CHAPTER V

PATAGONIA AND SHEEP-REARING

The arid tableland and the region of glacial lakes--The first settlements on the Patagonian coast and the indigenous population--Extensive breeding--The use of pasture on the lands of the Rio Negro--Transhumation.

The northern limit of the Patagonian region passes to the north of the Colorado, in the latitude of the Cerro Payen and of the ridge which leads from Malarüe to the Rio Grande in the sub-Andean zone (36° S. lat)., and to the Sierra de Lihuel Calel in the southern part of the Pampa province. South of this line, from the Andes to the Atlantic, on the territory of the Neuquen, the Rio Negro, the Chubut, and the Santa Cruz, is the region of the sheep farms, their refuge since more profitable branches of farming have driven the sheep from the Pampa. The extensive breeding practised on these poor lands is not profitable enough to justify much expenditure, and is therefore all the more controlled by the physical conditions. It is true that cattle-breeding was once undertaken in the Spanish settlements of the lower Negro, and still exists in western Patagonia at the foot of the Andes, but one never finds there the particular combination of cattle-breeding and sheep-breeding which is characteristic of the Pampean region, in which the main function of the cattle is to improve the pasture and make it ready for sheep.

The climate is trying. The west winds are violent during the greater part of the year, especially on the coast, and merely relax a little in the winter. The mean temperature on the Atlantic coast falls nearly one degree for each degree of latitude (14.6° at San Antonio, below 41° S. lat.; 8.5° at Santa Cruz, below 50° S. lat.; and 5.3° at Ushuaia, below 55° S. lat.). The summer temperature falls even more steeply, but the difference is less notable in winter (21.4° at San Antonio, 14° at Santa Cruz, and 9.2° at Ushuaia). The low summer temperature does not allow cereals to ripen south of the Chubut. In the sub-Andean valleys the summer is comparatively warm (16° in January at Diez y seis de Octubre at a height of 1,800 feet), but there is severe frost, especially at the beginning of the winter, and no month of the year is quite free from it.

Rain is plentiful in the Cordillera, and on its western border: 800 millimetres at Junin, nearly two metres at San Martin (which the wet westerly winds reach by the gap of Lake Lacar), and nearly a metre at Bariloche, on Lake Nahuel Huapi. It diminishes rapidly, however, as soon as one leaves the mountainous region and goes further east over the tableland. The whole tableland has a rainfall of less than 200 millimetres (Las Lajas 180, Limay 150, San Antonio 180, Santa Cruz 135). It is only south of the Rio de Santa Cruz that the rainfall rises once more (Gallegos 400 millimetres, Ushuaia 500 millimetres). Hence Patagonia as a whole is, with the exception of a narrow belt at the foot of the Andes, a semi-arid region with a sub-desert climate. In the Patagonian Andes the rain falls, as on the coast of Chile, mainly in winter. Between Mendoza, which has the summer-rain feature of central and tropical Argentina, and Chosmalal, in the Neuquen Andes, the contrast is absolute. The summer months there (January and February) are dry, and the rain is confined to the winter months, from May to August. It is the same further south, at Bariloche and at Diez y seis de Octubre. On the Atlantic coast the winter-rain feature is less regular and uniform. At San Antonio the heaviest rains fall in autumn (April and May). There is a secondary maximum in August, and a few more showers in the spring (September and October). South of San Antonio the winter maximum, which is always marked, is cut by a short dry period (July and August at Camerones, June at Deseado and Santa Cruz).[50] In the interior, on the other hand, the winter-rain system remains unchanged. The predominance of the precipitations of the cold season is of great importance to the breeders. As a rule, they come down in the form of snow, which melts slowly, and the small quantity of moisture is at least all absorbed in the soil. South of the Santa Cruz the humidity increases, but the rainy season alters. At Gallegos the wettest month is December; at Ushuaia, the rains last from September to March. The snow-season (May-August) is the dry season, and the snowfalls are not heavy enough to interfere with breeding.

[50] This anomaly is doubtless due to the proximity of the sea and the respite of the westerly winds in winter. The coast, with its cold waters and the land-winds causing the deeper water to rise, has a special climate of fogs and mists. These, which remind us of the _garuas_ of the coast of Peru, do not penetrate into the interior.

The surface of the Patagonian tableland is very uneven, though it bears traces of having been much worn by the agencies of its desert climate, which seems to have lasted through the whole Tertiary Era. Going up the Rio Negro, one sees the grey sandstones and Tertiary tufas which form the cliffs, on both sides of the lower valley. They give place higher up to the variegated marls and red sandstones of the Cretaceous which form the tableland at the foot of the first Andean chains. The core of ancient granites and porphyries crops up at places from under the mantle of Cretaceous and Tertiary sandstones. The horizon of the peneplain passes from the Tertiary and Cretaceous tableland to level masses of crystalline rock, the contour of which has been almost entirely effaced. Volcanic eruptions have occurred until quite recent times, and so eruptive areas are the salient features of the tableland, at Añecon and at Somuncurra, south of the district of the Rio Negro, in the ridge on the left bank of the middle Senguerr, in the Chubut province. The basalts have spread out in sheets, the surface of which seems to have cooled not long ago. Basalt flows are found as far as northern Patagonia, south of Valcheta and Maquinchao; but their chief seat is in eastern Patagonia. They cover the inhospitable tablelands to the east of Lakes Buenos Aires and Pueyrredon. The Rio Chico and the Santa Cruz cross them for the upper two-thirds of their course, South of Coile and Gallegos they spread almost to the coast, and the Tertiary Pampas in this part are dominated by an archipelago of small volcanic cones.

The tableland is crossed from west to east by deep and broad valleys, enclosed between high cliffs, often strangled by ridges of basaltic or crystalline rock, and very little ramified. The ravines (_cañadones_), which make breaches in their cliffs on both sides, go only a little way into the sandstone Pampa or the lava tableland. Only a certain number of these valleys are occupied by important rivers (the Rio Negro and the Santa Cruz, for instance) which are born in the Andes, but receive little addition from the light rains of eastern Patagonia. Most of the valleys have only intermittent streams (Sheuen, Coile) or are altogether dry and sown with salt lakes (Deseado). The west wind is now the ruler of this network of fossil valleys. It carves their slopes, and brings into them sand, with which it makes dunes.

We must not confuse with these dead valleys the long depressions, with no outlet, which are scattered over the granite and sandstone tableland (_bajos_, _valles_, _cuencas_). Some have obstinately, but wrongly, sought in these the traces of rivers that have disappeared; and the _bajos_ of Gualicho and Valcheta have wrongly been regarded as the former bed of the Rio Negro and the Limay. Erosion by wind seems to have had something to do with these depressions. Their persistence, at all events, is one of the effects of the aridity which prevents normal erosion from moulding the surface of the tableland. The chief of them are centres for collecting running water. There is a group of valleys all round them, and alluvial beds accumulate in them.

The climate determines the character of the soil in Patagonia. The rounded pebbles of granite and eruptive rock, so often described since the time of Darwin, sometimes free and sometimes embedded in red sand or limestone,[51] are spread over the tableland like aureoles round the masses of rock, and they are particularly abundant in the coast region. On the Rio Negro they seem to be confined to the vicinity of the valley; they disappear as one goes away from it. The progressive reduction in the volume of the Rio Negro gravels, as one goes downward, has been observed to begin in the Andean zone, and it is from the Andes that they come. South of Santa Cruz, in a moister climate, in which the circulation of the water is less localized, the bed is more continuous, and it covers the Tertiary sandstones and clays. It is of fluvio-glacial origin, and comes from the destruction of the old moraines, before the excavation of the actual valleys. But it is the wind that explains the concentration of the gravel at the surface. It separates the pebbles from the more mobile material about them. Wherever the outcrop-strata contain pebbles, the wind eventually converts the place into a field of shingle. It has done this with the terraces of the Limay. The Tertiary marine deposits of the coast region also are rich in pebbles torn from the rocky promontories of the shore; hence the extent of stony soils in the coast region. The wind similarly strips naked the angular stones, of local origin and incompletely worn, round the isolated rocks of the desert tableland or on the flanks of the secondary ravines.

[51] The calcareous flag-stone of La Tosca, which is characteristic of the south-west province of the plain of the Pampa, stretches in the south as far as the Rio Negro in the coast-district. On the other hand, it is almost entirely absent a hundred miles to the west, between the Colorado and the Rio Negro, along the line of the railway from Fortin Uno to Choele Choel.

On the other hand, the bedding action of the wind creates deposits consisting of small and uniform elements from the sands of the dunes to the finest dust. The lightest particles, caught up repeatedly by the squalls and carried to a great height in the atmosphere, go beyond the Patagonian region and reach the bottom of the Atlantic or the plain of the Pampa. Some of this, however, is deposited in the depressions of the tableland, where the moisture fixes it and prevents the wind from regaining it. These æolian deposits in the depressions, a dark-grey clay, which hardens when it is dry, but is softened by water, form two entirely different kinds of soil. If the depression is closed in, or if the circulation of the water is too slight, there is a concentration of the mineral salts; this is the _salitral_, either naked or sustaining a halophytic vegetation, which the saline efflorescences cover with a white coat during the dry season. If on the other hand, the underground waters have a free course, the æolian clay forms the _mallin_. Bushes and fine grasses grow on it, and, as they decay, gradually give it a darker shade and modify its composition. The soil above the _mallin_ is rich in organic elements. It covers the bottom of the valleys between low terraces, covered with faceted pebbles, and dominated by the vertical cliffs of tufa and lava. The contrast between the verdure of the _mallin_ and the arid, dusty, yellow steppe of the tableland is one of the most characteristic features of Patagonian scenery. The area in which _mallin_ has been formed coincides with the most humid districts in the vicinity of the Andes and round the higher hills. On the road that runs along the right bank of the Limay, at some distance from the river, on the surface of the tableland, the limit between the country of the _salitrales_ and that of the _mallinas_ passes between Tricaco and Chasico, a hundred miles south-east of Neuquen; it almost tallies with the curve of a 200 millimetres rainfall.[52] Though the word _mallin_ is not used at Santa Cruz, similar æolian soils are found in the western part of the tableland up to this latitude. Further south glacial deposits, clays with moraine-blocks, fill the valleys, and from Gallegos onward, cover the greater part of the tableland.

[52] G. Rovereto, "Studi di geomorfologia argentina: la valle del Rio Negro," _Bull. Soc. Geol. Ital._, xxxi. 1912, pp. 101-142 and 181-237.

On the eruptive flows of recent date the rock is naked. The wind carries away the products of its decomposition, and the dust accumulates only in the fissures. Traffic is difficult, sometimes impossible.

* * * * *

Toward the west the tableland is separated from the Cordillera by a longitudinal depression, though the continuity of this has been exaggerated. This depression, which outlines the contact between the folded zone of the Andes and the flat zone of the tableland, is very important from the point of view of colonization. Just at the frontier of the steppe and the forest, it is the most hospitable part of Patagonia, the richest in natural resources. Amidst the glacial lacustrine deposits which are accumulated on it there rise masses of different kinds of rock which break it up into compartments, granitic ridges of laccolites exposed to view, eruptive structures that have been dismantled. In the south the sub-Andean depression forms a broad passage between Lake Maravilla and Puntá Arenas, about two hundred miles long, enclosed between the basalt cliffs of the tableland on the east and the mountains of the Brunswick Peninsula and William IV Land. The bottom of it is a singular glacial landscape, sown with lagoons, punctuated by scattered hills, with an impermeable soil of drift and mud. From Lake Argentina to Lake Buenos Aires the elevated tablelands, which rise to a height of 5,000 feet, back upon the Cordillera, and the sub-Andean depression is interrupted. Similarly, between Lake Buenos Aires and Lake General Paz the contour of the Patagonian tableland is not very marked above the sub-Andean zone. The glacial alluvia at the foot of the Cordillera rise to the level of the tableland, which sinks steadily eastward toward the Genua and the Senguerr. To the north, between Carrenleufu and Lake Nahuel Huapi, the retreat of the lakes has left long narrow beds right in the Cordillera, such as the Valle Nuevo del Bolson, the bed of which has been taken over by the Futaleufu west of the Cerro Situación. Further east the topographical features of the edge of the tableland (the valleys of the Chubut, Tecka, and Norquineo) lie from north to south. Hence within a space of little more than a hundred kilometres the sub-Andean zone has a series of parallel roads, communicating with each other by means of broad, transverse gaps, which at one time were occupied by the lower lobes of the glaciers. The sub-Andean depression does not go north of Lake Nahuel Huapi.

The morphological features of the Patagonian Andes begin at 36° S. lat.[53] The edge of the Cordillera, in the Malargüe depression, below 35° S. lat., still presents the typical scenery of the central Andes. The dejection-cone of the Atuel resembles that of the Mendoza. The fringe of torrential deposits, distributed in cones over which the waters spread, is due to the rapidity of the disintegration of the rocks in a desert climate. Keidel has pointed out the part played by the summer rains in transporting mobile elements, which the water drops as soon as the slope diminishes; the amount of precipitation being too slight to permit the formation and spread over the plain of a regular network of streams. From the Rio Grande onward the dejection cones disappear. The streams tend to become permanent, and sink into narrow valleys. The summer rains cease, and the water produced by the melting of the snows has only a feeble capacity for transporting stuff. The soil of the Cordillera is protected by a denser vegetation. The first thickets of _molle_ appear in the valleys, the first scattered cypresses on the slopes, at the Rio Agrio, a tributary of the Neuquen. Then the forest invades the mountain: at first, from 38° S. lat. to 39° 30' S. lat., a resinous forest of araucarias. At length, at Lake Nahuel Huapi, the forest assumes the general appearance which it has as far as the Magellan region. It is chiefly made up of different kinds of beeches. The _coihue_ (_Notofagus dombeyi_) is the most conspicuous for about three quarters of a mile, rising above an impenetrable undergrowth of bamboo. Higher up the domain of the _lenga_ (_Notofagus pumilio_) extends as far as the fringe of the Alpine forests. The forest does not reach the eastern limit of the lakes. In the sub-Andean depression it is reduced to thickets of _ñirre_ (_Notofagus antarctica_) and _mayten_ and clumps of _calafate_ (something like myrtles).

[53] The great mass of the Patagonian Andes differs considerably in geological structure from the Argentinian Andes. The Paleozoic sedimentary rocks and the lofty chains of the pre-Cordillera cease at 36° S. lat. The Mesozoic beds--variegated breccie and porphyritic conglomerates, sandstones, limestones, and marls--which form the western slope of the Andes in central Chile, pass to the eastern slope at 35° S. lat., where they develop in regular folds, in the direction south-south-east, obliquely to the general line of the range. These folds account for the orientation of the interior valleys, which is remarkably uniform from the Rio Negro to the Collon Cura. They pass in the south-west under the sandstones of the tableland. West of this sedimentary zone, the zone of the sub-Andean granites and diorites, which have not been exposed further north except at the base of the western slope, opens out in the Patagonian Andes, of which it is the main body between Lake Lacar and the Gulf of Ultima Esperanza. In fine, the Patagonian Andes are characterised by volcanic formations. They are seen on the eastern slope about 36° S. lat., in the lava-flows and ashes of Payen and Tromen. Further south volcanoes with acid lava and characteristic cones are restricted to the central zone (Lanin, etc.) and the Chilean flank, but flows of fluid basic lava cover enormous stretches at the eastern fringe of the Andes, and they have spread over a good deal of the Patagonian tableland outside the Andean region.

It is on the Alumine, about 39° S. lat., that we find traces of glacial erosion, as they spread over the landscape. At present there is no ice on the mountain except on the peaks of Lanin and Tronador, but from the Rio Puelo onward (42° S. lat.), glaciers clothe all the summits which rise above 6,500 feet. North of the Aisen they form a narrow, but almost continuous, line. From the Aisen to the Calen fiord, and beyond the gap of the fiord as far as 52° S. lat., the ice spreads in a considerable sheet which in some places attains a breadth of eighty miles. The tongues of the glaciers reach the Pacific below 46° S. lat., and Lake San Martin on the Argentine slope below 49° S. lat. In Tierra del Fuego the snow-line is at 2,300 feet, and the glaciers which the snows feed, reach as far as the fiords and Lake Fagnano.

Lake Carri Lauquen, on the Barrancas (36° 20' S. lat.), which was almost entirely drained in 1914 through the breaking down of the natural dam of soft earth which confined its waters, is not a glacial lake.[54] The chain of glacial lakes stretches from the Alumine to the Seno de la Ultima Esperanza, and is continued southward by Skyring Water, Otway Water, and Useless Bay--genuine lakes in communication with the Pacific by means of narrow channels. The lakes sometimes lie in a narrow and deep glacial valley, the bottom of which they fill; sometimes they branch out into the neighbouring valleys; at other times they advance eastward beyond the zone of the mountains and spread into round basins surrounded by circles of moraines. The largest of them include groups of ramified fiords, which represent their western half, while the eastern half spreads between lower banks.[55]

[54] Pablo Groeber, _Informe sobra las causas que han producido las crecientes del Rio Colarado en_ 1914. Dir. Gen. de Minas, Geol. e Hidrol., Bol. No. 11, series B, Geologia (Buenos Aires, 1916).

[55] Most of the lacustrine depressions are continued eastward across the Patagonian tableland in the shape of distinct valleys. The eastern part of the Straits of Magellan is merely a submerged valley on the axis of Otway Water. Useless Bay also is continued eastward by the hollow which ends in the Bay of San Sebastian. Sometimes the waters of the lakes flow eastward, toward the Atlantic, along these valleys. Generally, however, the lakes of the western slope are drained on the west by means of narrow defiles across the Cordillera, or on the north and south by rivers which follow the sub-Andean depression and thread them together in the manner of a rosary. The valley which joins the lake to the Atlantic is in those cases a dead valley, and the inter-oceanic dividing line of the waters is marked by the frontal moraine of the old glacier, which confines the lake on the east. This arrangement is found, with surprising regularity, from the Alumine and the Lacar to the Neuquen, and as far as Lake Buenos Aires and the Seno de la Ultima Esperanza at Santa Cruz. The capture of the waters of the eastern slope by the rivers of the Pacific across the Cordillera is fairly ancient, and certainly pre-glacial. But during the Glacial Period the glaciers obstructed the transverse valleys of the Cordillera, and the waters of the eastern slope found their way to the Atlantic once more. With the retreat of the glaciers the valleys of the Cordillera were successively cleared. The lakes, dammed by the glaciers, were suddenly released and their level lowered. The valleys of the Patagonian tableland were finally abandoned, and the topographical accident of secondary importance, which the ancient frontal moraine of the glacier represents, came to mark the limit of the domain of the Pacific. The freshness of the contours of the dead valleys of Patagonia bears witness to the recent date of this conquest, which was too sudden or rapid to be called a "capture" in the proper sense. It has not been accomplished everywhere. From Lake San Martin to Lake Buenos Aires all the lakes of the eastern slope are drained into the Pacific by rivers which flow into the Culen fiord. But further south, Lakes Viedma and Argentino are still tributaries of the Atlantic. They correspond to the zone of the Patagonian Andes which is still covered by inland ice. To the north, in the basin of the Puelo and the Yelcho, where the trans-Andean valleys long ago ceased to be obstructed by ice, the lakes of the eastern slope which drain toward the Pacific are small in size. Their level to-day is much lower than it used to be, and a network of streams has developed east of them, on the earlier lacustrine region, which is now dry.

Pastoral colonization has now spread over almost the entire surface of Patagonia. The parts that are not yet occupied are of slight extent; they consist only of the most desolate regions in the south of the Rio Negro district and north of Santa Cruz. The expansion of white colonization began only about 1880. Until then the interior was abandoned to the indigenous tribes and was almost entirely unknown. The Atlantic coast alone had been explored. The travels of Villarino along the Rio Negro and the Limay as far as Lake Nahuel Huapi had left only a faded memory.[56] North of the Rio Negro, Woodbine Parish (1859), making use of the notes left by Cruz, who had crossed the Andes and the Indian territory between Antuco and Melincue in 1806, was the first to publish definite information, to which no addition would be made during the next forty years.[57]

[56] _Diario de D. Basilio Villarino del reconocimiento que hizo del Rio Negro en el año de_ 1782 (Coll. de Angelis, vi).

[57] It is Woodbine Parish who corrects Villarino's mistake in confusing the Neuquen, at its confluence with the Limay, with the Rio Diamante, known in the south of the Mendoza province.

The settlements founded on the coast by the Spaniards at the close of the eighteenth century (S. José and P. Deseado) were ephemeral. Only one of them maintained an obscure existence, Carmen de Patagones, some miles above the mouth of the Rio Negro. One of its chief resources was the export of salt. Expeditions for this purpose began on the Patagonian coast about the middle of the eighteenth century (_Journey from San Martin to Puerto San Julian about_ 1753, Coll. de Angelis, V). After the revolution, Buenos Aires finally abandoned these costly expeditions by land to the salt districts of the Pampa, and was supplied with salt by schooners from Carmen. During the war with Brazil and the blockade of the Rio de la Plata, Carmen, protected by the bar of the Rio Negro, and the Bay of San Blas were the harbours in which Argentine, English and French privateers concealed their prizes and did their repairs after the storms of the Gulf of Santa Catarina. D'Orbigny visited Carmen during this period of equivocal prosperity. One of the most curious effects of the hospitality offered to the privateers was the unloading upon the Patagonian coast of blacks, intended for Brazil, who were taken from the slave-traders. Thus an unforeseen eddy brought to the south of the Pampean region part of the current of the slave-trade intended for the sugar-cane plantations in tropical America. A number of the Carmen ranches had coloured workers at this time.

Breeding, in fact, was just beginning to spread in the neighbourhood of Carmen at the time. The cattle had been brought by land from Buenos Aires, and had multiplied along the coast and the river above Carmen. South of Carmen, at San José, the cattle had run wild after the fort was abandoned. The Carmen herds were estimated, before the revolution, at 40,000 head. They disappeared during the revolutionary period, but were reconstituted immediately afterwards, and even during the war with Brazil there was an active export of hides and salt beef. Carmen profited mainly by trade with the Indians. It lived in terror of them, and had garrisons to give the alarm on the routes by which they could approach. But this state of chronic warfare did not prevent trade. Near Carmen there was a group of peaceful Indians who served as intermediaries with the tribes of the interior, who were jealous and hostile. Guides and interpreters were found in this colony, and through it came the first news of the interior. The traffic with the Indians continued for a long time to be of great use to the colonists. In 1865 the Welsh colony established on the Chubut, which had many difficulties at first, was saved from complete disaster by its trade with the Indians.

The indigenous population comprised two groups: the Tehuelches, or Patagonians proper, men of tall stature, and the Araucans, the Ranqueles, the Pehuenches and the Pampas. There was no fixed geographical limit between them. The Tehuelches lived in southern Patagonia; but the Araucans advanced eastward as far as the Pampas region and southward beyond the Chubut. The Indian population of the valley of the Genua and the Sanguerr, south of the colony of San Martin, comprised in 1880,[58] and still comprises,[59] a mixture of Araucans and Tehuelches. The Araucans were acquainted with agriculture, but, once they had tamed the horse, they became mainly a pastoral and hunting people, like the Tehuelches.

[58] Carlos M. Moyano, "Informe sobre un viaje a traves de la Patagonia," _Bol. Instit. Geog. Argent._, ii. 1881, pp. 1-35.

[59] W. Vallentin, _Chubut_ (Berlin, 1906).

In so far as they were hunters, the Indians of Patagonia were nomadic. The taming of the horse only made it easier for them to shift from place to place, and gave them a greater range. Their nomadism has too often been regarded as an aimless wandering. They had laws, settled by the physical conditions; and we can gather a few of these. They kept away from the coastal districts except in winter; that is the season when the rains provide water-courses there. It has been observed that names of Indian origin are lacking on the coast of Patagonia. The Spanish navigators who landed there during the summer found the country deserted and the camps abandoned. On the other hand, the share of the Indians in giving names is very considerable in the interior, as far as the foot of the Andes. During the summer the Indians approached the mountains, where they found good hunting grounds. In particular they chased the young guanacos in the breeding season, December and January. Popper has indicated similar migrations amongst the Onas of Patagonia; they approach the coast in winter, and leave it in summer, to hunt in the interior.[60] The district of Lake Nahuel Huapi and Collon Cura had some attraction from afar. The forest of araucarias produced seeds (_pinones_) which the Indians went to gather; and they also liked the wild apples which ripened on the former estates of the old Jesuit missions. The clusters of bamboo on the Cordillera provided the lances of the Aucas and Tehuelches.

[60] J. Popper, "Exploracion de la Tierra del Fuego," _Bol. Instit. Geog. Argent._, viii. 1887, pp. 74-93.

Lake Nahuel Huapi is the first stage of the busiest of the routes used by the Indians. It came from the lower Santa Cruz, went up the Rio Chico, and from there northward followed the foot of the Cordillera. D'Orbigny was told about it: "All the Indians who live near the Andes go along the eastern foot of the mountains in their journeys, because they find water there, whereas they would find none if they went by the coast; in that way they travel from the Straits of Magellan to the Rio Negro." The Indian track only left the sub-Andean depression between the Rio Chico and Lake Buenos Aires, in the district where the high basalt _mesetias_ extend as far as the Cordillera, and on the Pampa of the Sanguerr.

From Lake Nahuel Huapi the Indians of the south descended the Limay and the Rio Negro, and reached the island of Choele Choel, some 230 miles above Carmen, where they met the Aucas and Puelches. There they exchanged their _guanacos_ hides for woollen fabrics made by the Aucas. Choele Choel was the only large, purely indigenous market; the whites never visited it. Geographical reasons fixed the site of this market of the nomads. In the latitude of Choele Choel the Rio Negro approaches the Colorado and the archipelago of the Sierras of the southern Pampa, which mark so many stages on the routes from the Pampa to the Andes. To the south the coast-route, less exposed to snow than the sub-Andean track, began from Choele Choel. The Indians followed this to reach the Gulf of San Jorge and the Santa Cruz in winter, during the rainy season. Darwin notes the importance of the site and the ford of Choele Choel. Villarino had suspected it, and had, as early as 1782, pleaded for the building of a fort there. By holding this point, he said, they could prevent the tribes from attacking Buenos Aires, or from approaching the Patagonian coast in the district of San José.[61]

[61] _Informe de D. Basilio Villarino à Fr. de Viedma_, Coll. de Angelis, v.

As far back as we can go, the life of the Indians seems to have been deeply influenced by their relations with the whites. The Aucas brought to Choele Choel, not only the products of their industry, but also objects stolen or bought from the Christians on the Pampa. The report of Musters, who followed a Tehuelche tribe from Santa Cruz to the country of the Manzanas ("land of apples"), shows clearly that the attraction of the Nahuel Huapi region for the Indians was less due to its natural resources than to the presence of the Chilean settlements at Valdivia, from which came across the passes of the Cordillera certain quantities of brandy.

The Indian never took to cattle-breeding. His herd never consisted of more than mares and a few sheep. But trade in stolen cattle quickly became the chief occupation of the tribes. It would, however, be a mistake to imagine that the thievish Indian was merely and always a dreaded enemy of the ranches of Carmen. They sometimes had recourse to his services and profited by his misdeeds. After the Revolution, it was the Indians who helped to fill once more the ranches of the Rio Negro, bringing runaway cattle which had remained in the San José district. Later, Carmen bought the cattle stolen by the Indians at Buenos Aires. From 1823 to 1826 the number of the cattle sold by the Indians to the colonists on the Rio Negro is estimated at 40,000. Hence the breeders of Carmen had, as regards the Indians, alternate periods of armed conflict and complicity.

But Chile was always the great market for stolen cattle. Raids (_malones_) and the crossing of the Cordillera by convoys began in the eighteenth century, and continued throughout the nineteenth, until 1880, when the consolidation of Argentine authority on the eastern side gave a more regular form to the cattle-trade. The convoys came to a halt at Antuco and Chillan from which the Chilean buyers sometimes accompanied the Indian tribes as far as the _tolderias_ on the edge of the Pampa. The trade in stolen cattle made use of all the passes of the Cordillera, from the Planchon pass below 35° S. lat., which Roca had covered in 1877 by the fortress of Alamito, to the source of the Bio Bio. The one most used was the Pichachen or the Antuco pass. On the tableland the cattle-tracks formed a regular network with innumerable strands, spreading over a width of about two hundred miles. The most northern route started east of the Poitague district and, after fording the Salado and the Atuel, and passing the _aguadas_ of Cochico and Ranquilco, entered the Cordillera at the bend of the Rio Grande. Another track ascended the Colorado and then reached the high valley of Neuquen. A third crossed from the Colorado to the Rio Negro, and, above the confluence of the Limay, to the Rio Agrio or the Alumine.

The first exact information about the range of the Patagonian Indians is supplied by a group of bold travellers who followed their tracks from 1870 to 1880: Musters, Moreno, Moyano, Ramon Lista, etc. Their discoveries had already outlined the geographical survey of Patagonia when the campaign of 1879-1883 opened it to colonization.

The story of white colonization since 1880 shows us several distinct streams of population. The first, starting from the region of the Pampa, went from north to south along the Atlantic coast, and gradually extended its sphere toward the interior. The breeders used the sea-route, the ancient Indian track with recognized sources of water, to convey their first herds. In 1884, the only spot inhabited on the coast between the Rio Negro and the Deseado was the Welsh colony on the Chubut. In 1886 Fontana reports ranches in the Puntá Delfin district, south of the Chubut.[62] About 1890 the whole district round the Gulf of San Jorge was occupied; and a little later the stream from the north met the stream from the south about San Julian and Santa Cruz. The expansion of colonization was less rapid in the interior. Ambrosetti tells us of the establishment of the first ranches round the Sierra de Lihuel Calel in 1893,[63] and at the same time Siemiradzki still found few traces of colonization on the Colorado.[64]

[62] L. J. Fontana, "Exploracion en la Patagonia austral," _Bol. Instit. Geog. Argent._, vii, 1886, pp. 223-239.

[63] J. B. Ambrosetti, "Viage a la Pampa central," _Bol. Instit. Geog. Argent._, xiv. 1893, pp. 292-368.

[64] J. V. Siemiradzki, _Eine Farschungsreise in Patagonien_, Petermann's _Mitteilungen_, xxxix. 1893, pp. 49-62.

The second stream of colonization came from the Magellan region. It started in Chilean territory, about Puntá Arenas. It was about 1878 that sheep-breeding spread round Puntá Arenas, and between 1885 and 1892 was the most rapid growth of the ranches of the Magellan district. North of the Straits they occupied the lowlands round Skyring Water and Otway Water, then the plateau south of Gallegos. They spread along the Atlantic as far as the Santa Cruz. In 1896 the limit of the sheep-region was on the Santa Cruz about forty miles from the coast.[65] To the west, Puerto Consuelo was founded in 1892, and in 1896 colonization came up against the mountain barrier which the Cerro Payen and the basalt tableland of the Cerro Vizcachas interpose between Lake Argentine and Ultima Esperanza fiord.

[65] J. B. Hatcher, _Reports of the Princeton University expeditions to Patagonia_ 1896-9 (_Narrative of the Expeditions_ and _Geography of Southern Patagonia_, Princeton, 1903).

The spheres of primitive colonization in southern Patagonia on the coast still differ from each other in regard to density of population. But breeders in search of unoccupied land have not hesitated to push beyond. In 1895 and 1900 they passed west of the Gulf of San Jorge toward the basin of the Sanguerr and the Genua, (establishment of the Sarmiento colony, south of Colhuapi, 1897: establishment of San Martin on the Genua 1900). Since 1900 the population has also advanced up the Santa Cruz and the Rio Chico as far as the zone of the Andes, and the lagoon which still existed twenty years ago, between the district of the Sanguerr and that of Lake Argentino, and is easily recognized on the maps of the Frontier Commission, has been almost entirely filled up.

The story of colonization in the northern part of the Patagonian Andes is more complicated. Immediately after the campaign of 1883 the valleys of the Neuquen were invaded by Chilean immigrants, half-breeds of the frontier, who cannot always be easily distinguished from pure Araucans. A certain number of Chilotes, and even Germans from the southern colonies of Chile, were mixed with the half-breeds. This stream of immigration had begun before the conquest. As early as 1881 Host notices that there are at Chosmalal various families of Chilean farmers who held their lands from the Indian _cacique_. During the summer they took care of the migratory herds from the Chilean plain. Once the country was pacified, they grew rapidly in number. It was they who provided the manual labour for the placer miners of the Neuquen, where gold began to be worked in 1890. The area of Chilean colonization extends from the Rio Atuel, where Villanueva found Chilean immigrants in 1884, to the south of Lake Nahuel Huapi, where Chileans were still met by Vallentin in 1906, on the Rio Pico, close to 44° S. lat.[66] South of Nahuel Huapi there is no regularly used route across the Cordillera.[67] The Chilean colonists of the southern zone came from the north, therefore, along the eastern foot of the Andes. Bailey Willis calculated that there were 2,000 Chileans in a total population of 3,500 in the sub-Andean area from Nahuel Huapi to Diez y seis de Octubre. The total number of Chilean immigrants may be about 20,000. It is not on the increase, as immigration from Chile was suspended from 1890 to 1895. Since the reconstruction of the frontier the Chilean Government has tried to bring back part of the emigrants to its own territory. Many have gone to settle in the valley of the Lonquimay. In 1896 Moreno saw traces everywhere in the valley of the Collon Cura of the departure of Chilean colonists who had left the country.

[66] C. Villanueva, "De Mendoza a Narguin," _Bol. Instit. Geog. Argent._, v. 1884, pp. 171-4.

[67] Chilean woodcutters have sometimes got as far as the eastern valleys in search of larch, but these were nomads who did not settle.

At first it was only the Argentinians of the western provinces, San Juan and Mendoza, who vied with the Chileans for the soil. It is they whom Furque found in 1888 at Roca, on the Rio Negro. But beginning with 1890-95, immigrants of various nationalities have settled on the Neuquen and the Negro.[68] Foreign capitalists organized their first ranches there. In 1888, on the other hand, the Welsh of the lower Chubut, led by Indian guides, went from the coast to the sub-Andean region, and settled in the valley of Diez y seis de Octubre. Between 1895 and 1900 the neighbouring valleys began to be inhabited, and the colonization areas of Nahuel Huapi and the Sanguerr came into contact.[69]

[68] Furque, "Descripción del Pueblo General Roca," _Bol. Instit. Geog. Argent._, ix. 1888, pp. 124-132.

[69] In spite of their importance we must regard as mere episodes in the story of Patagonian colonization the influx of population caused on the eastern coast by the discovery of placer-gold at Cape Virgenes and on the Atlantic coast of Tierra del Fuego (1884), and the discovery of petroleum at Rivadavia (1907) in the course of drilling in search of water. Rivadavia is already, with its 3,000 inhabitants, one of the chief centres in Patagonia.

* * * * *

The most striking feature of colonization in Patagonia is the very low density of population. The Census of 1914 gives 81,000 inhabitants altogether for the territories of the Rio Negro, the Neuquen, the Chubut, the Santa Cruz, and Tierra del Fuego. A well-kept ranch of 25,000 square kilometres has only a staff of about a hundred men at the most, counting strangers, settled on its land; three hundred inhabitants, or scarcely more than one to ten square kilometres. This population falls into two distinct classes. One is the class of proprietors with regular titles: a rooted and stable class. At first the Government granted enormous concessions, which were taken up especially by English buyers, but it now seeks to break up the land, and the plots which it puts on the market for new pastoral colonies have not more than 625 hectares. This is too small for breeding, no matter how good the situation may be, and there will inevitably be, one would think, a concentration of estates in the hands of a few proprietors. The other part of the population occupy lands which they do not own. They are displaced steadily as the regular concessions are sold to new ranches. They live, so to say, on the margin of colonization, and are more and more restricted to the poorest lands. Sometimes these _intrusos_ or _pobladores_ get hospitality for their herds on the land of some ranch in return for their services. They have little capital, and never make material improvements. They take no care to nurse the pasture, and it matters little to them if it is impoverished.

The climate divides Patagonia into two distinct regions. In the west, the moist Andean zone is suitable for cattle-breeding. About 1870 the Chileans of Valdivia hunted wild cattle in the Nahuel Huapi district. Similarly the Frontier Commission met large herds of wild cattle on the shores of Lake San Martin, which were not yet occupied. Sheep do not get on well in the moist zone, where the rains have washed out the soil and carried away the salts which seem to be indispensable to the sheep. It is the arid tableland that is the land of the sheep. There it has displaced cattle, even in the area which the early breeders at the end of the eighteenth century had filled with cattle. Between the sheep-area and the cattle-area is a mixed region, where the two are combined. It extends more or less according as the transition from a moist to a desert climate is gradual or sudden. It is especially important in the districts where colonization is already old, as in the Fuegian and Neuquen regions. It is lacking in districts where the colonization is recent (Chubut and Santa Cruz), where the sheep-breeders have had a free run as far as the Andes. The ranches of the Cordillera, which specialize in cattle-breeding, all have small flocks of sheep for their own use, their staff being so small that it does not pay to kill the cattle.

The sheep-area is by far the more extensive of the two. The patches of agricultural colonization are very scattered and small on its surface. They are restricted to the river-oases of the Rio Negro and the Chubut. These small tilled districts have preserved a remarkable economic independence as regards the pastoral zone, in which they seem lost. Thus the farmers on the Chubut exported their wheat to Buenos Aires until about 1900, and they still send their bales of dry lucerne there. Some of the ranches have tilled small oases in suitable places, but these are merely intended to increase their stores of fodder; not for their flock of sheep, but for the saddle-horses used in watching the estate and the draught-horses used for transport.

The pastoral capacity of the Patagonian scrub is, on the average, from 800 to 1,200 head of sheep to 25 square kilometres: less than a tenth that of the prairies of the eastern Pampa. The ranch fixes its residence in the best part of the estate, where there is least fear of a shortage of water, and where pasture is most plentiful. To this the sheep are brought periodically to receive disinfecting baths against the scab, and for shearing. These incessant movements toward the centre of the ranch cause an almost permanent strain on the pasture, and this is one of the chief anxieties of the breeder. The area of the estate is divided as soon as possible into sections (_potreros_) by steel-wire fences, which enables them to watch over the reproduction and improvement of the flock and make the best use of the pasture. Fencing is more advanced near the Cordillera, as timber for the posts is found there.

Certain districts are still uninhabited on account of the lack of water. Some of the sources of water are permanent. The water issues at the base of the volcanic rocks, when the underlying rock is impermeable, and above the various levels of the marl in the Patagonian swamps; for instance, in the _cañadones_ round the Gulf of San Jorge. Besides this, the rain and melting snow leave on the surface of the tableland a great number of pools, which evaporate in the dry season. These are temporary supplies, the _manantiales_, to which the breeders are reduced over large areas of the tableland. Most of the stagnant sheets of water which are permanent are saline. The proportion of salt in them is very variable, and changes in each case according to the cycle of dry and wet years. The water of the Carilaufquen was fresh in 1900, and in 1914 it had become brackish, though it could still be used for the flocks.

Finding permanent sources of water is the first concern of the breeder. In some districts he has succeeded in tapping sheets of fresh water by means of wells. There are none of these wells in the crystalline zones, the closed hollows, where the sheets of water are often large, but they are always saline. Neither are there any in the red sandstone district, the dryest of all. In the western region the wells are sunk in the arid valleys, along the track of the underground stream. Thus the Picun Leufu, the visible course of which is lost seventeen miles above its confluence with the Limay, may be traced by a continuous line of wells. It is especially in the coastal districts that the wells have transformed the conditions of breeding. Water was first discovered at the foot of the dunes, along the coast itself (district of Viedma, San José, etc.). Since then deep borings have been made over the whole of the Tertiary platform on both sides of the lower part of the Rio Negro, north of San Antonio. There every ranch has its sheet-iron tank, sheltered by a clump of tamarinds, with a windmill to fill it.

All pastures are not equally available in every season. Those which are at a height of more than 4,000 feet in the north, and 2,300 to 2,600 feet in the south, are covered in winter with a thick mantle of snow. These are summer pastures. During the winter the animals are brought down to the principal valleys or to sheltered _cañadones_ below the level of the tableland. The _mallin_ is, as a rule, a winter pasture. When it is too wet, however, it is treacherous, and the animals are buried in it. They have to wait for fine weather before going into it. The pastures, too, which have no permanent water supply, or have only _manantiales_, which dry up at the beginning of summer, can only be used during the winter. Hence each ranch has to have, besides its assured water supply, a suitable combination of summer and winter pasturage, and it is far from certain that this will be found on every estate, cut up geometrically for colonization, as they were, by the administration of lands.

The constitution of the flock and the first occupation of the land have compelled breeders to undertake difficult journeys, and more than one of these proved disastrous. The earliest arrivals, driving their sheep along little-known tracks, could not avoid losses in crossing the arid parts of the tableland: parts which D'Orbigny, translating literally the Spanish word _travesia_, calls "crossings."[70] When the ranch is established, the breeding does not necessitate any further movements of the flocks to a great distance, apart from certain special migrations, or "transhumations," which I will consider later. It is on each ranch, sometimes on each group of ranches combined in a single estate, that they pass alternately from winter to summer pasture. The only transport necessary is that of wool. The fleeces, which the west wind has heavily laden with dust, are collected in the sheds belonging to the ranch, or, in the case of the _intrusos_, on the premises of certain small traders (_bolicheros_) who are scattered over the tableland even at its extreme limits. Convoys of wagons then take them to the ports on the coast.

[70] The search for possible routes for cattle in the districts that were not yet colonized helped in the study of Patagonia. Moyano was doing this when he explored the route from Santa Cruz to Lake Nahuel Huapi.

For some years now, however, wool has ceased to be the sole product of the ranches. A little before 1895 the first slaughter-houses, for killing the older sheep that were no longer fertile, were erected on the Straits of Magellan. Refrigerators have succeeded these, and were opened at Puerto Callegos and San Julian. A third refrigerator is being constructed (1915) at Puerto Deseado. In southern Patagonia, also, part of the flock is sent to the refrigerators or to the slaughter houses of the Pampean region. The creation of the refrigerator has compelled breeders to adapt their work to the new economic conditions. The merino breed is being eliminated by the Lincoln in all districts which feel the influence of the refrigerator; the Lincoln is of greater weight and quicker growth, but the merino survives in arid northern Patagonia.

Besides this, the establishment of the refrigerators has caused important movements of transport. The flocks which are to go to the refrigerators or the northern railways are moved in the good season, after the shearing, from November to April. The routes they take are not invariable. One of the most frequented, leading from the sub-Andean tablelands to San Julian, follows the Santa Cruz valley. When the land was cut up, there was no reason to foresee these movements, and nothing was done to facilitate them. The roads cross the ranches, which are compelled to allow it. It is a serious burden for some of them, unless they can make a profit out of their situation on the road by hiring pasture for the flocks as they pass.

The Andean zone itself is still mainly pastoral, but it is nevertheless far more varied and richer in possibilities of development than the tableland. Agriculture is already combined with breeding in that area.

The name _vegas_, which in the Puna and at San Juan means alpine pasture, is applied here to tilled patches in the Andean valleys. They are found in the north in the valley of the Neuquen, round Chosmalal. In the south, the valley of the Rio Pico marks the limit of cultivation. Irrigation is almost always necessary north of Lake Nahuel Huapi, where the _vegas_ have, as a rule, a soil of coarse alluvia or permeable tufa, which dries up quickly. Water is plentiful, it is true, and increases in quantity rapidly as one travels southward. The chief obstacle to the extension of cultivation is the frequency of frost in spring and summer. The deep hollows of the sub-Andean depression south of Lake Nahuel Huapi, the height of which drops to 1,000 feet at the Bolson, and 1,600 feet at Diez y seis de Octubre, have no frosts in summer, and they sustain small agricultural communities. At higher levels, in the basin of the lake or on the _vegas_ of the Traful and Lake Lacar, at an altitude of about 2,600 feet, the distribution of the summer frosts is closely related to the contour and lie of the land, which may facilitate or impede the circulation of the layers of cold air, and the play of what has been called atmospheric drainage. The valleys which are very open from west to east, at the outlet of the lakes, where the west winds have a free passage, are little liable to frost. Wherever frost is frequent, cultivation has to be restricted to fodder plants. The more favoured cantons, which grow wheat, rye and potatoes, help to feed the local pastoral population, and export part of their produce to some distance on the tableland.

Cattle-breeding is, like sheep-breeding on the tableland, practised both by the _pobladores_ on public lands and by ranchers who have settled on regular concessions, which they have worked up and fenced round. The high alpine pastures, above the fringe of the forest, are partly used, from December to March, as summer-pasture. The forest also serves for pasture; it is a sort of common land, available both in winter and summer. Below the height of 3,500 feet the clumps of bamboos in the underwood provide shelter during the winter and fodder which is not buried under snow. The fires lit by the breeders have changed part of the primitive forest into a scrub which has been invaded by a leguminous climbing fodder, and it has superior pastoral capacity to the forest. East of the forest, the prairie, which is too much exposed to the winds, is not generally suitable for winter-pasture. The cattle take refuge in sheltered valleys and in the _mayten_ thickets which follow the depressions. Bailey Willis puts the pastoral capacity of the virgin forest at 400 cattle to each 2,500 hectares, 600 for the burnt forest, and 350 for the sub-Andean prairies. The essential problem in connection with the question of completely developing the pastoral resources of the sub-Andean region is the problem of transit. There are no roads from one district to another and to the higher prairies. The fallen trunks which lie about the forest obstruct the way of the cattle. Collecting the animals for sale and watching them are both difficult.

It seems that the profit of exploiting the timber must necessarily be small. The forest, thinned by fire and difficult of access, is partly composed of trees that are too old. The _libocedrus_ has disappeared from one-third of it. The larch, which is the most valuable, passes into Argentine territory at few places. Saw-mills are not so numerous on the eastern slope of the Andes as they are in the Magellan area.

The essential function of the forest is, according to Argentine experts on forestry, to control the water-circulation. In this land of glacial erosion and recent captures, where the water-courses have always a great variety of form, and there are lakes to make their output more regular, it is particularly easy to make use of hydraulic power. "White coal" will, Bailey Willis says, make a great industrial region of it, and plant an urban life in it. Bailey Willis, whose optimism and prophetic gift will not fail to surprise the European reader, has drawn the plans in detail of a future town of 40,000 souls at the eastern end of Lake Nahuel Huapi. The Patagonian land will supply the raw material of its industries; timber, leather, and wool.

One, at least, of the indispensable conditions of the development of urban life is fully realized in the district of Lake Nahuel Huapi and the Limay. It is a remarkable meeting-place of natural roads, and its economic value will increase in the future. It is the point where the road from eastern Patagonia by the sub-Andean depression, from the Gulf of San Antonio on the Atlantic, and from the Rio Negro by the Limay, and the roads that lead to Chile across the Cordillera, meet. The whole zone of the Andes between 36° S. lat. and 42° S. lat., the latitude of the southern part of the Chilean plain, has numerous and easy passes. There has always been close communication between the two slopes, and people have emigrated freely from one to the other. But north of 39° S. lat. the passes are rarely lower than 5,000 feet. They are covered with snow in the winter, and can be used for traffic only in certain seasons. It is not the same south of the volcano Lanin. That is the beginning of the glacial valleys which go to the heart of the Cordillera, some of them crossing the mountains from east to west. They have not yet been entirely explored. The Bariloche pass, south of the Tronador, by which the Chilean missionaries reached Nahuel Huapi in the eighteenth century, is no longer used. The Cajon Negro pass, west of Lake Traful, through which Bailey Willis traces the line of a southern trans-Andean railway, was only recently discovered, and the valleys which run into it on the Chilean side are not yet well known. The two best-known trans-Andean routes to-day are the Perez Rosales road, which leads from Chile to Nahuel Huapi by the north of the Tronador, and further north, the road from Lake Lacar to San Martin. Both these have received some attention, and the lakes are connected by telegraph or telephone. The frequent need to unload and reload makes the traffic costly, but it is permanent and is not interrupted in winter. The reduction of the export of cattle to Chile has cut down the traffic for a time, but it is sure to recover. The permanent importance of it is one of the facts most clearly written by nature upon the soil of South America.

* * * * *

It is not easy, in the absence of documents, to attempt to give for Patagonia as a whole a detailed description of the pastoral industry, and to follow step by step on the spot its efforts to adjust itself to the natural conditions. But the analysis may be attempted in regard to the region between San Antonio and Lake Nahuel Huapi south of the Rio Negro,[71] the valley of the Rio Negro, and the tableland which stretches westward between the Neuquen and the Limay. This part of Patagonia is now easily accessible, and it is entered by two parallel railways. One starts from San Antonio on the Atlantic, and goes westward to Lake Nahuel Huapi. It has (1914) reached Maquinchao, on the tableland, mid-way across the Andes. The other starts from Bahía Blanca. At Choele Choel it enters the valley of the Rio Negro, and ascends it as far as the confluence of the Neuquen. Then it goes 130 miles westward as far as Zapala, at the foot of the first sub-Andean chains. Each of these lines is ambitious to attract the trans-Andeans. At all events, they are in a hurry to reach the humid zone at the foot of the Andes, which could maintain a busier traffic than the desolate tableland.

[71] This was the area studied by the Commission of which Bailey Willis was chairman.

The railway from San Antonio, and the road which is a continuation of it west of Maquinchao, cover a distance of 320 miles from the Atlantic to the Andes, and cross five distinct regions. The first is the coastal plain, composed of horizontal marine Tertiary sedimentary rocks, both of sand and clay. The plain rises slowly toward the west, and it attains a height of 650 feet at a distance of seventy miles from the coast. This coastal platform divides, on the north-west, the enclosed hollow of the Bajo del Gualicho from the Gulf of San Antonio. Its surface is very even. The gravel on it has formed a sort of conglomerate, and in spite of appearances, this gravelly soil is not bad for vegetation. It quickly absorbs the rain-water, which thus escapes evaporation. The vegetation is comparatively rich. There are no springs, but the autumn rains sustain _manantiales_ in the marly surface, and these do not dry up until the spring. During the summer the plain is deserted, and there is no water. But the flocks return in the winter and remain there until spring. There is very little snow, as the temperature is moderate. In spite of the density of the pastoral population in winter, the pasturage is not injured. The grass grows plentifully amongst the thickets. This is because the flocks leave the district before the season when the grasses flower and reproduce, so the next generation is secured. Part of the flocks which winter on the coastal plain pass the summer in the south-west, on the high basaltic tablelands of Somuncura. However, the whole of the surface of the tableland cannot be used permanently, or during the entire summer. There is plenty of water in spring, when the snows have melted. In the middle of the summer the flocks collect round the permanent springs, and they scatter once more over the mountain pastures during the autumn rains, before they return to the plain.

The second region is that of Valcheta. From Aguada Cecilia to Corral Chico the railway follows for sixty miles the edge of the outpour of lava from the south, which covers the Tertiary clays. In front of the basalt cliff the land dips in the north toward a closed depression, the Bajo de Valcheta, the bottom of which consists of clays impregnated with salt. Tertiary marine strata surround this hollow in the west and north, where they divide it from the Bajo del Gualicho, but here they form only a thin skin which covers the crystalline platform. The line of contact of the basalt and the Tertiary marls is marked by a series of good springs, and these give rise to permanent streams, such as the Arroyo Valcheta and the Nahuel Niyeu. At first they flow in a narrow valley crowned by basalts, with peaty prairies at the bottom, then over Tertiary marls, and, in the latitude of the railways, they pass into a gorge cut through the granites before losing themselves to the north in the _salitral_. A small agricultural oasis is sustained by the waters of the Valcheta.

The site of Valcheta has an exceptional importance in the story of Patagonian colonization. It marks a necessary stage in the Indian track from the Atlantic to Nahuel Huapi, which is now followed by the line of the railway. Musters halted there. The track from Choele Choel, on the Rio Negro, to the southern coast and the Santa Cruz also passed by there. It was so much used, says Ezcurra, that the hoofs of the horses had hollowed it.[72] The Argentine village dates from 1890. At first it lived by supplying fodder to the convoys of wagons which carried the wool. The railway has suppressed this traffic, and the only outlet of the oasis to-day is the small port of San Antonio, where the wool is shipped, and where the district is unsuitable for any kind of cultivation.

[72] Pedro Ezcurra, "Camino indio entre los rios Negro y Chubut: la travesia de Valcheta," _Bol. Instit. Geog. Argent._, xix. 1898, pp. 134-38.

Like the coast region, the Valcheta district seems marked out by its moderate altitude to serve as winter pasture. In point of fact, it is used during the whole year. The springs do not dry up in summer. The streams which flow from the south toward the Bajo de Valcheta are permanent. In addition, a few wells have been bored in the Tertiary strata. Contrary to experience on the coast, therefore, cattle can be kept here during the summer. There is less chance for the grasses to reproduce, and the pasture tends to become impoverished.

The third zone, 130 miles from the coast, is that of the tableland of the Cerros Colorados, where low masses of red granite rise like an archipelago amongst the Tertiary formations deposited in the intervening depressions. In the west its altitude rises from 650 to 1,300 feet. It is one of the poorest parts of the tableland, and the size of the flock is reduced to 600 head to the square league. The naked rock crops up, not covered, as it is further east, by a bed of gravel. In the valleys there is little water, and it lies very deep. There are no periodical removals of the animals. Winter and summer they remain within range of a few poor springs, which are caused by various outcrops of lava of limited extent; and they leave these, and wander over the tableland, only in the rainy season.

Beyond the Cerros Colorados the line rises rapidly, and at Maquinchao it reaches the basin of Lake Carilaufquen. This occupies the bottom of a closed depression, at an altitude of 3,000 feet, dominated on every side by a plateau of lava, toward which, in the south, a number of important valleys run (Nahuel Niyeu, Quetriquile, Maquinchao). These valleys rise in the south in the basalt plateau, at a height of 4,000 and 4,700 feet, and have no running water except at their upper ends. South of Carilaufquen they open upon a broad plain, round which there is a sombre cornice of lava, about 350 feet high. Water has collected on the plain, which consists of alluvial beds redistributed by wind: angular pebbles from the terraces, fine dust from the _mallinas_, and sand from the dunes round the lake.

This region is much better than that of the Cerros Colorados. There are many springs at the base of the lava-flows, on the sides of the valleys, and it has as yet not been necessary to look for the subterranean sheets which accompany some of the valleys. The elevated basin of the Quetriquile, though it is only occupied by _intrusos_, seems to have a particularly high pastoral density, and, I am told, feeds 500,000 sheep. In the western part of the region the spring is late, and there is risk of snow during the lambing season. There are, however, no rams there; the lambs are brought from Maquinchao. This arrangement of special zones for the multiplication of the flock enables them rapidly to improve the breed. Here again there are no removals of the animals to a great distance in order to use the pasture. The vegetation of the valleys suffered from the continuous presence of the flocks during the years of drought before 1914; the reproduction of useful grasses was prevented. There is, however, less danger here than on the Cerros Colorados, because the _mallinas_ are extensive, and they suffice for feeding the sheep during the periods when the _manantiales_ of the tableland dry up, and the animals are confined to the valleys.

The fifth region comprises the high ridge which divides the basin of the Carilaufquen from Nahuel Huapi, the water of which flows northward toward the Limay and southward toward the Chubut: successive eruptions have covered the surface with lava and ash, which at Añecon rise to a height of 6,700 feet. The granite platform which emerges in the north, at the Cerro Aspero and the Quadradito, rises to a height of 4,400 and 4,700 feet, and in some places presents a bold and rejuvenated aspect. The whole has been cut up in all directions by erosion, and it affords comparatively easy means of getting about, which the Indian tracks have followed. Below the higher slopes the valleys deepen into gorges, and these broaden out in the soft tufa and are lost at the cross-streams of lava or the outcrops of the granite. In so varied a land, with such marked differences of altitude, the winter and summer pastures are always close together. Precipitation is more plentiful than at a distance from the Cordillera; the pasturage is richer, and the size of the flock rises to 1,600 sheep to the league. The sheep pass the winter on the lower slopes, where they are sheltered from the winds and the snow. They descend to the _mallin_ when the dry season sets in and makes the soil firm. In summer they go on to the tablelands, where the pastures extend to a height of 5,000 feet.

Bailey Willis, studying the improvements that might be made in the pastoral processes, concluded that the essential point was to use each pasturage in its best season, and establish a carefully considered rotation on the various lands. This system, which alone would enable them to nurse the natural resources of the scrub in the way of plants for fodder, is used to-day in only a small number of districts--in the east, where the flocks winter on the coastal plain and spend the summer on the Somuncura tableland, and in the west, round the Añecon, where the summer and winter pastures are not far from each other. The custom ought to be general. The area which ought to be reserved for winter pasture comprises the coastal plains, the whole of the low-lying district round Valcheta, and the lower part of the valleys to the south of the Carilaufquen. They are less extensive than the available summer pastures, but their capacity could be enlarged by developing the irrigated areas in the Bajo de Valcheta, and sowing lucerne in the _mallinas_ of the basin of the Carilaufquen. The low valleys round the Carilaufquen ought to be reserved for winter pasture. In the summer the sheep would be taken south to the higher-level valleys, which afford permanent pasture. From there they would spread after the melting of the snow, and after the first rains in autumn, over the high tablelands which surround them.

This plan is obstructed in the first place by the actual terms of ownership, which were imprudently fixed before the examination of the country in detail had been concluded. Thus the Maquinchao ranch, in the lower valley, does not own the upper valley with the summer pastures that ought to belong to it. A more serious obstacle is that it is extremely difficult to remove the sheep. It is not merely roads that are wanting, but a water supply at the various stages.[73]

[73] The district of the Rio Negro is not the only part of Patagonia which faces the problem of increasing the winter pasture. Attention has been drawn to the possibility of enlarging the lucerne farms in the district of Colonia Sarmiento, south of Lake Musters, and making this a great wintering area for the Santa Cruz flocks.

Between the railway that runs from San Antonio to Lake Nahuel Huapi and the Rio Negro, there is a desert region about seventy miles in width. Red sandstone predominates in it, and it remains uninhabited. North of this _travesia_ the valley of the Rio Negro opens. Its width between Neuquen and Patagones ranges from five to fifteen miles. Its slope diminishes gradually toward the bottom (from 0.67 to 0.49 per 1,000 above Chelfaro; from 0.45 to 0.29 per 1,000 above Conesa).

The sandstone and marl cliffs which enclose it become gradually lower as one goes downward. They dominate the valley at a height of 650 feet at the confluence of the Neuquen, and are only 100 to 130 feet high at Patagones. At the foot of them are broad terraces cut by dissymetrical ravines, in which the beds of sandstone outcrop on the western slope, exposed to the winds, while the eastern slopes are covered with gravel. On the banks of the river there is a strip about two miles wide with abundant herbaceous vegetation between lines of willows. This is covered by the normal floods. The remainder of the river plain, to the foot of the cliffs, has only a thin scrub, with dunes at intervals. Saline clays here overlie the river gravels. The level of the underground water, which is fed by the river, sinks lower as one goes from the banks toward the cliffs. Few parts of the tableland have so desolate an aspect as the bottom of these great Patagonian valleys, when they have not been transformed by irrigation. The pasturage is poor. At Conesa, however, the valley (_costa_) is used as summer pasture when there is a shortage of water on the surrounding tableland (_planeza_).

The water-supply is good, the volume of the river ranging from 200 to 900 cubic metres a second. Low water lasts from February to April (end of the summer). From May to July the river has sudden and violent floods--an effect of the autumn rains. The curve sinks again in August and September, to rise once more in October and December, when the snow melts on the Andes. The Limay, the upper basin of which contains large, lacustrine sheets, is more regular than the Neuquen, which has very pronounced low-water, as well as dangerous floods in the autumn. The first attempts at irrigation date from 1885, when the canal of the Roca colony was dug. Others were made lower down at a later date. The co-operative groups organized for the administration of the canals have not been quite as successful as might have been expected. The advance of agricultural colonization has been slow. Costly preparatory work is needed to level the ground and organize the drainage, otherwise saline patches form and spread like leprosy at the expense of the cultivable areas. Lastly, the centre of the valley is exposed to floods.[74]

[74] The work now (1914) in hand will reduce the risk of floods, and will enable them to enlarge considerably the extent of the tilled land. The Cuenca Vidal, which opens amongst the sandstone, below the level of the valley, on the tableland to the north of the Neuquen, will be arranged so as to absorb the flood-water, and it will feed a canal which will serve the left bank over an area of 100 miles. The waters of the Limay will be available for the lower valley.

The chief crops are lucerne, cereals, and the vine. All the efforts and hopes of the colonists are now centred upon the vine. It is for the purpose of extending the vineyards that they are endeavouring to secure more workers. These are a singularly mixed lot, Chileans from the Neuquen rubbing shoulders with Latin immigrants (Italian and Spanish) from the region of the Pampas.

The lucerne is made up in bales and exported by rail to Bahía Blanca and Buenos Aires. The economic life of the agricultural oasis of the Rio Negro is no more connected with that of the pastoral tableland than is life on the Chubut. Neither sheep nor cattle are fattened on the Rio Negro. It is a curious contrast to the spectacle offered by the Andean regions of western and north-western Argentina, where for generations there has been a close association between the breeding industry of the scrub and the fattening on the lucerne-farms. This is because the currents of the cattle-trade are not here as permanent and stable as they are in the north. The time when the convoys of Pampean cattle bound for Chile used the valley of the Rio Negro preceded the agricultural colonization of the banks of the river. The conquest of Patagonia put an end to this traffic. There was an interval of twenty-five years between the period of the export of Pampean cattle to Chile and the export of cattle from the Neuquen to Buenos Aires, to which I will refer presently. As to sheep-breeding, it did not for a long time rear the animals for the meat-market, and it is only a few years since it found transport necessary. The farmers of the Rio Negro, who have little capital, and who sell and are paid in advance for their dry fodder, have not yet been able to take advantage of the reorganization of the cattle-trade.

West of the confluence of the Neuquen and the Limay the railway ascends the sandstone tableland, from 1,700 to 3,000 feet high, and goes as far as the foot of the first sub-Andean chain, the Zapala ranch. The eruptive rocks here have thrown up the sandstone, and the profiles raised north and south of Zapala, across the Sierra de la Vaca Muerta and the Cerro Lotena, cut through folds of Mesozoic strata which have been reduced by erosion to the level of the plateau. One already feels the vicinity of the Cordillera. Pasture is plentiful, the _mallin_ is thick, and springs abound. The sheep-area stretches westward of Zapala, as far as the Rio Cataluin and the Rio Agrio. East of Zapala, on the other hand, the desolate condition of the country gets worse and worse. The supplies of water dry up in the summer, and the entire zone that lies east of 70° W. long. is useless, on account of the lack of permanent water, except as winter commonage. Hence, transhumation is here indispensable. It has been practised for a long time on the Chilean slope of the Cordillera from the latitude of Coquimbo and San Juan to the north of Lake Quillen. At present it tends to disappear from the Andes of the Neuquen.[75] But there is still transhumation on the Argentine side. The sheep of the plateau, driven from their winter pasture when the water dries up, ascend the Cordillera. Sometimes the mountains are not yet free from snow. In that case the journey is delayed, and the sheep feed on the way, to the great detriment of the land they cross.

[75] As a matter of fact, of recent years there has been a practice on this slope of disguising the smuggling of animals under the name of "transhumation," as the removal of the sheep facilitated it and helped to maintain it. The shepherds got certificates exaggerating the number of their sheep from the Chilean officials before they crossed the frontier, and under cover of these they came back to Chile with additions to their flocks which they had bought on Argentine territory.

There are many routes, and frequently they coincide with those which were formerly taken by the cattle of the Pampas in ascending to the passes of the Cordillera. Groeber mentions a transhumation track south of the Rio Barrancas and Lake Carri Lauquen. From the left bank of the Neuquen the flocks ascend by Chosmalal and Butamallin to the pasture of the Pichachen pass, or by Las Lajas to the Pino Hachado pass. From Zapala and the tableland further south they go to spend the summer in the Cataluin Cordillera, where the number of sheep in summer is calculated to be 70,000. Others go still further, to the source of the Alumine and the Arco pass. The volcano Lanin almost marks the southern limit of the zone of transhumation. The chief group of migrating sheep comes from the district of the Coyunco, the Cañadon Grande, and the Picun Leucu.

Transhumation is practised only by the _intrusos_. They go from the unowned lands of the tableland to the unowned lands of the Cordillera. The renting of winter pasture to owners is quite exceptional. The concessions of land granted by the Argentine Government are steadily reducing the area of the migrators in the Cordillera, and also the ways of communication between the tableland and the mountains. The proprietors do not care to receive the migrating flocks, and they put obstacles in their way by enclosing the land. The routes of the transhumation are now fixed by the spaces which remain open between the enclosed ranches. Moreover, the migrating _intrusos_ are haunted by the fear of finding the winter pasture occupied by others during their absence, and they have no proprietary title. The splitting up of the land and the organization of ownership will before long lead to the extinction of the practice of transhumation, and the greater part of the winter pasturage will be turned into permanent pasture by boring wells and nursing the water-supply.

The district round the Zapala ranch has become very busy since the construction of the railway, which has deeply affected the conditions of life there. It has made a sort of capital of Zapala. It is curious to contrast the renaissance which has followed upon the appearance of the railway in this district with the much less material changes which it has made at Maquinchao. The life which the railway concentrates at Zapala includes not only the wool trade, as at Maquinchao, but also the cattle trade. The herds which are to be exported gather round the ranch at the same time as the _tropas_ of wagons, and a good price is paid for the right of pasturage. While the Maquinchao line ends at the port of San Antonio, which is merely fitted up for the export of wool, the Zapala railway feeds the refrigerator at Bahía Blanca. It joins up with the network of railways of the Pampa. Sheep arrive at Zapala, not only from the surrounding district and from the Neuquen, but from a good part of the Rio Negro, and even the Chubut. The convoys of animals coming from the south find it best to keep near the Cordillera, where the pasturage is better. Only a few of them descend the Limay as far as Senillosa. From Zapala to Senillosa there is no suitable road in connection with the railway, and further east it is necessary to go as far as Choele Choel to find tracks which lead to it. The exporting of the sheep lasts five months, from November to March.

Zapala station is also a point of convergence of herds of cattle. There are people at Zapala who still remember the time when the cattle brought from the Pampa to go to Chile passed through their valley. Although these exports of Pampean cattle to Chile ceased after 1885, the whole Andean region of the Neuquen still lived entirely on the Chilean market until very recently. The attraction of the Chilean market is one of the reasons for the survival of transhumation. It was to the advantage of the Argentine breeders to keep near the Cordillera and the passes through which the buyers came from Chile in the summer. The life of the small centres in the upper valleys which developed rapidly after the conquest (Chosmalal, Ñorquin, Codihue, Junin, and San Martin) was bound up with the Chilean cattle trade, and was reflected on the opposite side of the Andes in the prosperity of the corresponding markets in Chile.

In the years immediately preceding 1914, a sudden revolution upset the cattle traffic on the Neuquen, and the attraction of Buenos Aires took the place of that of the Chilean market. The commercial influence of Buenos Aires was first felt in the wool-market. The _tropas_ of wagons which brought wool to Zapala loaded up, in exchange, with the flour and salt that were needed for sheep-breeding in the pastures of the Cordillera (_pastos dulces_). The import trade followed the path traced by the export trade. The small Chilean wagons which still cross the Cordillera now only bring to the Neuquen the coarse flour of Chile, haricot beans, and wine. They return empty to Chile. After the wool-buyers, the cattle-merchants of Buenos Aires next found their way to the Cordillera. The centres where the sales of cattle for Chile used to be held are now in decay, and have lost part of their population. The cattle are sent to the fattening centres on the Pampa, or to the Bahía Blanca and Buenos Aires markets. Thus we have under our eyes, unexpectedly, in the north of Patagonia a transformation that occurred gradually half a century ago in all the western and north-western parts of Argentina. In its many forms it is the essential fact in the modern history of Argentine colonization. The more distant provinces are detached in succession from foreign markets, and the whole national life is being organized round the great economic focus which the region of the Pampas has become.