The syndic in Cumanà had recommended us to the missionaries who run the Chaima Indian mission. This recommendation was all the more useful to us as the missionaries, zealous for the purity of their parishioners' morals, and wary of the indiscreet curiosity of strangers, tended to apply an ancient rule of their order according to which no white secular person could remain more than one night in an Indian village. In general, to travel agreeably in Spanish missions it would not be wise to trust solely to the passport issued by the Madrid Secretary of State: you must arm yourself with recommendations from the ecclesiastical authorities, especially the custodians of convents, or the generals of orders residing in Rome, who are far more respected by missionaries than bishops. Missions have become, in reality, a distinct, almost independent, hierarchy, despite being primitive or canonical institutions.

South of the ravine, in the plain that stretches to the lake shore, another less hot and less gassy sulphurous spring gushes out. The thermometer reached only 42°C. The water collects in a basin surrounded by large trees. The unhappy slaves throw themselves in this pool at sunset, covered in dust after working in the indigo and sugar-cane fields. Despite the water being 52°C to 14°C warmer than the air the negroes call it refreshing. In the torrid zone this word is used for anything that restores your strength, calms nerves or produces a feeling of well-being. We also experienced the salutary effects of this bath. We had our hammocks slung in the trees shading this pond and spent a whole day in this place so rich in plants. Near this bãno de Mariara we found the volador or gyrocarpus. The winged fruits of this tree seem like flying beings when they separate from the stem. On shaking the branches of the volador, we saw the air filled with its fruits, all falling together. We sent some fruit to Europe, and they germinated in Berlin, Paris and Malmaison. The numerous plants of the volador, now seen in hothouses, owe their origin to the only tree of its kind found near Mariara.

From the 28th of October to the 3rd of November the reddish mist was thicker than usual: at night the heat was stifling yet the thermometer did not rise beyond 26°C. The sea breeze, which usually refreshed the air from eight to nine at night, was not felt at all. The air was sweltering hot, and the dusty, dry ground started cracking everywhere. On the 4th of November, around two in the afternoon, extraordinarily thick black clouds covered the tall Brigantín and Tataraqual mountains, and then reached the zenith. At about four it began to thunder way above us without rumbling; making a cracking noise, which often suddenly stopped. At the moment that the greatest electrical discharge was produced, twelve minutes past four, we felt two successive seismic shocks, fifteen seconds from each other. Everybody ran out into the street screaming. Bonpland, who was examining some plants, leaning over a table, was almost thrown to the floor, and I felt the shock very clearly in spite of being in my hammock. The direction of the earthquake was from north to south, rare in Cumanà. Some slaves drawing water from a well, some 18 to 20 feet deep next to the Manzanares river, heard a noise comparable to artillery fire, which seemed to rise up out of the well; a surprising phenomenon, though quite common in American countries exposed to earthquakes.

We left finally on the 22nd of September, with four mules carrying our instruments and plants. Leaving the Caripe valley we first crossed a chain of hills running north-east from the convent. The road was uphill across a wide savannah towards the tableland of Guardia de San Agustín. We waited there for the Indian who carried our barometer. We stood at 533 toises above sea-level. The savannahs, or natural grasslands, excellent pasture for the monastery cows, were totally devoid of trees and shrubs. It is the kingdom of the monocotyledonous plants, with only a few maguey (Agave americana), whose flowers rise some 26 feet above the grass. On the high tableland of Guardia we felt we had been carried to an ancient lake flattened by the prolonged effect of the waters. You think you can recognize the curves of the ancient shore, the prominent tongues of land, the steep cliffs forming islands. Even the distribution of flora seems to refer to that primitive state. The bottom of the basin is a savannah, while its edges are covered with tall trees. It is sad that a place so favored by its climate, and fit for wheat, should be totally uninhabited.

Only after Diamante do you enter territory inhabited by tigers, crocodiles and chiguires, a large species of Linnaeus's genus Cavia (capybara). We saw flocks of birds pressed against each other flash across the sky like a black cloud changing shape all the time. The river slowly grew wider. One of the banks is usually arid and sandy due to flooding. The other is higher, covered with full-grown trees. Sometimes the river is lined with jungle on both sides and becomes a straight canal some 150 toises wide. The arrangement of the trees is remarkable. First you see the sauso shrubs (Hermesia castaneifolia), a hedge some 4 feet high as if cut by man. Behind this hedge a brushwood of cedar, Brazil-wood and gayac. Palms are rare; you see only scattered trunks of corozo and thorny piritu. The large quadrupeds of these regions, tigers, tapirs and peccaries, have opened passages in the sauso hedge. They appear through these gaps to drink water. They are not frightened of the canoes, so we see them skirting the river until they disappear into the jungle through a gap in the hedge. I confess that these often repeated scenes greatly appeal to me. The pleasure comes not solely from the curiosity a naturalist feels for the objects of his studies, but also to a feeling common to all men brought up in the customs of civilization. You find yourself in a new world, in a wild, untamed nature. Sometimes it is a jaguar, the beautiful American panther, on the banks; sometimes it is the hocco (Crax alector) with its black feathers and tufted head, slowly strolling along the sauso hedge. All kinds of animals appear, one after the other. 'Es como en el paradiso' ('It is like paradise') our old Indian pilot said. Everything here reminds you of that state of the ancient world revealed in venerable traditions about the innocence and happiness of all people; but when carefully observing the relationships between the animals you see how they avoid and fear each other. The golden age has ended. In this paradise of American jungles, as everywhere else, a long, sad experience has taught all living beings that gentleness is rarely linked to might.

Further to the south-west the soil turns dry and sandy. We climbed a relatively high range that separates the coast from the great plains or savannahs bordering the Orinoco. That section of the mountains through which the road to Cumanacoa leads is devoid of vegetation and falls steeply both to the south and north. It has been called Imposible because this impenetrable mountain ridge would offer a refuge to the inhabitants of Cumanà during a hostile invasion. We reached the top just before sunset. I scarcely had the time to take a few horary angles with my chronometer to calculate the geographic longitude of the place.

The chemical operation, whose importance is exaggerated by the master of the curare, seemed to us very simple. The bejuco used to make the poison in Esmeralda has the same name as in the Javita jungles. It is the bejuco de mavacure, which is found in abundance east of the mission on the left bank of the Orinoco. Although the bundles of bejuco that we found in the Indian's hut were stripped of leaves, there was no doubt that they came from the same plant of the Strychnos genus that we examined in the Pimichin jungles. They use either fresh mavacure or mavacure that has been dried for several weeks. The sap of a recently cut liana is not considered as poisonous; perhaps it only really works when it is very concentrated. The bark and part of the sapwood contain this terrible poison. With a knife they grate some mavacure branches; the bark is crushed and reduced to thin filaments with a stone like those used to make cassava flour. The poisonous sap is yellow, so all this matter takes on that color. It is thrown into a funnel some 9 inches high and 4 inches wide. Of all the instruments in the Indian's laboratory, this funnel is the one he was most proud of. He several times asked if por alla (over there, in Europe) we had seen anything comparable to his embudo. It was a banana leaf rolled into a trumpet shape, and placed into another rolled trumpet made of palm leaves; this apparatus was held up by a scaffolding made of palm-leaf stalks. You begin by making a cold infusion, pouring water on the fibrous matter that is the crushed bark of the mavacure. A yellow water filters through the leafy funnel, drop by drop. This filtered water is the poisonous liquid; but it becomes strong only when concentrated through evaporation, like molasses, in wide clay vessels. Every now and then the Indian asked us to taste the liquid. From its bitterness you judge whether the heated liquid has gone far enough. There is nothing dangerous about this as curare only poisons when it comes into contact with blood. The steam rising from the boiler is not noxious, whatever the Orinoco missionaries might say.