face, and is called l'Isle de Planier, on which stands a light-house. The position of these two points seemed very favourable for the proposed experiment, which was to be made by determining the difference of latitude of the two points, the convent and the light-house, by astronomical observation, and then connecting them together by a series of triangles, in order to ascertain the same difference by trigonometrical measurement. At N. D. des Anges the hill of Mimet would exert its full action on the plumb-line, or on the liquor in the spirit level. At Isle de Planier, on the contrary, on the surface of an insulated rock, and at the distance of 8000 toises from the land, and 16,000 from the mountain, the action of the hill must amount to nothing; and, consequently, the difference between the amplitude of the arch of the meridian, determined by celestial observation, and inferred from terrestrial measurement, would give the effect of the attraction, uncombined with any other force either assisting or opposing it. The instruments with which the Baron was furnished, were, a repeating circle of twelve inches radius by Reichenbach, with which he proposed to measure the distances from the zenith; a repeating theodolite of eight inches radius by the same artist, for observing azimuths and terrestrial angles; an English sectant of nine inches radius by Troughton, for taking corresponding altitudes, to regulate four chronometers, three constructed

by Josiah Emery of London, and one by Louis Berthoud of Paris. With these he began his observations at the station of Notre Dame des Anges; and, by 874 altitudes, determined the true zenith distances of a great number of stars, all which was done between the 11th and 24th of July 1810.

The situation of this station did not allow the observation of stars on both sides of the zenith, as the mountain rose very perpendicularly to the north of the convent. Such stars might indeed have been observed with an instrument, like the sector, calculated for making observations near the zenith. The repeating circle has not that advantage; for its perpendicularity to the horizon not being very accurately ascertained, and the error arising from that source being greatest near the zenith, the instrument is ill adapted to the observations which, in such a case as the present, would have been the most eligible. The altitudes, therefore, observed, were of stars considerably distant from the zenith, and all of them to the south. Indeed, though Baron de Zach appears to be well pleased with the advantages which both the locality already described, and the instruments that have been enumerated carried with them, we do not think that in either they were remarkable; and if the result, after all, has turned out favourable, it is more to be attributed to the skill, diligence, and accuracy of the observer, than to the particular advantages which

he enjoyed. In one thing, indeed, we cannot but admire the power which an astronomer derives from the fine climate of Marseilles, compared with that of our island. Dr Maskelyne, in a residence on the side of Schehalien of four months, could hardly find the means of placing his sector in the meridian; and, with all that patience or industry could perform, could only make 337 observations. The Baron de Zach, in 13 days, on the shores of the Mediterranean, was able to make 874 observations. That Greenwich should afford, nevertheless, a greater number of observations to be completely depended on, than any other observatory in Europe, is a strong instance how, on some occasions, the moral causes can control the physical.

The data necessary in this way to determine the attraction of Mimet, required observations to be made for finding the difference of latitude between the two extreme points already mentioned; and this must be done, as has been said, not only by celestial observation, but by terrestrial measurement. For these purposes, a great number of observations were made, which the Baron has given, not only in their original state, but also as reduced and prepared for the final calculation, with a degree of order and correctness altogether exemplary. We have never seen any work of the same kind, where there is more method and order in the arrangement, more accuracy in the detail, and

more fairness in striking the mean, where there is any difference among the observations. It is a book, for these reasons, which no one, engaged in similar pursuits can study with too much care.

In his discussion concerning the merit of the instruments employed in his observations, a fact occurs concerning the repeating circle, which is certainly of importance; and, to us who are but little acquainted with the nature of that instrument, seems difficult to be explained. It appears, that these circles are subject to certain variations or anomalies, which may extend to three or four seconds, from causes altogether unknown. Our author

tells us, that he had formerly remarked, in a letter addressed to the Editors of the Bibliothéque Britannique, that one cannot answer, within three or four seconds, for latitudes inferred from a long series of observations agreeing well with one another, and made with the same repeating circle; for another circle will offer another series of observations, agreeing as well with one another, but differing constantly from the first series by 3" or 4". This remark, when first made by Baron de Zach, appears to have drawn upon him a good deal of animadversion, though the fact itself was not disputed.

"They have made it," says the Baron," a kind of reproach, that I had not pointed out the precise source of these variations. My answer was not H h

VOL. IV.

ready, but will appear in due time. In the meanwhile, I have the satisfaction to think, that I have awakened and directed the attention of astronomers and of artists, to an important point which requires their attention."

He goes on to remark, that this defect, from whatever cause it may arise, had no chance of affecting the determination of the attraction of the mountain which was the present object of research. It was not the absolute latitude, either of N. D. des Anges, or of l'Isle de Planier, that was now required; but it was the difference between them, which the constant error above described could have no tendency to affect.

The preceding remarks, however, cast a little uncertainty on the determinations made by these most commodious and useful instruments. We have certainly no right to offer any opinion about an anomaly, which those who are best acquainted with the subject seem hitherto at a loss to explain. It has always appeared to us, that the smallness of the telescopes with which instruments held in the hand must be provided, is a considerable defect, and may perhaps have given rise to the inconsistency just mentioned.

The second article in this Treatise, relates to the difference of longitude between the two stations, and contains several remarks of great value to those engaged in similar pursuits. Having regulated