with a sheer fall of forty feet. The gold-seekers of the Challuma have penetrated further into the forests, and nearer to the main stream of the Purus, than any other explorers; and their discovery of the Challuma, and of the auriferous hills near its banks, has added something to our geographical knowledge of this region.

The remaining villages on the eastern slopes of the Caravayan Andes are Patambuco, Sandia, Cuyo-cuyo, Quiaca, Sina, and the farm of Saqui, on the frontier of Bolivia. The river of Sandia has one of its sources near the pass twenty miles north-east of Crucero, whence it flows past Sandia, and for many leagues down a narrow gorge, with magnificent mountains rising up abruptly on either side. At a distance of twenty miles below Sandia, in a part of the ravine called Ypara, the coca and coffee plantations commence, at a height of 5000 feet above the sea. Beyond Ypara cultivation ceases, and the river, now increased to double its former size by its junction with the Huari-huari, flows for many leagues between mountains covered from their summits with a dense tropical forest. This region is known as San Juan del Oro, once famous for its gold-washings; and here the royal town of the same name stood, founded by the fugitive Almagristas, and afterwards tenanted by the Señores Mulattos, but long since destroyed and abandoned. The forests contain chinchona-trees of valuable species, and, until the last fourteen years, they were frequented by bark-collectors.

While flowing through the forests of San Juan del Oro the river takes a turn to the westward, and, at a distance of sixty miles from Sandia, enters the Hatun-yunca, or Valle Grande, where the people of Sandia have very extensive

3 Lijera descripcion que hace Juan Bustamante, de su viaje a Carabaya, y del estado actual de sus lavaderos y minerales. Arequipa, 1850. Bustamante says that, at the time of his

visit, there were a hundred people at the lavaderos of the Challuma, and that the Indians received 4 rials a day.

coca and coffee plantations. The curve here made by the river is so considerable that the people from Sandia reach their farms in the Valle Grande by leaving the ravine above Ypara, and making their way across the grass-covered mountains. The coffee-plants in these farms receive no attention whatever from the time they are planted, so that, instead of the dense well-pruned bushes of India or Ceylon, they grow into tall straggling trees about twelve feet high, with a very small harvest of berries on each, but each berry well exposed to the sun. The coffee is certainly ex

cellent.

Passing through the Valle Grande the river flows on past Versailles, where it receives the golden Challuma, and, uniting with all the other rivers of Caravaya, becomes that great Ynambari which finally effects a junction with the Madre de Dios, and forms the main stream of the mighty Purus.

The river Huari-huari, which is formed by two streams flowing from the villages of Sina and Quiaca, joins the river of Sandia about thirty miles below that town, and their united streams compose the Ynambari. Finally the river Tambopata rises near a farm called Saqui, just within the boundary between Peru and Bolivia, at the foot of a ridge of the Eastern Cordillera. After a course of forty miles it receives the river of San Blas, on the banks of which the people of the Sina village have their coca-plantations. Eighty miles lower down the Tambopata unites with the river Pablobamba, on its right bank, at a place called Putina-puncu. The Pablo-bamba rises in a hill called Corpa-ychu on the very frontier of Bolivia, and is only divided from the Tambopata, during its whole course, by a single range of hills. The frontier between the two republics has never been surveyed. Below Putina-puncu the united waters of the two rivers enter the vast forest-covered plains into which the spurs of the

Andes finally subside, and henceforth its course is entirely unknown. I think it probable, however, that the Tambopata finds its way direct to the Purus, without previously uniting with the Ynambari.

The respective distances and populations of the villages of Caravaya are as follows:

But some of these villages are at greater distances from the foot of the Andes than others; thus they are not in a straight line, and the direct distance from Ollachea to the Bolivian frontier is a good deal under 180 miles. The valleys in which the Caravaya villages are situated are separated from each other by spurs of the Andes, many of them so wild and precipitous as to be quite inaccessible; and there is no means of passing from village to village, in many instances, without crossing the Andes to Crucero or Macusani, and descending again by another pass. For this reason Crucero, being in the most central position, has been chosen as the site of the capital of the province, though in a bleak and intensely cold region.

The geological formation of Caravaya is composed of non

fossiliferous schists, micaceous and slightly ferruginous, with veins of quartz. It is a portion of the extensive system of rocks which Mr. Forbes has grouped together as belonging to the Silurian epoch, and which extends almost continuously over an extent from north-west to south-east of more than seven hundred miles, forming the mountain-chain of the Eastern Andes, continuous from Cuzco, through Caravaya, to Bolivia. These rocks throw off spurs along the eastern side of the main chain. Of this formation, too, are the loftiest mountain-peaks in South America :-Illampu, or Sorata (24,812 feet), and Illimani (24,155 feet). Illampu, Mr. Forbes assures us, is fossiliferous up to its very summit.'

Such is a brief account of the geography of Caravaya, and especially of the streams which combine to form the great river Purus, from the rivers of the Paucartambo valley on the extreme north-west, to the Pablo-bamba on the frontier of Bolivia. The streams flowing from the Eastern Andes to the north-west of the Paucartambo system combine to swell the Ucayali, while those to the south-east of the Pablo-bamba fall into the Beni, one of the chief tributaries of the Madeira. The intermediate streams are the sources of the unknown Purus, they are all more or less auriferous, they flow through forests abounding in valuable products, and through countries of inexhaustible capabilities. Yet the courses of very few of them have been explored to distances of seventy miles from their sources, and the main stream of the Purus, one of the principal affluents of the Amazon, may be said to be entirely unknown to geographers.

4 On the Geology of Bolivia and Southern Peru, by David Forbes, Esq., in the Journal of the Geological Society for Feb. 1861, p. 53.

Mr. Forbes had, of course, personally examined only a portion of this great Silurian region. At Tipuani, in Bolivia, there is a very rich auriferous country, composed of blue-clay slates, with no fossils; while the beds near

Sorata contain fossils, and consist of blue-clay shales, micaceous slates, grauwacke, and clay slates, with goldbearing quartz, metallic bismuths, ironore, and argentiferous galena. "The whole of this Silurian formation is eminently auriferous, and contains everywhere frequent veins of auriferous quartz, usually associated with iron pyrites."

CHAPTER XIII

CARAVAYA.-THE VALLEY OF SANDIA.

On the 18th of April I left Crucero, on my way to the chinchona forests, rather late in the afternoon, accompanied by Mr. Weir the gardener, a young mestizo named Pablo Sevallos, and two cargo-mules. After a ride of three leagues along the bleak plain of Crucero, covered with coarse Stipa and stunted Cacti, we reached a little shepherd's hut, called Choclari-piña, at dusk. It was built of loose stones, with a sheepskin hung across the doorway, but with no plaster or mud between the interstices of the stones, so that the piercingly cold wind blew right through the hut.' The poor Indian family were kind and hospitable, and gave us plenty of fresh milk. Next morning we continued the journey along the same plain, with the snowy peaks of the Caravayan Andes on the left, and the glorious nevada of Ananea ahead, whence rise the rivers of Azangaro flowing into lake Titicaca, and of Ynambari finding its way to the Atlantic. A ride of twelve miles brought us to a hut called Acco-kunka (neck of sand), at the foot of long ridges of dark-coloured cliffs, with huge boulders of rock scattered over the sides of the hills. A hard white frost covered the ground.

of

At Acco-kunka I met a red-faced man, about fifty years age, who gave his name as Don Manuel Martel. He said that he had been a colonel, and had suffered persecution for being faithful to his party; that he had lost much money

The thermometer was at 25° Fahr. inside the hut.