no appearance of rain; and what struck me most was that this phenomenon, which perfectly resembled a lunar rainbow, was not in the direction opposite to the moon. The arch remained stationary, or at least appeared to do so, during eight or ten minutes; and at the moment when I tried if it were possible to see it by reflection in the mirror of the sextant, it began to move and descend, crossing successively the Moon and Jupiter. It was 12 hours 54 minutes (mean time) when the summit of the arch sank below the horizon. This movement of an arch, coloured like the rainbow, filled with astonishment the sailors who were on watch on the deck. They alleged, as they do on the appearance of every extraordinary meteor, that it denoted wind. M. Arago examined the sketch of this arch in my journal; and he is of opinion that the image of the moon reflected in the waters could not have given a halo of such great dimensions. The rapidity of the movement is no small obstacle in the way of explanation of a phenomenon well worthy of attention. On the 3rd of December we felt some uneasiness on account of the proximity of a small vessel supposed to be a pirate but which, as it drew near, we recognized to be the Balandra del Frayle (the sloop of the Monk). I was at a loss to conceive what so strange a denomination meant. The bark belonged to a Franciscan missionary, a rich priest of am Indian village in the savannahs (Llanos) of Barcelona, who had for several years carried on a very lucrative contraband trade with the Danish islands. M. Bonpland and several passengers saw in the night at the distance of a quarter of a mile, with the wind, a small flame on the surface of the ocean; it ran in the direction of south-west and lighted up the atmosphere. No shock of earthquake was felt and there was no change in the direction of the waves. Was it a phosphoric gleam produced by a great accumulation of mollusca in a state of putrefaction; or did this flame issue from the depth of the sea, as is said to have been sometimes observable in latitudes agitated by volcanoes? The latter supposition appears to me devoid of all probability. The volcanic flame can only issue from the deep when the rocky bed of the ocean is already heaved up so that the flames and incandescent scoriae escape from the swelled and creviced part without traversing the waters. At half-past ten in the morning of the 4th of December we were in the meridian of Cape Bacco (Punta Abacou) which I found in 76 degrees 7 minutes 50 seconds, or 9 degrees 3 minutes 2 seconds west of Nueva Barcelona. Having attained the parallel of 17 degrees, the fear of pirates made us prefer the direct passage across the bank of Vibora, better known by the name of the Pedro Shoals. This bank occupies more than two hundred and eighty square sea leagues and its configuration strikes the eye of the geologist by its resemblance to that of Jamaica, which is in its neighbourhood. It forms an island almost as large as Porto Rico.