The rays emanating from Tycho surpass in extent and interest any of the others. Hundreds of distinct light streaks originate round the grey margin of this magnificent object, some of them extending over a greater part of the moon's visible superficies, and "radiating," in the words of Professor Phillips, "like false meridians, or like meridians true to an earlier pole of rotation." No systematic attempt has yet been made to map them accurately as a whole on a large scale, for their extreme intricacy and delicacy would render the task a very difficult one, and, moreover, would demand a long course of observation with a powerful telescope in an ideal situation; but Professor W.H. Pickering, observing under these conditions at Arequipa, has recently devoted considerable attention both to the Tycho and other rays, with especially suggestive and important results, which may be briefly summarised as follows:-- (1.) That the Tycho streaks do not radiate from the apparent centre of this formation, but point towards a multitude of minute craterlets on its south-eastern or northern rims. Similar craterlets occur on the rims of other great craters, forming ray-centres. (2.) Speaking generally, a very minute and brilliant crater is located at the end of the streak nearest the radiant point, the streak spreading out and becoming fainter towards the other end. The majority of the streaks appear to issue from one or more of these minute craters, which rarely exceed a mile in diameter. (3.) The streaks which do not issue from minute craters, usually lie upon or across ridges, or in other similar exposed situations, sometimes apparently coming through notches in the mountain walls. (4.) Many of the Copernicus streaks start from craterlets within the rim, flow up the inside and down the outside of the walls. Kepler includes two such craterlets, but here the flow seems to have been more uniform over the edges of the whole crater, and is not distinctly divided up into separate streams. (5.) Though there are similar craters within Tycho, the streaks from them do not extend far beyond the walls. All the conspicuous Tycho streaks originate outside the rim. (6.) The streaks of Copernicus, Kepler, and Aristarchus are greyish in colour, and much less white than those associated with Tycho: some white lines extending south-east from Aristarchus do not apparently belong to the system. In the case of craterlets lying between Aristarchus and Copernicus the streaks point away from the latter. (7.) There are no very long streaks; they vary from ten to fifty miles in length, and are rarely more than a quarter of a mile broad at the crater. From this point they gradually widen out and become fainter. Their width, however, at the end farthest from the crater, seldom exceeds five miles.