yet, owing to the imperfection of the tables, these epochs are often found to be appreciably in error. There is yet another difficulty. The satellites are not mere points, but, being in reality also as large as or larger than our moon, they have disks of appreciable though small dimensions. Accordingly, they do not vanish or reappear instantaneously, but gradually, the process lasting in reality several seconds (a longer or shorter time, according to the particular satellites considered), and the estimated moment of the phenomenon thus comes to depend on the power of the telescope employed, or the skill or the visual powers of the observer, or the condition of the atmosphere, and so on. Accordingly, very little reliance could be placed on such observations as a mean for determining the longitude with any considerable degree of exactness.
No other celestial phenomena present themselves except those depending on the moon's motions.[1] All the planets, as well as the sun and moon, traverse at various rates and in different paths the sphere of the fixed stars. But the moon alone moves with sufficient
- ↑ If but one star or a few would periodically (and quite regularly) "go out" for a few moments, the intervals between such vanishings being long enough to insure that one would not be mistaken in point of time for the next or following one, then it would be possible to determine Greenwich or other reference time with great exactness. And here one cannot but recognize an argument against the singular theory that the stars were intended simply as lights to adorn our heavens and to be of use to mankind. The teleologists who have adopted this strange view can hardly show how the theory is consistent with the fact that quite readily the stars (or a few of them) might have been so contrived as to give man the means of travelling with much more security over the length and breadth of his domain than is at present possible. In this connection I venture to quote a passage in which Sir John Herschel has touched on the usefulness of the stars, in terms which, were they not corrected by other and better-known passages in his writings, might suggest that he had adopted the theory I have just mentioned: "The stars," he said, in an address to the Astronomical Society, in 1827, "are landmarks of the universe; and, amid the endless and complicated fluctuations of our system, seem placed by its Creator as guides and records, not merely to elevate our minds by the contemplation of what is vast, but to teach us to direct our actions by reference to what is immutable in his works. It is indeed hardly possible to over-appreciate their value in this point of view. Every well-determined star, from the moment its place is registered, becomes to the astronomer, the geographer, the navigator, the surveyor, a point of departure which can never deceive or fail him—the same forever and in all places, of a delicacy so extreme as to be a test for every instrument yet invented by man, yet equally adapted for the most ordinary purposes; as available for regulating a town-clock as for conducting a navy to the Indies; as effective for mapping down the intricacies of a petty barony as for adjusting the boundaries of transatlantic empires. When once its place has been thoroughly ascertained, and carefully recorded, the brazen circle with which the useful work was done may moulder, the marble pillar may totter on its base, and the astronomer himself survive only in the gratitude of posterity; but the record remains, and transfuses all its own exactness into every determination which takes it for a groundwork, giving to inferior instruments, nay, even to temporary contrivances, and to the observations of a few weeks or days, all the precision attained originally at the cost of so much time, labor, and expense." It is only necessary, as a corrective to the erroneous ideas which might otherwise be suggested by this somewhat high-flown passage, to quote the following remarks from the work which represented Sir John Her-
