I use the word 'savage' grudgingly because it implies a cultural difference between the tamed Indians living in missions and the free ones, which belies the facts. In the South American jungles there are Indian tribes who live peacefully in villages under their chiefs, who cultivate banana trees, cassava and cotton in large areas of land, and weave their hammocks with cotton fibers. They are not more barbarous than the naked Indians of the missions who have learned to make the sign of the cross. In Europe it is a common fallacy to assume that all Indians who are not tamed are nomadic hunters. In Terra Firma agriculture was known long before the arrival of the Europeans, and today is still practiced between the Orinoco and Amazon rivers in jungle clearings never visited by missionaries. What the missionaries have achieved is to have increased the Indians' attachment to owning land, their desire for secure dwelling places, and their taste for more peaceful lives. It would be accepting false ideas about the actual condition of South American Indians to assume that 'Christian', tamed' and 'civilized' were synonymous with 'pagan', 'savage' and free'. The tamed Indian is often as little a Christian as the free Indian is an idolater. Both, caught up in the needs of the moment, betray a marked indifference for religious sentiments, and a secret tendency to worship nature and her powers.

When you travel through Carib missions and observe the order and submission there it is hard to remind yourself that you are among cannibals. This American word, of doubtful origin, probably comes from the Haitian or Puerto Rican language. It passed into European languages from the fifteenth century as a synonym for anthropophagy. I do not doubt that the conquering island Caribs were cruel to the Ygneris and other West Indian inhabitants, who were so weak and unwarlike; but their cruelty has been exaggerated because the first discoverers listened only to stories from conquered tribes. All the missionaries that I asked assured me that the Caribs are perhaps the least cannibalistic of the New World tribes. Perhaps the desperate way in which the Caribs fought the Spaniards, which led in 1504 to a royal decree declaring them to be slaves, contributed to their fame for ferocity. It was Christopher Columbus who first decided to attack the Caribs and deny them their freedom and natural rights; he was a fifteenth-century man, and less humane than is thought today. In 1520 Rodrigo de Figueroa was appointed by the Spanish Court to decide which South American tribes were Caribs, or cannibals, and which were Guatiaos, or peaceful and friendly to Spain. His ethnographic piece, called El auto de Figueroa, is one of the most curious records of the early conquistadores' barbarism. Without paying attention to languages, any tribe that was accused of eating prisoners was called Carib. All the tribes that Figueroa called Carib were condemned to slavery; they could be sold at will or exterminated. It was after these bloody wars, and the death of their husbands, that Carib women, d'Anghiera says, became known as Amazons.

Until the second half of the eighteenth century the great Apure, Payare, Arauca and Meta rivers were hardly known in Europe by their names, and were obviously far less known than in preceding centuries when the valiant Felipe de Urre (90) and the conquerors of Tocayo crossed the llanos seeking the great city of El Dorado and the rich country of Omagua, the Timbuktu of the New World, beyond the Apure river. Such daring expeditions could take place only on a war footing. The weapons meant to protect the new colonizers were ceaselessly turned against the unhappy Indians. Following that period of violence and misery two Indian tribes, the Cabres and the Orinoco Caribs, became masters of those parts no longer being devastated by the Spaniards. Only poor monks were allowed to advance south of the steppes. The Venezuelan coast became isolated and the slow conquest of the Jesuit missionaries followed the banks of the Orinoco. It is hard to believe that the city of San Fernando de Apure, some 50 leagues from the coast, was not founded until 1789.

We began to load the new pirogue. It was, like all Indian canoes, made from one tree trunk, hollowed out by axe and fire. It was 40 feet long and 3 feet wide. Three people could not squeeze together from one side to the other. These pirogues are so unstable that the weight must be distributed very equally and if you want to stand up for a second you must warn the rowers (bogas) to lean over the other side. Without this precaution water would pour in on the lopsided side. It is difficult to form an idea of the inconveniences of these miserable boats.

A little above the Davipe mission the Río Negro receives a branch of the Casiquiare whose existence is a remarkable phenomenon in the history of river branching. This branch emerges from the Casiquiare, north of Vasiva, under the name Itinivi; and after crossing a flat, virtually uninhabited country some 25 leagues long, pours into the Río Negro under the name of Río Conorichite. It seemed to me, near its mouth, to be 120 toises wide, and added large quantities of white waters to the black waters. Even though the Conorichite current is very fast, you shorten the journey from Davipe to Esmeralda by three days using this canal. It is not surprising to find a double communication between the Casiquiare and the Río Negro when you recall that so many American rivers form deltas when they meet other rivers. In this way the Branco and the Jupura pour into the Río Negro and Amazon through many branches. At the confluence with the Jupura there is another more extraordinary phenomenon. Before joining the Amazon this river, which is its main recipient, sends three branches called Uaranapu, Manhama and Avateprana to the Jupura, which is none other than its tributary. The Amazon thus sends its waters into the Jupura before receiving the waters of the latter back.