1805.
September.
The district or quarter called Wilhems Plains, occupies a considerable portion of the interior of the island; its northern extremity borders on the sea by the side of the district of Port Louis, from which it is separated by the Grande Rivière; and it extends southward from thence, rising gradually in elevation and increasing in breadth. The body of the quarter is bounded to the N.E. by the district of Mocha,—to the S.E. by that of Port Bourbon or the Grand Port,—to the south by the quarter of La Savanne,—and to the west by the Plains of St. Piérre. Its length from the sea to the Grand Bassin at its southern extremity, is about five geographic leagues in a straight line, and mean breadth nearly two leagues; whence the superficial extent of this district should not be much less than ninety square miles. In the upper part is a lake called the Mare aux Vacouas, apparently so named from the number of pandanus trees, called vacouas, on its borders; and that part of Wilhems Plains by which the lake is surrounded, at the distance of a league, more or less, bears the appellation of Vacouas; in this part my residence was situate, in a country overspread with thick woods, a few plantations excepted, which had been mostly cleared within a few years.
In consequence of the elevation of Vacouas, the climate is as much different from that of the low parts of the island as if it were several degrees without the tropic; June, July, and August are the driest months at Port Louis, but here they are most rainy, and the thermometer stands from 7° to 12° lower upon an average throughout the year.[1] In a west direction, across that part of the Plains of St. Piérre called Le Tamarin, the sea is not more distant- ↑ The mean height of the thermometer in July 1805, which is the middle of winter, was 67¼°, and of the barometer in French inches and lines, 26.7¾; and during February 1806, the middle of summer, 76° and 26.5¾ were the mean heights. At M. Pitot's house in the town of Port Louis, the averages in the same February were 86° and 27.7¾. According to De Luc, the difference between the logarithms of the two heights of the barometer expresses very nearly the difference of elevation in thousand toises, when the ther-
