We may remark in passing that by one sect of ancient writers the moon was supposed to be a kind of mirror, receiving the image of the earth and reflecting it back to terrestrial spectators. Humboldt affirmed that this opinion had been preserved to his day as a popular belief among the people of Asia Minor. He says, “I was once very much astonished to hear a very well educated Persian from Ispahan, who certainly had never read a Greek book, mention when I showed him the moon’s spots in a large telescope in Paris, this hypothesis as a widely diffused belief in his country: ‘What we see in the moon,’ said the Persian, ‘is ourselves; it is the map of our earth.’” Quite as extravagant an idea, though perhaps a more excusable one, was that held by some ancient philosophers, to the effect that the spots on the moon were the shadows of opaque bodies floating in space between it and the sun. [Illustration: PLATE III. FULL MOON.] An observer watching the forms and positions of the lunar face-marks, from night to night and from lunation to lunation, cannot fail to notice the circumstance that they undergo no easily perceptible change of position with respect to the circular outline of the disc; that in fact the face of the moon presented to our view is always the same, or very nearly so. If the moon had no orbital motion we should be led from the above phenomenon to conclude that she had no axial motion, no movement of rotation; but when we consider the orbital motion in connection with the permanence of aspect, we are driven to the conclusion—one, however, which superficial observers have some difficulty in recognising—that the moon has an axial rotation equal in period to her orbital revolution. Since the moon makes the circuit of her orbit in twenty-seven days and one-third (more exactly 27d. 7h. 43m. 11s.) it follows that this is the time of her axial rotation, as referred to the stars, or as it would be made out by an observer located at a fixed position in space outside the lunar orbit. But if referred to the sun this period appears different; because the moon while revolving round the earth is, with the earth, circulating around the sun. Suppose the three bodies, moon, earth, and sun, to be in a line at a certain period of a lunation, as they are at full moon: by the time the moon has completed her twenty-seven days’ journey around the earth, the latter will have moved along twenty-seven days’ march of its orbit, which is about twenty-seven degrees of celestial longitude: the sun will apparently be that much distant from a straight line passing through earth and moon, and the moon must therefore move forward to overtake the sun before she can assume the full phase again. She will take something over two days to do this; hence the solar period of her revolution becomes more than twenty-nine days (to be exact, 29d. 12h. 44m. 2s. ·87). This is the length of a solar day upon the moon—the interval from one sunrise to another at any