise from a shrinking of the surface. The most ancient deposits, millions of ages ago, reveal mud cracks differing in no respect from those found to-day. We subjoin a few forms of cracks from various surfaces, to show their essential resemblance. It will be seen that the cracks in the Moon are identical in character to those found on the Mesa at Flagstaff. They start from some indefinite point, are irregular in outline and end as indefinitely. A poor asphalt pavement offers one of the best opportunities for the study of the formation of