The destruction of the forests, the clearing of the plains, and the cultivation of indigo over half a century has affected the amount of water flowing in as well as the evaporation of the soil and the dryness of the air, which forcefully explains why the present Lake Valencia is decreasing. By felling trees that cover the tops and sides of mountains men everywhere have ensured two calamities at the same time for the future: lack of fuel, and scarcity of water. Trees, by the nature of their perspiration, and the radiation from their leaves in a cloudless sky, surround themselves with an atmosphere that is constantly cool and misty. They affect the amount of springs by sheltering the soil from the sun's direct actions and reducing the rainwater's evaporation. When forests are destroyed, as they are everywhere in America by European planters, with imprudent haste, the springs dry up completely, or merely trickle. River beds remain dry part of the year and are then turned into torrents whenever it rains heavily on the heights. As grass and moss disappear with the brushwood from the mountainsides, so rainwater is unchecked in its course. Instead of slowly raising the river level by filtrations, the heavy rains dig channels into the hillsides, dragging down loose soil, and forming sudden, destructive floods. Thus, the clearing of forests, the absence of permanent springs, and torrents are three closely connected phenomena. Countries in different hemispheres like Lombardy bordered by the Alps, and Lower Peru between the Pacific and the Andes, confirm this assertion.

Despite their authority the missionaries could not persuade the Indians to go any further on into the cavern. The lower the vault the more piercing the screaming of the guàcharos became. Thanks to the cowardice of our guides we had to retreat. We found that a bishop of Saint Thomas of Guiana had gone further than us. He had measured 2, feet from the mouth to where he stopped, but the cavern went further. The memory of this feat was preserved in the Caripe convent, without precise dates. The bishop had used torches made from white Castile wax, while we had torches made of tree bark and resin. The thick smoke from our torches in the narrow underground passage hurt our eyes and made breathing difficult.

From the time we left Graciosa the sky remained so consistently hazy that despite the height of the mountains of Gran Canaria we did not make out the island until the evening of the 18th. It is the granary of the archipelago of the Fortunate Islands and, remarkably for an area outside the Tropics, there are two wheat harvests a year, one in February, the other in June. Gran Canaria has never been visited before by a geologist, yet it is worth observing because its mountains differ entirely from those of Lanzarote and Tenerife.

The farmers and their slaves cut a path through the jungle to the first Juagua river waterfall, and on the 10th of September we made our excursion to the Cuchivano crevice. Entering the cave we saw a disemboweled porcupine and smelled the stink of excrement, similar to that of European cats, and knew that a jaguar had been near by. For safety the Indians returned to the farm to fetch small dogs. It is said that when you meet a jaguar in your path he will leap on to a dog before a man. We did not follow the bank of the torrent, but a rocky wall overhanging the water. We walked on a very narrow ledge along the side of a precipice with a drop of some 200 to 300 feet. When it narrowed, so that we could not walk along it any further, we climbed down to the torrent and crossed it on foot, or on the backs of slaves, to climb up the other side. Climbing is very tiring, and you cannot trust the lianas, which, like thick rope, hang from tree-tops. Creepers and parasites hang loosely from the branches they grip; their stalks together weigh a lot, and if you slip and grab one of the lianas you risk bringing down a tangle of green branches. The vegetation became impenetrable the more we advanced. In some places the roots of trees grew in the existing cracks between strata and had burst the calcareous rock. We could hardly carry the plants we picked at each step. The canna, the heliconia with pretty purple flowers, the costus and other plants from the Amomum genus reach here the height of 8 to 50 feet. Their tender, fresh green leaves, their silky sheen and the extraordinary development of their juicy pulp contrast with the brown of the arborescent ferns whose leaves are so delicately jagged. The Indians made deep incisions in the tree trunks with their long knives to draw our attention to the beauty of the red-and gold-colored woods, which one day will be sought after by our furniture makers. They showed us a plant with composite flowers that reaches some 20 feet high (Eupatorium laevigatum), the so-called 'Rose of Belveria' (Brownea racemosa), famous for the brilliance of its purple flowers, and the local 'dragon's blood', a species of euphorbia not yet catalogued, whose red and astringent sap is used to strengthen the gums. They distinguished species by their smell and by chewing their woody fibers. Two Indians, given the same wood to chew, pronounced, often without hesitation, the same name. But we could not take advantage of our guides' wisdom, for how could they reach leaves, flowers and fruit (53) growing on branches some 50 to 60 feet above the ground? We were struck in this gorge by the fact that the bark of the trees, even the ground, were covered in moss and lichen.

A dreadful accident almost made me put off my Orinoco journey, or postpone it for a long time. On the 27th of October, the night before the eclipse, we were strolling along the gulf shore as usual, to take some fresh air and observe high tide. Its highest point in this area was no more than 12 to 13 inches. It was eight at night and the breeze had not begun. The sky was overcast and during this dead calm it was extremely hot. We were crossing the beach that separates the landing-stage from the Guaiquerí Indian village. I heard somebody walking behind me; as I turned I saw a tall man, the color of a mulatto, and naked to the waist. Just above my head he was holding a macana, a huge stick made of palm-tree wood, enlarged at the end like a club. I avoided his blow by leaping to the left. Bonpland, walking at my right, was less lucky. He had noticed the mulatto later than I had; he received the blow above his temple and fell to the ground. We were alone, unarmed, some half a league from any houses, in a vast plain bordered by the sea. The mulatto, instead of attacking me, turned back slowly to grab Bonpland's hat, which had softened the blow and fallen far from us. Terrified at seeing my travelling companion on the ground and for a few seconds unconscious I was worried only about him. I helped him up; pain and anger doubled his strength. We made for the mulatto who, either due to that cowardice typical of his race or because he saw some men far off on the beach, rushed off into the tunal, a coppice of cacti and tree aviccenia. Luck had him fall as he was running, and Bonpland, who had reached him first, began fighting with him, exposing himself to great danger. The mulatto pulled out a long knife from his trousers, and in such an unequal fight we would surely have been wounded if some Basque merchants taking the fresh air on the beach had not come to our aid. Seeing himself surrounded the mulatto gave up all idea of defending himself: then he managed to escape again and we followed him for a long time through the thorny cacti until he threw himself exhausted into a cow shed from where he let himself be quietly led off to prison.

I placed very active curare on the crural nerves of a frog without noticing any change, measuring the degree of its organs' irritability with an arc formed of heterogeneous metals. But these Galvanic experiments hardly worked on birds a few minutes after they had been shot with poison arrows. Curare works only when the poison acts on the vascular system. At Maypures, a colored man (a zambo, a cross between Indian and negro) was preparing one of those poison arrows that are shot in blowpipes, to kill small monkeys or birds for M. Bonpland. He was a carpenter of extraordinary strength. He stupidly rubbed the curare between slightly bleeding fingers and fell to the ground, dizzy for half an hour. Luckily it was a weak curare (destemplado), used for small animals, which may be revived later by placing muriate of soda in the wound. During our journey back from Esmeralda to Atures. I escaped from danger myself. The curare had attracted humidity and become liquid and spilled from a poorly closed jar on to our clothes. We forgot to check the inside of a sock filled with curare when washing our clothes. Just touching this sticky stuff with my hand I realized I should not pull on the poison sock. The danger was all the greater as my toes were bleeding from chigoe wounds.

We passed the night of the 20th of May, the last on the Casiquiare, near the bifurcation with the Orinoco. We hoped to make some astronomical observations as we saw extraordinary shooting stars visible through the mist. Indians, who do not embellish their imagination through words, call shooting stars the 'piss of the stars', and dew the 'spit of the stars'. But the clouds thickened and prevented us from seeing both meteors and stars.