high dark mountains, with abrupt cliffs cropping out. A flock of vicuñas dashed across our path, disappearing again in the driving sleet. After wading through snow and mud for several leagues the weather cleared up, and we began to descend a splendid gorge, exactly like some of the finest coombs on the north coast of Devon, on a gigantic scale. This led us down into a valley, where I parted with my young vicuña-hunter, who had been a very pleasant companion. Riding down the grassy valley, and passing many flocks of sheep, I rode through the village of Potoni, a dozen huts on the side of a hill; forded the river Azangaro, which is here but a small stream even in the rainy season; and riding up the opposite bank, got a magnificent view of the snowy mountains of Caravaya, with their sharp needle-like peaks. Two leagues brought me to Crucero, the capital of the province of Caravaya, so called from the cross-roads which here branch off to the various villages in the forests on the other side of the snowy barrier which rises up close to the town, to the eastward.

Crucero is a collection of comfortless mud-houses, with a small dilapidated church in the plaza, on a very elevated swampy plain. It was intensely cold, with heavy snowstorms during the nights, and the people sat wrapped up in cloaks without fires, shivering in a dreary helpless way, and going to bed soon after sunset, as the only comfortable place. I was most kindly received by the sub-prefect, Don Pablo Pimentel, a veteran soldier, and an official who had served many years at the head of the Government in Caravaya, and in Lampa. Dr. Weddell had named a new genus of chinchonaceous plants Pimentelia, in honour of the worthy old sub-prefect, which had pleased him very much. I remained a few days in Crucero, before setting out for the chinchonaforests in the valleys of Sandia and Tambopata; and during that time I obtained a good deal of information from Don

Pablo Pimentel, and from Señor Leefdael the Judge, respecting the province of Caravaya. Don Pablo had travelled over almost every part of it; and I also received much information at Arequipa from Don Agustin Aragon, a former sub-prefect, who has a large estate in the Caravaya forests. From these sources I am enabled to offer some account of those parts of Caravaya which I did not visit, and which will form the subject of the following chapter. Caravaya is a region of which little is known to European geographers, and, so far as I am aware, no traveller has yet given any account of it to the English public.

CHAPTER XII.

THE PROVINCE OF CARAVAYA.

A short Historical and Geographical Description.

THE Peruvian province of Caravaya is drained by streams which form part of the system of one of the largest and least known of the tributaries of the Amazon-the river Purus.

The Purus is the only great affluent flowing into the Amazon from the south, the course of which has never yet been explored. We have detailed accounts of the Huallaga from Maw, Smyth, Poeppig, and Herndon; of the Ucayali from Smyth, Herndon, and Castelnau; and of the Madeira from Castelnau and Gibbon; but of the Purus, the largest apparently, and one which, in course of time, will probably become the most important, we have next to nothing. Its mouth, and the course of its tributaries, near the base of the Andes, are alone described.

Condamine and Smyth, in descending the Amazon, mention the great depth and volume of water at the mouth of the Purus: Herndon heard from a Brazilian trader at Barra, who had ascended its stream for some distance, that it was of great size, and without obstructions; and Haënke, in the last century, arguing from reliable geographical data which he had collected from Indians, stated his conviction that a very large river, flowing from the Andes east of Cuzco, reached the Amazon to the westward of the mouth of the Madeira.

This is the sum of our knowledge of the mouth and lower course of the Purus. The tributaries which flow into it drain the eastern slopes of the Andes, from the latitude of Cuzco

quite to the frontier of Bolivia-that frontier dividing the streams flowing into the Purus, on the Peruvian side, from those which feed the Beni, on the Bolivian. These affluents of the Purus are divided into three distinct systems: the furthest to the north and west, consisting of the streams flowing through the great valley of Paucartambo, which unite under the name of the Madre de Dios, or Amaru-mayu; the middle system, draining the ravines of Marcapata and Ollachea; and the southern and eastern, being the numerous rivers in the province of Caravaya, as far as the Bolivian frontier, which unite as the Ynambari. The Madre de Dios and Ynambari together form the main stream of the Purus.

The Paucartambo system is the only one which has, as yet, been described by modern explorers. In Spanish times the streams which compose it were explored, and farms of cacao and coca were established on their banks; and in the end of the last century an expedition was sent to explore the course of the Madre de Dios, under an officer named Don Tiburcio de Landa. This must have been at some time previous to 1780, for Landa was killed in that year in the great rebellion of the Indians under Tupac Amaru.1 After the declaration of Peruvian independence, General Gamarra, the first Republican Prefect of Cuzco, sent an expedition to protect the farms in the valley of Paucartambo from the encroachments of the wild Chuncho Indians, and to explore the Madre de Dios. It was commanded by a Dr. Sevallos, now a very old man, retired to a farm in the Caravaya forests, but he has, unfortunately, lost his journal. General Miller made an expedition into the same region in 1835, and penetrated to a greater distance than any other explorer

1 Landa sent in a report of his expedition to the Corregidor of Cuzco. My friend Dr. Don Julian Ochoa, the rector of the university of Cuzco, has recently searched the archives of the

ancient municipality of that city, as well as private collections, for this interesting document, at my request, but without success.

before or since. A very brief account of his journey was published in the 'Royal Geographical Society's Journal' for 1836; but there is a much fuller and most interesting journal kept by this gallant veteran, which has never been printed. In 1852 Lieut. Gibbon, U.S.N., entered the valleys of Paucartambo; and in 1853 I explored a part of the course of its principal stream, the Tono.2 Another expedition to explore this region, under the sanction and with the aid of the Peruvian Government, was undertaken by some native adventurers, accompanied by a few Americans, and an English artist named Prendergast, in 1856, but it completely failed. Since that time the wild Chuncho Indians have continued to attack and encroach upon the few farms which existed in these valleys at the time of my visit in 1853, and at the present moment there is not one remaining. The rich valleys of Paucartambo, once covered with flourishing cacao and coca farms, have again become one vast uncultivated tropical forest.

Following the eastern slopes of the Andes to the south and east, we next come to the streams which drain the valleys of Marcapata and Ollachea, but of these very little is known. These valleys are in the province of Quispicanchi, in the department of Cuzco; and it is said that in times past they were cultivated with advantage, and contained many cocafarms. In the beginning of the last century a Jesuit found gold in a hill called Camante, in the Marcapata valley, situated between two ravines, in one of which, called Garrote, a Spanish company established gold-washings. The leading man of this company, named Goyguro, employed hundreds of Indians, and extracted gold from the Camante hill in lumps; but one day an immense landslip fell into the Vilca-mayu,3

2 See Cuzco and Lima, chap. viii.; also Roy. Geo. Soc. Journal for 1855.

3 This is not the great river which flows near Cuzco, and falls into the

Ucayali. The Indians call all rivers which serve as the trunk or centre of a system of streams Huilea or Vileamayu.