In principle, the method of finding the longitude from 'lunars,' that is to say, from measurements of the distances between the moon and certain stars, is an exceedingly simple one. In actual practice, it involves a very toilsome calculation, beside exact and careful observation. The principle, as already mentioned, is simply this: The moon travels round the sky, making a complete circuit of the heavens in between twenty-seven and twenty-eight days. It thus moves amongst the stars, roughly speaking, its own diameter, in about an hour. When once its movements were sufficiently well known to be exactly predicted, almanacs could be drawn up in which the Greenwich time of its reaching any definite point of the sky could be predicted long beforehand; or, what comes to the same thing, its distances from a number of suitable stars could be given for definite intervals of Greenwich time. It is only necessary, then, to measure the distances between the moon and some of these stars, and by comparing them with the distances given in the almanac, the exact time at Greenwich can be inferred. As has been already pointed out, the determination of the latitude of the ship and of the local time at any place where the ship is, is not by any means so difficult a matter; but the local time being known and the Greenwich time, the difference between these gives the longitude; and the latitude having been also ascertained, the exact position of the ship is known. There are, of course, difficulties in the way of working out this method. One is, that whilst it takes the sun but twenty-four hours to move round the sky from one noon to the next, and consequently its movements, from which the local time is inferred, are fairly rapid, the moon takes nearly twenty-eight days to move amongst the stars from the neighbourhood of one particular star round to that particular star again. Consequently, it is much easier to determine the local time with a given degree of exactness than the Greenwich time; it is something like the difference of reading a clock from both hands and from the hour hand alone. There are other difficulties in the case which make the computation a long and laborious one, and difficult in that sense; but they do not otherwise affect its practicability. During this voyage to St. Helena, both when outward bound and when returning, Maskelyne gave the method of 'lunars' a very thorough testing, and convinced himself that it was capable of giving the information required. For by this time the improvement of the sextant, or quadrant as it then was, by the introduction of a second mirror, by Hadley, had rendered the actual observation at sea of lunar distances, and of altitudes generally, a much more exact operation. This conclusion he put at once to practical effect, and, in 1763, he published the _British Mariner's Guide_, a handbook for the determination of the longitude at sea by the method of lunars.