his time-keepers at N. D. des Anges, by a series of observations of equal altitudes of the sun, the Baron proposed to determine the difference of time at that point and l'Isle de Planier, by signals made at the one station, and observed at the other. He enumerates, however, before describing these, the various ways in which the problem of ascertaining the difference of longitude of places had been attempted to be resolved. One is, as is well known, the occultation of stars by the moon, which, of all the methods purely astronomical, is certainly to be considered as the best. Yet if the star is small, and if it disappears behind the enlightened part of the lunar disk, it is often lost sight of from the comparative weakness of its light, before it actually touch the limb of the moon; and so also, at the emersion, it has perhaps got to some distance from the moon, before it can be distinguished. It is also remarked, that, in occultations, it sometimes happens that the star, after having touched the luminous disk of the moon, still appears for some seconds upon the disk. At first it seems to advance, and afterwards disappears altogether. This is known to have been experienced by several of the most eminent observers; by Cassini, Delahire, Feuillée, &c. Another cause, he adds, which may render such observations defective, is the moon's parallax; in consequence of which, the immersions and emersions of stars are made at different points of the limb for observers placed in different situations. Now, it is certain that the surface of the moon is unequal, and that there are mountains on it which, according to the observations of Messrs Herschell and Schröter, are not less than 4000 toises in height. With good telescopes, one may see the little asperities which their summits form on the limb of the moon. " It was thus that, in observing an occultation of Piscium on the 8th of September 1786, the star appeared to sink in the interval between two of these summits on the moon's limb, and disappeared in the hollow or valley between them. An observer, in another place of the earth, might have seen it hid behind one of these summits; and the two occultations supposed to have been made by the same part of the moon might have differed by several seconds. A singular observation, made by M. Koch, an astronomer of Dantzig, on the 7th of March 1794, on occasion of an occultation of Aldebaran, shows the effect which the mountains in the moon may produce in such cases. The star which just grazed as it were along the limb of the moon, was three times eclipsed by the mountains, before it totally disappeared behind the real disk. The immersion was near the superior horn; the star first disappeared, and in 10" appeared again in all its brightness; after some seconds, it disappear ed, and re-appeared a second time. It was soon after concealed for the third time by a mountain, but appeared once more before the real immersion behind the true disk of the moon." However excellent, therefore, the method of occultations may be for great distances, it is insufficient for affording the necessary accuracy when the distances are small. An error of 1" or 2", which, on an arch of the meridian of several degrees, might be counted as nothing, would become very considerable for a difference of longitude which was only a few seconds. In such cases, the celestial signals must be abandoned, and we must have recourse to such as we can make ourselves on the surface of the earth. The first person who attempted this was Picard, in determining the difference of meridians between the observatory at Copenhagen, and the ruined observatory of Tycho Brahe in the island of Huena. He kindled a fire on the tower at Copenhagen; but he does not tell by what means he made it disappear suddenly. Other methods have been since followed. A trial of the method of finding the difference of time, by means of signals, was made near London in 1775, with great success. The signals were made by the explosion of rockets in the air, which were thrown up from Loampithill, near London, where Mr Aubert, a wellknown lover of astronomy, had his observatory. Dr Maskelyne observed the explosion from the Royal Observatory; Mr Wollaston from Chislehurst in Kent; Mr Heberden from Pall-Mall in London; and Mr Ellicot from Horseley Lane. The longitudes of these five places were thus determined with the greatest precision; the differences at any one place not exceeding a fraction of a second. The greatest distance, however, between these places, was not more than 6 English miles, or 3 French leagues. In the case of greater distances, the same method probably could not be practised with equal success. " It seems singular," the Baron observes, " that for making such signals they have not long since had recourse to the most natural and simple expedient, and the most easy of execution withal, that of kindling a small quantity of gunpowder in the open air. This signal is the most visible and the most instantaneous that can be conceived. It is seen at all seasons, and across rain and fog, even by the naked eye. The sudden flash of the gunpowder strikes the eye, though it be not directed precisely to the point from which the light comes, and even when the place from which the signal is made is under the horizon of the observer. It is not only during the night that these signals may be made, but they may be seen in broad day, with telescopes directed to the place where the signal is made, as I have often experienced; and have, by that means, been relieved from the necessity of passing the night in bivouac, in the open air, as I must otherwise have done. The first use that was made of this method was in the year 1740, by Cassini and Lacaille, in measuring two degrees of longitude near Jet, in Languedoc, and Aix in Provence. These two stations are distant about 40 leagues. Towards the middle of that distance they took a station on the sea side, near the mouth of one of the branches of the Rhone. There, from a terrace on the roof of a church, they set fire, evening and morning, to 10 lib. of powder. The flashes were seen distinctly at both extremities of the line; and the difference of longitude concluded accordly. Cassini proposed to do the same for determining the difference of meridians between Paris and Vienna; but his proposal has never been executed. The quantity of gunpowder which these academicians used, was much too great; and, beside the useless expence, the signals so made were more uncertain and less instantaneous. Even with a single pound of powder, I have observed that the flame lasted for 2" or 3"; and on that account I have never used above 6 or 8 ounces. In 1803, I made these signals on the Brocken, one of the highest mountains of the Hartz, 535 toises above the level of the sea; and the signals were seen at the distance of more than 50 French leagues all round. What is most extraordinary is, that they were seen at the distance of nearly 55 French leagues on the small |