The changes which appear to take place in the lunar formations owing to the change in their illumination are much more striking and varied than would be anticipated. But the question arises whether all the changes that are associated with the progress of the lunar day can be ascribed to this effect. Thus, Prof. W. H. Pickering writes concerning a well-known pair of little craters of about nine miles in diameter, "known as Messier and Messier A, situated side by side not far from the centre of the Mare Fecunditatis. When the Sun rises first on them, the eastern one, A, is triangular and larger than Messier, which latter is somewhat pear-shaped. About three days after sunrise they both suddenly turn white, Messier rapidly grows in size, soon surpasses A, and also becomes triangular in shape. Six days after sunrise the craters are again nearly of the same size, owing to the diminution of Messier. The shape of A has become irregular, and differs in different lunations. At nine days after sunrise the craters are exactly alike in size and shape, both now being elliptical, with their major axes lying in a nearly N. and S. direction. Just before sunset A is again the larger, being almost twice the size of Messier."[12] Some observers explain this cycle of changes as due merely to the peculiar contour of the two objects, the change in the lighting during the lunar day altering their apparent figures. Prof. W. H. Pickering, on the other hand, while recognizing that some portion of the change of shape is probably due to the contour of the ground, conceives that, in order to explain the whole phenomenon, it is necessary to suppose that a white layer of hoar frost is formed periodically round the two craters. It is also alleged that whereas Mädler described the two craters as being exactly alike eighty years ago, Messier A is now distinctly the larger; but it is very doubtful whether Mädler's description can be trusted to this degree of nicety. If it could, this would establish a permanent change in the actual structure of the lunar surface at this point. There are several other cases of the same order of ambiguity. The most celebrated is Linné, a white spot about six miles in diameter on the Mare Serentatis. This object appears to change in size during the progress of the lunar day, and, as with Messier, some selenographers consider that it has also suffered an actual permanent change in shape within the last sixty or seventy years. Here again the evidence is not decisive; Neison is by no means convinced that a change has taken place, yet does not think it impossible that Linné may once have been a crater with steep walls which have collapsed into its interior through the force of gravity.