In the midst of the mist I made trial of the electrometer of Volta, armed with a smoking match. Though very near a thick wood of heliconias, I obtained very sensible signs of atmospheric electricity. It often varied from positive to negative, its intensity changing every instant. These variations, and the conflict of several small currents of air, which divided the mist, and transformed it into clouds, the borders of which were visible, appeared to me infallible prognostics of a change in the weather. It was only two o'clock in the afternoon; we entertained some hope of reaching the eastern summit of the Silla before sunset, and of re-descending into the valley separating the two peaks, intending there to pass the night, to light a great fire, and to make our negroes construct a hut with the leaves of the heliconia. We sent off half of our servants with orders to hasten the next morning to meet us, not with olives, but with a supply of salt beef.

We had scarcely made these arrangements when the east wind began to blow violently from the sea. The thermometer rose to 12.5 degrees. It was no doubt an ascending wind, which, by heightening the temperature, dissolved the vapours. In less than two minutes the clouds dispersed, and the two domes of the Silla appeared to us singularly near. We opened the barometer in the lowest part of the hollow that separates the two summits, near a little pool of very muddy water. Here, as in the West India Islands, marshy plains are found at great elevations; not because the woody mountains attract the clouds, but because they condense the vapours by the effect of nocturnal refrigeration, occasioned by the radiation of heat from the ground, and from the parenchyma of the leaves. The mercury was at 21 inches 5.7 lines. We shaped our course direct to the eastern summit. The obstruction caused by the vegetation gradually diminished; it was, however, necessary to cut down some heliconias; but these arborescent plants were not now very thick or high. The peaks of the Silla themselves, as we have several times mentioned, are covered only with gramina and small shrubs of befaria. Their barrenness, however, is not owing to their height: the limit of trees in this region is four hundred toises higher; since, judging according to the analogy of other mountains, this limit would be found here only at a height of eighteen hundred toises. The absence of large trees on the two rocky summits of the Silla may be attributed to the aridity of the soil, the violence of the winds blowing from the sea, and the conflagrations so frequent in all the mountains of the equinoctial region.

To reach the eastern peak, which is the highest, it is necessary to approach as near as possible the great precipice which descends towards Caravalleda and the coast. The gneiss as far as this spot preserves its lamellar texture and its primitive direction; but where we climbed the summit of the Silla, we found it had passed into granite. Its texture becomes granular; the mica, less frequent, is more unequally spread through the rock. Instead of garnets we met with a few solitary crystals of hornblende. It is, however, not a syenite, but rather a granite of new formation. We were three quarters of an hour in reaching the summit of the pyramid. This part of the way is not dangerous, provided the traveller carefully examines the stability of each fragment of rock on which he places his foot. The granite superposed on the gneiss does not present a regular separation into beds: it is divided by clefts, which often cross one another at right angles. Prismatic blocks, one foot wide and twelve long, stand out from the ground obliquely, and appear on the edges of the precipice like enormous beams suspended over the abyss.

Having arrived at the summit, we enjoyed, for a few minutes only, the serenity of the sky. The eye ranged over a vast extent of country: looking down to the north was the sea, and to the south, the fertile valley of Caracas. The barometer was at 20 inches 7.6 lines; the thermometer at 13.7 degrees. We were at thirteen hundred and fifty toises of elevation. We gazed on an extent of sea, the radius of which was thirty-six leagues. Persons who are affected by looking downward from a considerable height should remain at the centre of the small flat which crowns the eastern summit of the Silla. The mountain is not very remarkable for height: it is nearly eighty toises lower than the Canigou; but it is distinguished among all the mountains I have visited by an enormous precipice on the side next the sea. The coast forms only a narrow border; and looking from the summit of the pyramid on the houses of Caravalleda, this wall of rocks seems, by an optical illusion, to be nearly perpendicular. The real slope of the declivity appeared to me, according to an exact calculation, 53 degrees 28 minutes.* (* Observations of the latitude give for the horizontal distance between the foot of the mountain near Caravalleda, and the vertical line passing through its summit, scarcely 1000 toises.) The mean slope of the peak of Teneriffe is scarcely 12 degrees 30 minutes. A precipice of six or seven thousand feet, like that of the Silla of Caracas, is a phenomenon far more rare than is generally believed by those who cross mountains without measuring their height, their bulk, and their slope. Since the experiments on the fall of bodies, and on their deviation to the south-east, have been resumed in several parts of Europe, a rock of two hundred and fifty toises of perpendicular elevation has been in vain sought for among all the Alps of Switzerland. The declivity of Mont Blanc towards the Allee Blanche does not even reach an angle of 45 degrees; though in the greater number of geological works, Mont Blanc is described as perpendicular on the south side.

At the Silla of Caracas, the enormous northern cliff is partly covered with vegetation, notwithstanding the extreme steepness of its slope. Tufts of befaria and andromedas appear as if suspended from the rock. The little valley which separates the domes towards the south, stretches in the direction of the sea. Alpine plants fill this hollow; and, not confined to the ridge of the mountain, they follow the sinuosities of the ravine. It would seem as if torrents were concealed under that fresh foliage; and the disposition of the plants, the grouping of so many inanimate objects, give the landscape all the charm of motion and of life.

Seven months had now elapsed since we had been on the summit of the peak of Teneriffe, whence we surveyed a space of the globe equal to a fourth part of France. The apparent horizon of the sea is there six leagues farther distant than at the top of the Silla, and yet we saw that horizon, at least for some time, very distinctly. It was strongly marked, and not confounded with the adjacent strata of air. At the Silla, which is five hundred and fifty toises lower than the peak of Teneriffe, the horizon, though nearer, continued invisible towards the north and north-north-east. Following with the eye the surface of the sea, which was smooth as glass, we were struck with the progressive diminution of the reflected light. Where the visual ray touched the last limit of that surface, the water was lost among the superposed strata of air. This appearance has something in it very extraordinary. We expect to see the horizon level with the eye; but, instead of distinguishing at this height a marked limit between the two elements, the more distant strata of water seem to be transformed into vapour, and mingled with the aerial ocean. I observed the same appearance, not in one spot of the horizon alone, but on an extent of more than a hundred and sixty degrees, along the Pacific, when I found myself for the first time on the pointed rock that commands the crater of Pichincha; a volcano, the elevation of which exceeds that of Mont Blanc.* (* See Views of Nature, Bohn's edition, page 358.) The visibility of a very distant horizon depends, when there is no mirage, upon two distinct things: the quantity of light received on that part of the sea where the visual ray terminates; and the extinction of the reflected light during its passage through the intermediate strata of air. It may happen, notwithstanding the serenity of the sky and the transparency of the atmosphere, that the ocean is feebly illuminated at thirty or forty leagues' distance; or that the strata of air nearest the earth may extinguish a great deal of the light, by absorbing the rays that traverse them.

The rounded peak, or western dome of the Silla, concealed from us the view of the town of Caracas; but we distinguished the nearest houses, the villages of Chacao and Petare, the coffee plantations, and the course of the Rio Guayra, a slender streak of water reflecting a silvery light. The narrow band of cultivated ground was pleasingly contrasted with the wild and gloomy aspect of the neighbouring mountains. Whilst contemplating these grand scenes, we feel little regret that the solitudes of the New World are not embellished with the monuments of antiquity.

But we could not long avail ourselves of the advantage arising from the position of the Silla, in commanding all the neighbouring summits. While we were examining with our glasses that part of the sea, the horizon of which was clearly defined, and the chain of the mountains of Ocumare, behind which begins the unknown world of the Orinoco and the Amazon, a thick fog from the plains rose to the elevated regions, first filling the bottom of the valley of Caracas. The vapours, illumined from above, presented a uniform tint of a milky white. The valley seemed overspread with water, and looked like an arm of the sea, of which the adjacent mountains formed the steep shore. In vain we waited for the slave who carried Ramsden's great sextant. Eager to avail myself of the favourable state of the sky, I resolved to take a few solar altitudes with a sextant by Troughton of two inches radius. The disk of the sun was half-concealed by the mist. The difference of longitude between the quarter of the Trinidad and the eastern peak of the Silla appears scarcely to exceed 0 degrees 3 minutes 22 seconds.* (* The difference of longitude between the Silla and La Guayra, according to Fidalgo, is 0 degrees 6 minutes 40 seconds.)

Whilst, seated on the rock, I was determining the dip of the needle, I found my hands covered with a species of hairy bee, a little smaller than the honey-bee of the north of Europe. These insects make their nests in the ground. They seldom fly; and, from the slowness of their movements, I should have supposed they were benumbed by the cold of the mountains. The people, in these regions, call them angelitos (little angels), because they very seldom sting. They are no doubt of the genus Apis, of the division melipones. It has been erroneously affirmed that these bees, which are peculiar to the New World, are destitute of all offensive weapons. Their sting is indeed comparatively feeble, and they use it seldom; but a person, not fully convinced of the harmlessness of these angelitos, can scarcely divest himself of a sensation of fear. I must confess, that, whilst engaged in my astronomical observations, I was often on the point of letting my instruments fall, when I felt my hands and face covered with these hairy bees. Our guides assured us that they attempt to defend themselves only when irritated by being seized by their legs. I was not tempted to try the experiment on myself.

The dip of the needle at the Silla was one centesimal degree less than in the town of Caracas. In collecting the observations which I made during calm weather and in very favourable circumstances, on the mountains as well as along the coast, it would at first seem, that we discover, in that part of the globe, a certain influence of the heights on the dip of the needle, and the intensity of the magnetical forces; but we must remark, that the dip at Caracas is much greater than could be supposed, from the situation of the town, and that the magnetical phenomena are modified by the proximity of certain rocks, which constitute so many particular centres or little systems of attraction.* (* I have seen fragments of quartz traversed by parallel bands of magnetic iron, carried into the valley of Caracas by the waters descending from the Galipano and the Cerro de Avila. This banded magnetic iron-ore is found also in the Sierra Nevada of Merida. Between the two peaks of the Silla, angular fragments of cellular quartz are found, covered with red oxide of iron. They do not act on the needle. This oxide is of a cinnabar-red colour.)